Handelshistorik

Horisont: 12-18 months Begrundelse: Strong cross-feed BUY consensus with 33% analyst upside and improving price momentum (258 vs 246 ten days ago); deploy freed capital into the highest-conviction blue-chip name available. === Session-analyse === [main/US] Portfolio is cash-frozen at $88.95 spendable with no US name fitting the buy window directly. The BUY WINDOW mechanism indicates a full exit of NVDA or MSFT would unlock capital. NVDA is the better exit: cross-feed consensus is SELL (1 BUY/2 SELL/3 HOLD) despite a bullish analyst price target, it's my smallest US position, and I've only held it one session without a developed long-term thesis. MSFT remains a strong long-term holding with BUY consensus, +29% unrealized gain, and continued analyst support (+13% upside to target) — no reason to sell a quality compounder just to free cash. Holding DSV.CO, NOVO-B.CO, and MSFT unchanged — MSFT remains a high-conviction long-term compounder with BUY consensus and strong gains; the Danish positions are untouched this US-only session.

Horisont: n/a - full exit Begrundelse: Full exit of small NVDA position to break the cash freeze; cross-feed consensus is SELL despite high analyst target, and position is too small to add further conviction. === Session-analyse === [main/US] Portfolio is cash-frozen at $88.95 spendable with no US name fitting the buy window directly. The BUY WINDOW mechanism indicates a full exit of NVDA or MSFT would unlock capital. NVDA is the better exit: cross-feed consensus is SELL (1 BUY/2 SELL/3 HOLD) despite a bullish analyst price target, it's my smallest US position, and I've only held it one session without a developed long-term thesis. MSFT remains a strong long-term holding with BUY consensus, +29% unrealized gain, and continued analyst support (+13% upside to target) — no reason to sell a quality compounder just to free cash. Holding DSV.CO, NOVO-B.CO, and MSFT unchanged — MSFT remains a high-conviction long-term compounder with BUY consensus and strong gains; the Danish positions are untouched this US-only session.

Horisont: 1-2 years Begrundelse: Clean uptrend over last 10 sessions (250 to 260), blue-chip pharma quality at reasonable entry after recent pullback stabilized. === Session-analyse === [main/DK] Cash is frozen at 420 DKK — below the DK fee floor for any BUY, so no new position can be opened without first unlocking capital. The BUY WINDOW explicitly lists DSV.CO and ORSTED.CO full exits as the unlock paths. DSV.CO carries the strongest cross-feed conviction of all four flagged names (analyst upside +68%, BUY consensus) and is only a small 3-share position already at a loss (-2.8%) — not a name I want to abandon on a dip. ORSTED.CO, by contrast, is a larger, newer position (11 shares, bought just yesterday) sitting flat at -0.3%, with decent but not standout analyst upside (+28%), and exiting it frees ~1,456 DKK — enough to redeploy into higher-conviction names like DSV.CO (which I'd keep) or add to NOVO-B.CO which has shown a clean uptrend. Holding DSV.CO (strongest cross-feed conviction, +68% analyst upside target) despite its small loss, and holding MSFT/NVDA positions in the US market unchanged this session. Deploying freed ORSTED.CO cash into NOVO-B.CO which shows the clearest technical momentum in the DK universe.

Horisont: n/a - full exit Begrundelse: Full exit of a newly-opened flat position to unlock trapped cash and break the cash-freeze, redeploying into stronger-conviction names. === Session-analyse === [main/DK] Cash is frozen at 420 DKK — below the DK fee floor for any BUY, so no new position can be opened without first unlocking capital. The BUY WINDOW explicitly lists DSV.CO and ORSTED.CO full exits as the unlock paths. DSV.CO carries the strongest cross-feed conviction of all four flagged names (analyst upside +68%, BUY consensus) and is only a small 3-share position already at a loss (-2.8%) — not a name I want to abandon on a dip. ORSTED.CO, by contrast, is a larger, newer position (11 shares, bought just yesterday) sitting flat at -0.3%, with decent but not standout analyst upside (+28%), and exiting it frees ~1,456 DKK — enough to redeploy into higher-conviction names like DSV.CO (which I'd keep) or add to NOVO-B.CO which has shown a clean uptrend. Holding DSV.CO (strongest cross-feed conviction, +68% analyst upside target) despite its small loss, and holding MSFT/NVDA positions in the US market unchanged this session. Deploying freed ORSTED.CO cash into NOVO-B.CO which shows the clearest technical momentum in the DK universe.

Horisont: 1-2 years Begrundelse: Strongest cross-feed BUY consensus with 43% analyst upside target, entering on recent pullback to ~$225-231 range. === Session-analyse === [main/US] Cash is essentially frozen at $46 spendable, below the fee-floor for any name except the hypothetical post-sale scenarios shown. The BUY WINDOW explicitly notes a FULL EXIT of MSFT would free ~$519, opening legal buys across the board including NVDA, AMZN, GOOGL, and AAPL. However, MSFT is my strongest winner (+29.8%) with a BUY consensus and 13% analyst upside — selling it purely to unlock cash would abandon a quality compounder for no new thesis, which violates the long-term blue-chip mandate. AMZN, by contrast, is a smaller position (+0.7%, barely a day old), and a full exit of AMZN frees ~$248, enough to open exactly 1 share of NVDA per the BUY WINDOW guidance — satisfying the cash-recycling escape hatch while preserving my best long-term holding. Holding MSFT (strong winner, BUY consensus, 13% upside), DSV.CO and ORSTED.CO (DK positions untouched this US session) — no reason to abandon quality names; freed cash redeployed into the highest-conviction cross-feed BUY signal instead.

Horisont: n/a Begrundelse: Full exit of smallest, newest position to break the cash freeze per BUY WINDOW guidance, recycling capital into higher-conviction NVDA. === Session-analyse === [main/US] Cash is essentially frozen at $46 spendable, below the fee-floor for any name except the hypothetical post-sale scenarios shown. The BUY WINDOW explicitly notes a FULL EXIT of MSFT would free ~$519, opening legal buys across the board including NVDA, AMZN, GOOGL, and AAPL. However, MSFT is my strongest winner (+29.8%) with a BUY consensus and 13% analyst upside — selling it purely to unlock cash would abandon a quality compounder for no new thesis, which violates the long-term blue-chip mandate. AMZN, by contrast, is a smaller position (+0.7%, barely a day old), and a full exit of AMZN frees ~$248, enough to open exactly 1 share of NVDA per the BUY WINDOW guidance — satisfying the cash-recycling escape hatch while preserving my best long-term holding. Holding MSFT (strong winner, BUY consensus, 13% upside), DSV.CO and ORSTED.CO (DK positions untouched this US session) — no reason to abandon quality names; freed cash redeployed into the highest-conviction cross-feed BUY signal instead.

Horisont: 1-2 years Begrundelse: Cross-feed BUY consensus with analyst target +27% upside; renewable energy utility trading near multi-month lows offers attractive entry. === Session-analyse === [main/DK] Cash is tight (1,832 DKK) and the BUY WINDOW confirms limited options. Among the legal choices, ORSTED.CO stands out with a cross-feed BUY consensus and a strong analyst-rating upside (+27%), and it fits comfortably within spendable cash at 9-13 shares. DSV.CO also has BUY consensus with impressive +67% target upside, but I already hold 3 shares at a loss and adding only 1 share (the max allowed) wouldn't meaningfully improve the position while using significant cash. CARL-B.CO has mixed signal (1 BUY, 1 HOLD) and only 2 shares fit, which is a small allocation for the capital used. Holding DSV.CO (already own 3 shares, recent realized loss, limited capital to add meaningfully), AMZN and MSFT (US holdings untouched this DK session). Deploying available cash into ORSTED.CO which has the strongest cross-feed conviction and fits the buy window without over-concentrating in any single name.

Horisont: 1-2 years Begrundelse: Strongest cross-feed conviction among affordable names post-exit (BUY consensus, analyst target +34% upside) with reasonable entry near recent range lows. === Session-analyse === [main/US] Cash is essentially frozen at ~1,186 DKK ($179), too small to buy even the cheapest US name (NVDA at $230). The BUY WINDOW analysis shows a full exit of GOOGL (single share, +2.0% P&L, modest analyst upside +26%) would unlock enough cash to open several legal buys this cycle, while MSFT (+28.5% unrealized gain, strongest position, analyst target still +13% upside with BUY consensus) is the stronger long-term holding to keep. GOOGL is a single share with a thin thesis at this point — closing it is the cleanest recycling move since it fully bypasses the fee-guard and unlocks the freeze without disturbing the better-performing MSFT position. Keeping MSFT (strong momentum, +28.5% gain, still BUY-rated with upside) and DSV.CO untouched. After the GOOGL exit and AMZN buy, remaining cash stays as reserve; no other US names fit the fee-optimal window this cycle.

Horisont: N/A - exit Begrundelse: Full exit of single-share GOOGL position to break the cash freeze and recycle capital into a higher-conviction buy this session. === Session-analyse === [main/US] Cash is essentially frozen at ~1,186 DKK ($179), too small to buy even the cheapest US name (NVDA at $230). The BUY WINDOW analysis shows a full exit of GOOGL (single share, +2.0% P&L, modest analyst upside +26%) would unlock enough cash to open several legal buys this cycle, while MSFT (+28.5% unrealized gain, strongest position, analyst target still +13% upside with BUY consensus) is the stronger long-term holding to keep. GOOGL is a single share with a thin thesis at this point — closing it is the cleanest recycling move since it fully bypasses the fee-guard and unlocks the freeze without disturbing the better-performing MSFT position. Keeping MSFT (strong momentum, +28.5% gain, still BUY-rated with upside) and DSV.CO untouched. After the GOOGL exit and AMZN buy, remaining cash stays as reserve; no other US names fit the fee-optimal window this cycle.

Horisont: 6-12 months Begrundelse: Cross-feed analyst consensus shows strong BUY with +64% target upside on DSV.CO, reinforcing conviction after yesterday's initial buy. === Session-analyse === [main/DK] Cash is frozen: only 1,183.59 DKK available and the fee floor requires 1,160 DKK spendable, which is barely not met, and the BUY WINDOW confirms no name fits with current cash. The BUY WINDOW explicitly lists that a full exit of DSV.CO or NOVO-B.CO would reopen legal buys this cycle. DSV.CO has a strong cross-feed BUY consensus (analyst target +64% upside) and was just bought yesterday, so exiting it now would abandon a fresh, well-supported thesis. NOVO-B.CO, by contrast, has no cross-feed coverage shown this cycle, is roughly flat/slightly negative since purchase, and is a smaller, well-diversified holding (5 shares) — exiting it is the lower-conviction sacrifice that frees ~1,229 DKK. Holding DSV.CO (adding), GOOGL, and MSFT unchanged — MSFT remains a strong long-term winner (+27%) and GOOGL is a small stable holding. Freed cash from the NOVO-B.CO exit is redeployed into DSV.CO per the strongest cross-feed signal this cycle.

Horisont: n/a - full exit Begrundelse: Full exit of NOVO-B.CO frees trapped cash to break the buy-freeze; position was flat/slightly negative with no fresh cross-feed conviction versus DSV.CO's strong BUY consensus. === Session-analyse === [main/DK] Cash is frozen: only 1,183.59 DKK available and the fee floor requires 1,160 DKK spendable, which is barely not met, and the BUY WINDOW confirms no name fits with current cash. The BUY WINDOW explicitly lists that a full exit of DSV.CO or NOVO-B.CO would reopen legal buys this cycle. DSV.CO has a strong cross-feed BUY consensus (analyst target +64% upside) and was just bought yesterday, so exiting it now would abandon a fresh, well-supported thesis. NOVO-B.CO, by contrast, has no cross-feed coverage shown this cycle, is roughly flat/slightly negative since purchase, and is a smaller, well-diversified holding (5 shares) — exiting it is the lower-conviction sacrifice that frees ~1,229 DKK. Holding DSV.CO (adding), GOOGL, and MSFT unchanged — MSFT remains a strong long-term winner (+27%) and GOOGL is a small stable holding. Freed cash from the NOVO-B.CO exit is redeployed into DSV.CO per the strongest cross-feed signal this cycle.

Horisont: 1-2 years Begrundelse: Strong cross-feed BUY consensus (analyst target +25% upside) and stable price action make GOOGL the best single-share deployment after unlocking NVDA proceeds. === Session-analyse === [main/US] Cash is frozen at ~412 DKK ($61 spendable) — no BUY fits under current cash per the BUY WINDOW. The portfolio is small and both MSFT and NVDA are viable full-exit candidates to unlock capital. MSFT has strong analyst consensus (BUY +0.61, +13% upside) and a solid +25.8% unrealized gain — this is a quality core holding I'd rather keep long-term. NVDA has a mixed-but-leaning-positive cross-feed signal (2 BUY, 3 HOLD, 1 SELL) with strong analyst upside (+43%), but it's a small, newer position with weaker conviction than MSFT, and it's basically flat/slightly negative since entry, making it the weaker-conviction, more expendable holding to trim for cash recycling. Holding DSV.CO, NOVO-B.CO, and MSFT as long-term blue-chip core positions — MSFT especially has strong analyst backing and a healthy gain, no reason to exit a winner. Using the NVDA full-exit to recycle trapped cash into GOOGL, the highest-conviction BUY signal available in this session's universe with only $61 of spendable cash.

Horisont: n/a - full exit Begrundelse: Full exit of small, weaker-conviction NVDA position to break the cash freeze and redeploy into higher-conviction names like GOOGL/AMZN with stronger cross-feed BUY consensus. === Session-analyse === [main/US] Cash is frozen at ~412 DKK ($61 spendable) — no BUY fits under current cash per the BUY WINDOW. The portfolio is small and both MSFT and NVDA are viable full-exit candidates to unlock capital. MSFT has strong analyst consensus (BUY +0.61, +13% upside) and a solid +25.8% unrealized gain — this is a quality core holding I'd rather keep long-term. NVDA has a mixed-but-leaning-positive cross-feed signal (2 BUY, 3 HOLD, 1 SELL) with strong analyst upside (+43%), but it's a small, newer position with weaker conviction than MSFT, and it's basically flat/slightly negative since entry, making it the weaker-conviction, more expendable holding to trim for cash recycling. Holding DSV.CO, NOVO-B.CO, and MSFT as long-term blue-chip core positions — MSFT especially has strong analyst backing and a healthy gain, no reason to exit a winner. Using the NVDA full-exit to recycle trapped cash into GOOGL, the highest-conviction BUY signal available in this session's universe with only $61 of spendable cash.

Horisont: 1-2 years Begrundelse: Blue-chip pharma pulled back ~10% from recent highs, offering a better entry into a long-term core holding. === Session-analyse === [main/DK] Cash is frozen at 502.82 DKK with no legal BUY possible in this market — the fee floor requires ~1,160 DKK and nothing in the universe fits. Per the BUY WINDOW guidance, a full exit of GN.CO would free ~3,609 DKK and immediately reopen several legal BUYs this same cycle (CARL-B.CO, COLO-B.CO, DEMANT.CO, DSV.CO, NOVO-B.CO). GN.CO is a small position (34 shares, +3.9% P&L) with only modest cross-feed conviction (analyst target +6% upside, thin margin), making it the weakest-thesis holding relative to DSV.CO's standout cross-feed signal (BUY consensus, analyst target +65% upside) and NOVO-B.CO's blue-chip quality at a pulled-back price. Selling GN.CO fully frees trapped capital to redeploy into DSV.CO and NOVO-B.CO, both with stronger cross-feed conviction and quality blue-chip characteristics, resolving the cash freeze while keeping MSFT and NVDA positions untouched for the US session.

Horisont: 1-2 years Begrundelse: Strongest cross-feed conviction (analyst target +65% upside) in a quality logistics blue-chip after a pullback from highs. === Session-analyse === [main/DK] Cash is frozen at 502.82 DKK with no legal BUY possible in this market — the fee floor requires ~1,160 DKK and nothing in the universe fits. Per the BUY WINDOW guidance, a full exit of GN.CO would free ~3,609 DKK and immediately reopen several legal BUYs this same cycle (CARL-B.CO, COLO-B.CO, DEMANT.CO, DSV.CO, NOVO-B.CO). GN.CO is a small position (34 shares, +3.9% P&L) with only modest cross-feed conviction (analyst target +6% upside, thin margin), making it the weakest-thesis holding relative to DSV.CO's standout cross-feed signal (BUY consensus, analyst target +65% upside) and NOVO-B.CO's blue-chip quality at a pulled-back price. Selling GN.CO fully frees trapped capital to redeploy into DSV.CO and NOVO-B.CO, both with stronger cross-feed conviction and quality blue-chip characteristics, resolving the cash freeze while keeping MSFT and NVDA positions untouched for the US session.

Horisont: n/a - full exit Begrundelse: Full exit of GN.CO to break the cash freeze and redeploy into higher-conviction names; weakest analyst upside (+6%) among current holdings. === Session-analyse === [main/DK] Cash is frozen at 502.82 DKK with no legal BUY possible in this market — the fee floor requires ~1,160 DKK and nothing in the universe fits. Per the BUY WINDOW guidance, a full exit of GN.CO would free ~3,609 DKK and immediately reopen several legal BUYs this same cycle (CARL-B.CO, COLO-B.CO, DEMANT.CO, DSV.CO, NOVO-B.CO). GN.CO is a small position (34 shares, +3.9% P&L) with only modest cross-feed conviction (analyst target +6% upside, thin margin), making it the weakest-thesis holding relative to DSV.CO's standout cross-feed signal (BUY consensus, analyst target +65% upside) and NOVO-B.CO's blue-chip quality at a pulled-back price. Selling GN.CO fully frees trapped capital to redeploy into DSV.CO and NOVO-B.CO, both with stronger cross-feed conviction and quality blue-chip characteristics, resolving the cash freeze while keeping MSFT and NVDA positions untouched for the US session.

Horisont: 12-18 months Begrundelse: Strong cross-feed consensus (3 BUY votes) with analyst target implying +46% upside makes NVDA the top redeployment target after freeing capital from GOOGL exit. === Session-analyse === [main/US] Cash is frozen at ~$200 spendable, below the floor for any meaningful US buy given current prices (cheapest NVDA at $232). The BUY WINDOW confirms a full exit of the 1-share GOOGL position (P&L -0.5%, no strong thesis, smallest USD exposure) would free enough capital to deploy into a higher-conviction name. NVDA has the strongest cross-feed consensus this cycle (3 BUY votes, analyst target +46% upside, HOLD unchanged elsewhere), making it the best redeployment target versus holding a stagnant single GOOGL share. Holding GN.CO (Danish position, +3.0%, no session eligibility to trade) and MSFT (+26.1%, strong long-term winner, no reason to trim a small profitable position). The GOOGL-to-NVDA rotation addresses the cash freeze while shifting toward the highest-conviction cross-feed signal available this session.

Horisont: n/a - full exit Begrundelse: Full exit of the single-share GOOGL position to break the cash freeze and redeploy into higher-conviction NVDA, since GOOGL is roughly flat and too small to trim partially. === Session-analyse === [main/US] Cash is frozen at ~$200 spendable, below the floor for any meaningful US buy given current prices (cheapest NVDA at $232). The BUY WINDOW confirms a full exit of the 1-share GOOGL position (P&L -0.5%, no strong thesis, smallest USD exposure) would free enough capital to deploy into a higher-conviction name. NVDA has the strongest cross-feed consensus this cycle (3 BUY votes, analyst target +46% upside, HOLD unchanged elsewhere), making it the best redeployment target versus holding a stagnant single GOOGL share. Holding GN.CO (Danish position, +3.0%, no session eligibility to trade) and MSFT (+26.1%, strong long-term winner, no reason to trim a small profitable position). The GOOGL-to-NVDA rotation addresses the cash freeze while shifting toward the highest-conviction cross-feed signal available this session.

Horisont: n/a - full exit Begrundelse: Full exit of small, mixed-signal DSV position to break the cash freeze and redeploy into higher-conviction GN.CO. === Session-analyse === [main/DK] Cash is frozen at 146 DKK — no BUY fits without an exit. The BUY WINDOW analysis shows a full exit of DSV.CO (1 share, ~1,175 DKK back) would open several legal buys including GN.CO (12 shares) which has the strongest cross-feed conviction (BUY consensus, +7% momentum, range_pos 85%). DSV.CO is only a single share (below the 6-share/1,160 DKK partial-trim floor anyway) and mixed signal (1 buy/1 sell), so a full exit is clean and fee-guard-exempt, while GN.CO shows unambiguous 2/2 BUY consensus with strong momentum — a good redeployment target and it also adds to an already-profitable position. Holding GOOGL and MSFT unchanged (US session not active). Freed capital from DSV exit redeployed into GN.CO per BUY WINDOW guidance, keeping remaining cash buffer for fees.

Horisont: 1-2 years Begrundelse: Strong analyst consensus (BUY, +25% target upside) and broad cross-feed agreement support adding a quality mega-cap at a reasonable valuation given limited cash. === Session-analyse === [main/US] Cash is tight (~$367 spendable) but the BUY WINDOW confirms legal trades exist, so no freeze here. Among the cross-feed BUY consensus names, GOOGL and AMZN stand out with strong analyst upside (+25% and +32% respectively) and broad multi-feed agreement, while NVDA carries a SELL consensus despite one bullish analyst outlier — skip that. Given the tiny cash pool, I can only afford one meaningful share purchase without breaching the fee-guard on a second trade, so I'll concentrate on the single highest-conviction name. Cash is very limited this cycle; after the GOOGL buy there won't be room for a second qualifying US trade. Holding MSFT (strong analyst upside, +28.5% unrealized gain) and existing DK positions (DSV.CO, GN.CO) for the DK session — no changes warranted there today given this is a US-only session.

Horisont: 1-2 years Begrundelse: Cross-feed consensus BUY with strong analyst target upside (+66%) on DSV; logistics blue-chip trading below recent highs, small starter position given limited freed cash. === Session-analyse === [main/DK] Cash is frozen at 1,124.91 DKK — no BUY clears the 1,160 DKK fee floor with current spendable cash. Per the cash-recycling protocol, I need to free capital via a SELL in this same basket so the BUYs can execute after. My smallest/weakest-thesis position by absolute conviction is NOVO-B.CO: only +1.1% P&L, and while it's not screened in the cross-feed block this cycle, GN.CO is actually flagged BUY by analyst consensus (target +8% upside) and I already hold a large GN.CO position bought yesterday, so trimming NOVO-B (10 shares, holding since inception, no fresh catalyst, momentum weakening as price has drifted down from 280 to 254 over 10 days) is the more sensible rotation than cutting a stock with a live BUY signal. A full exit of the 10 NOVO-B.CO shares at ~254.35 DKK = ~2,543 DKK, well above the fee floor and bypasses the fee-guard anyway as a full exit, freeing enough cash to redeploy into a stronger cross-feed-endorsed name. Exiting NOVO-B fully to unlock trapped cash and redeploy into DSV.CO and GN.CO, both flagged BUY by cross-feed analyst consensus with meaningful upside. Keeping MSFT and existing GN.CO shares as long-term holds; MAERSK, CARL-B, VWS, DEMANT, COLO-B, ORSTED, PNDORA remain on the sidelines this cycle absent fresh conviction or available cash beyond the basket above.

Horisont: n/a - exit Begrundelse: Full exit of stale, small NOVO-B position (weak momentum, no cross-feed BUY signal this cycle) to unfreeze cash and re-open the BUY window this cycle. === Session-analyse === [main/DK] Cash is frozen at 1,124.91 DKK — no BUY clears the 1,160 DKK fee floor with current spendable cash. Per the cash-recycling protocol, I need to free capital via a SELL in this same basket so the BUYs can execute after. My smallest/weakest-thesis position by absolute conviction is NOVO-B.CO: only +1.1% P&L, and while it's not screened in the cross-feed block this cycle, GN.CO is actually flagged BUY by analyst consensus (target +8% upside) and I already hold a large GN.CO position bought yesterday, so trimming NOVO-B (10 shares, holding since inception, no fresh catalyst, momentum weakening as price has drifted down from 280 to 254 over 10 days) is the more sensible rotation than cutting a stock with a live BUY signal. A full exit of the 10 NOVO-B.CO shares at ~254.35 DKK = ~2,543 DKK, well above the fee floor and bypasses the fee-guard anyway as a full exit, freeing enough cash to redeploy into a stronger cross-feed-endorsed name. Exiting NOVO-B fully to unlock trapped cash and redeploy into DSV.CO and GN.CO, both flagged BUY by cross-feed analyst consensus with meaningful upside. Keeping MSFT and existing GN.CO shares as long-term holds; MAERSK, CARL-B, VWS, DEMANT, COLO-B, ORSTED, PNDORA remain on the sidelines this cycle absent fresh conviction or available cash beyond the basket above.

Horisont: 6-12 months Begrundelse: Strong 5-day momentum (+13.3%) and unanimous cross-feed BUY consensus align with an existing profitable position, entry near range highs confirms breakout strength. === Session-analyse === [main/DK] Cash is limited but not frozen (2,297 DKK spendable, and BUY WINDOW confirms legal trades exist). Cross-feed screening shows GN.CO with strong momentum (+13.3% 5d, at range highs) and unanimous BUY consensus, and I already hold GN.CO profitably - a modest add fits with existing thesis strength. ORSTED.CO also shows analyst BUY conviction with 25% upside to target, and DSV.CO shows the strongest analyst upside (+59%) though I can only afford 1 share which is below the fee floor, so I'll skip DSV this cycle. CARL-B.CO also has decent analyst support but GN.CO's momentum plus existing position edge gives it priority. Deployed available cash into GN.CO which has the strongest cross-feed momentum signal and fits the fee floor cleanly. Holding NOVO-B.CO (long-term thesis intact despite recent pullback) and MSFT (strong US winner, US session only). Keeping remaining cash buffer for fees.

Horisont: n/a - full exit Begrundelse: Full exit of the smallest, least-differentiated US position to break the cash freeze and redeploy into higher-conviction names like NVDA or AMZN next cycle. === Session-analyse === [main/US] Cash is essentially frozen at ~$8.70 spendable, well below the $40 fee-floor for any US buy — no ticker fits regardless of conviction. Per the cash-recycling protocol, a full exit is the correct unlock, and GOOGL is the natural candidate: it's my smallest US position (1 share, ~2,270 DKK), only marginally profitable (+1.3%), and while the cross-feed shows BUY consensus, MSFT's stronger momentum (+26% P&L, similar analyst upside) and NVDA's higher-conviction score make GOOGL the weakest single-share holding to recycle. Closing it fully bypasses the fee-guard and frees roughly $346 (~2,270 DKK) of trapped capital. After the GOOGL full exit frees ~$346 USD, I'll hold the proceeds as cash this cycle rather than immediately rebuying — the portfolio is already well-diversified across MSFT, NOVO-B, GN, and this freed capital can be deployed with better sizing (e.g., a full NVDA or AMZN share) in the next US session once the cash has settled. GN.CO and NOVO-B.CO remain solid long-term blue-chip holds with positive momentum; MSFT's strong +26% gain and BUY consensus warrant continued holding rather than trimming.

Horisont: 6-12 months Begrundelse: Cross-feed consensus BUY (2/2) plus strong 5-day momentum off a range high, adding to an already profitable core position while cash allows. === Session-analyse === [main/DK] Cash is tight (1,266 DKK) with only two legal buys available: 1 share of DSV.CO or up to 11 shares of GN.CO. GN.CO already shows strong cross-feed conviction (2/2 BUY, yfinance +0.44 on a sharp 5-day rally) and I already hold 12 shares with a solid +10.9% gain — adding more chases momentum into an already large relative position and risks breaching sizing discipline. DSV.CO has the strongest fundamental signal here: analyst consensus BUY with +62% upside to target, a quality logistics blue-chip, but only 1 share fits the budget which is a marginal fee-guard trade (1,329 DKK vs 1,237 spendable — actually check: 1 share DSV at ~1,329 exceeds spendable, so DSV is not actually affordable at current price shown in the window despite the stated limit; I'll defer to GN.CO which is confirmed affordable). Holding NOVO-B.CO, GOOGL, and MSFT unchanged — all have positive theses intact (NOVO-B analyst target +19% upside, MSFT strong long-term winner). Deploying remaining cash into GN.CO, the only fully affordable and cross-feed-confirmed buy this session, rather than sitting in a frozen cash position.

[stop-loss triggered at $111.37 — broker-fill $111.52]

Horisont: 1-2 weeks Begrundelse: Fresh 318-comment WSB DD post ('follow the smart money') on UBER, risk-on tape (QQQ +0.76%), tight gulv-stop of 4.1% fits the risk budget cleanly. === Session-analyse === [saxomoon/US] MU is the day's clearest WSB catalyst — reports today, JPMorgan calling for beat-and-raise, a $50K earnings YOLO post, and the interpreted signal shows net-sentiment +0.94 with actual disclosed positions (calls+shares), not drive-by hype. But MU itself is untradeable at this account size (>$1000/share, min-entry would blow the $270 budget and the position cap), so I can't act on that thesis directly. UBER has a fresh 318-comment DD post ('follow the smart money') plus a calm gulv-stop (4.1%) that actually fits the risk budget — but it's flagged CAP-FULL... wait, checking again, UBER is NOT in the cap-full list (only HOOD/RDDT are), so it's tradeable. Cross-feed doesn't cover UBER explicitly but the WSB DD is fresh and RISK-ON regime plus VIX calm at 16 supports leaning in on a reasonable size. HOOD and RDDT are both CAP-FULL — no adding. RDDT has a fresh WSB post (net-sentiment +1.00, fundamental thesis, disclosed calls) plus cross-feed consensus BUY (analyst target +47% upside) — this supports holding, not cutting, despite the -3.8% P&L; rule #4's cross-feed exception applies. HOOD is red on both WSB attention (not mentioned) and overnight gap (-1.8%), down -8.3% — but it's a small single-share position where a full exit would only marginally help and no cross-feed BUY line was surfaced for HOOD in this session's block, meaning weaker support than RDDT. I'll tighten stops on both to protect capital given no fresh bullish catalyst. Given cash is thin (~$270 spendable) and both current holdings are cap-full, the only actionable slot is a fresh UBER position sized to the risk budget (~2 shares, gulv-stop 4.1%). MU thesis is real but inaccessible; I'll skip forcing it into an untradeable name. No crypto-mining exposure add given the -6.6% 5d tape-fade and prior -620 DKK theme history (theme-facit discipline holds). HOOD and RDDT are both position-cap-full so no adds are possible regardless of conviction. RDDT keeps its cross-feed BUY consensus (analyst +47% target upside) and a fresh WSB post with disclosed calls-positioning, so I'm holding per rule #4's exception rather than cutting the -3.8% drawdown. HOOD lacks that cross-feed support and is down -8.3% with no fresh WSB mention, but a full exit on a single share isn't worth the churn versus tightening the stop close to current levels to cap further downside. MU is the session's strongest WSB catalyst (earnings today, JPMorgan beat-and-raise call, $50K YOLO post) but is priced entirely out of reach for this account's budget — noted for tracking, not actionable. Sitting mostly in cash given thin spendable capital and two cap-full core positions; will reassess at 14:00 for any post-MU-earnings reaction that opens new opportunities in adjacent semiconductor names.

Horisont: 1-2 weeks Begrundelse: WSB YOLO post ($146K RDDT Jan2028 calls, net-sentiment +0.75) plus cross-feed consensus BUY (analyst target +40% upside) — exit if momentum thesis fades or price breaks below the $140 shelf. === Session-analyse === [saxomoon/US] Small book ($553), mostly cash (79%), one small HOOD position down -4.5%. WSB is dominated by NKE (top post, 15k score) but the actual sentiment reading is negative loss-porn/joke content ('The only way to rescue Nike'), not a bullish catalyst — the interpreted signal is net -0.25 with 100% loss-porn framing. ORCL has 2 mentions and interpreted +0.50 but only 1 post, thin sample = near-zero signal per rule #15, and the force-majeure headline is a real overhang even though cross-feed doesn't cover it directly. RDDT has a genuine YOLO ($146K calls post) plus cross-feed consensus BUY (analyst target +40% upside) — best combo of WSB hype + cross-feed agreement I see today, but gulv-stop is 8.5% and budget only allows 1 share (~$153), which clears the $100 min-trade floor barely and is a real dollar position at this price tier (rule #3 exception for ≥$50 names). META also shows interpreted +0.92 with disclosed 3-share position and cross-feed consensus BUY, but it's UNTRADEABLE at min size per the params (too expensive even at 1 share within budget) — skip. HOOD is CAP-FULL, can't add; it's down -4.5% but only mentioned in comment-flow, not fresh WSB hype, and no cross-feed shown for it directly — I'll leave the stop where it is since the daily noise band (±3.3%) means current 6.9% stop is already outside noise, no need to tighten yet given it's a small loss, not a winner to ratchet. Given QQQ is neutral/flat, VIX calm (14.21), and most of the loud posts today are jokes/memes (NKE rescue meme, Bessent meme, Bears meme) rather than fresh actionable catalysts, I'm taking one modest, cross-feed-supported bet: RDDT at 1 share with a stop anchored to the gulv-stop floor. Everything else — ORCL (thin/negative catalyst), NKE (loss-porn meme), AMZN (mixed/ironic, tiny sample) — fails the signal-quality bar. Sitting mostly in cash remains correct; this is a small nudge, not a big swing, consistent with the tiny book size and the need to avoid manufacturing activity (rule #6). HOOD stays as-is — CAP-FULL means no add possible, and it's down only -4.5% with the daily noise band (±3.3%) already narrower than the current 6.9% stop, so no ratchet needed on a small loser. NKE's top post is loss-porn/meme framing (interpreted net -0.25, 100% loss-porn) despite the huge score, so it fails the signal-quality bar (rule #15) even though it's technically trending. ORCL's single post is too thin a sample (rule #15: 1-2 posts = zero signal) and carries a real negative catalyst (force-majeure headline) working against it. AMZN's interpreted read is mixed/ironic with a tiny sample too. Cash stays the default given QQQ neutral, VIX calm but no name clearing both the WSB-freshness and cross-feed-agreement bar convincingly enough for more than the one RDDT nibble.

[stop-loss triggered at $22.51 — broker-fill $22.52]

Horisont: 2-4 weeks Begrundelse: WSB post citing $300-by-2027 HOOD target on BTC/SEC catalyst, cross-feed shows no SELL votes and RISK-ON tape; sized to 1 share given last-cycle overnight risk and floor-stop constraints. === Session-analyse === [saxomoon/US] Last cycle of the day, RISK-ON (QQQ +2.81%), VIX calm at 15.4, crypto Fear&Greed at 70 (Greed). RIOT is our only holding, up +14% and crypto_mining tape is ripping (+17.5% 5d basket). Rule #2/#13 says let it run, don't trim — ratchet the stop instead. RIOT's floor-stop from the server sits at 11.6% under price (avg dagsmove ±4.3%), and current resting stop is already 22.10 (9.4% under, tighter than floor doesn't apply since floor is a minimum not max) — actually the floor requires stop >= 11.6%, so current 9.4% stop is inside the noise floor per the server's own gulv-stop rule. I'll widen protection isn't allowed (raise-only, and 11.6% is WIDER than current 9.4%, which the engine would reject as a widen). So I leave RIOT's stop as-is this cycle since I can't legally tighten it below the noise floor, and the ladder (+15% -> 12% trail) is roughly where the floor already sits — no action needed, current stop stands. On new entries: it's cycle 4/4, last cycle before close, so any BUY carries overnight risk with a weakened stop (44% of stop-fires historically hit in the first 30min after open). AMD/META/MU/INTC are the loudest WSB names today (AMD hit $1T, META getting mixed loss-porn love, MU spreads gain post) but ALL are on the UNTRADEABLE-at-min-size list (AMD, META, MU) or have a floor-stop too wide for budget (INTC gulv 8.9%, max 1 share ~$121 — technically tradeable but marginal). HOOD is trending (+2 mentions, NEW/RISING) with a clean cross-feed absence of SELL votes and a WSB post citing $300 2027 target, floor-stop 11.2%, max 1 share (~$124) — that's a real dollar position at $124/share (>=$50 threshold allows 1-2 share entries) and clears the $100 min-trade floor. Risk-sizing at 3% (~$17) against an 11.2% floor stop only supports ~$150 notional, so 1 share (~$124) fits within budget and floor. However, given it's the LAST cycle of the day, a fresh 1-share HOOD position needs a reason to hold overnight — the WSB catalyst ($300 target by 2027) is a slow-burn thesis, not an urgent overnight one, and the position size (1 share, ~$124) is small enough that overnight gap risk is limited in dollar terms. I'll take a small starter HOOD position with a wide stop reflecting the floor, sized down deliberately per rule #19. No other name clears both the fee-guard floor and the risk-budget cleanly at this session's cash level. Sitting mostly in cash/RIOT into the close with a small HOOD nibble is the disciplined call — RIOT is already my crypto_mining exposure at 21%, well under the 50% cap, and adding more mining here into a Monday close after a big already-run 5-day tape (+17.5%) would be chasing, not investing. RIOT's current stop (9.4% under price) sits inside the server's own noise-floor (11.6% gulv-stop for RIOT), so I cannot legally raise it further without violating the raise-only/floor logic — leaving it unchanged rather than manufacturing a move. The loudest WSB names today (AMD, META, MU, INTC) are either UNTRADEABLE at this account's min size or only offer 1-share entries at marginal risk-budget fit; AMD/META also show mixed-to-bearish INTERPRETED signal (loss-porn 50-60%, META net-sentiment near zero with 4 bear posts) which per rule #15 lowers conviction rather than raising it. Crypto_mining tape has already run +17.5% in 5 days — chasing more RIOT/mining exposure into the close would be adding to a stretched move, not fresh conviction. Staying mostly in cash with the existing RIOT position and one small HOOD nibble reflects the last-cycle discipline (#19): don't force size into positions that need to survive an unpaid overnight session on thin catalysts.

[stop-loss triggered at $10.10 — broker-fill $10.01]

Horisont: 1-3 weeks Begrundelse: Cross-feed consensus BUY (analyst target +41% upside) on a cheap name gapping up overnight (+3.3%) in a risk-on tape, uncorrelated to my crypto_mining/ai_infra overweight — exit if the gap fades and it drops back below $10. === Session-analyse === [saxomoon/US] WSB's hot list today is dominated by macro loss-porn (bond market chaos, CPI, treasury debt repurchases) rather than a fresh single-name squeeze thesis — the actual meme content (AAPL calls meme, oil superglut, MSTR Jordans joke) is more comedic than a directional catalyst. That said, the cross-feed screen is unusually constructive: MARA, RIOT, CLSK, WULF all show consensus BUY with short-interest and technical strength (RIOT +12.4% 5d, CLSK +13.0% 5d, WULF +120% analyst upside), and QQQ/SPY are green (+0.93%/+0.99%) on a RISK-ON day — exactly the regime where WSB momentum names pay. My own theme-facit shows crypto_mining has been a net loser (-717 DKK over 9 exits) which is a real scar, but that history is on MARA/RIOT/CLSK/WULF specifically during choppier tape — today's setup is different: multiple names in the group are simultaneously breaking out with short-interest fuel (MARA 32.7% short, ratio 2.4) and analyst upside, which is a materially stronger multi-feed confirmation than prior attempts. Given the crypto_mining scar tissue, I'll size small and concentrated rather than spreading across all four miners (that would just be one correlated bet wearing four costumes per the theme-facit rule). RIOT has the cleanest signal: 4 BUY votes, strong yfinance momentum reading (range_pos 71%, +12.4% 5d) and it's not one of my repeat-loser names in the current session's overnight data. I'll also take a small NOK position — cross-feed BUY (analyst target +41% upside), cheap enough for full share-count minimums, uncorrelated to the crypto_mining/ai_infra themes I'm already scarred on, giving portfolio diversification. Cash is tiny (~$556 spendable) so I'm keeping this to 2 disciplined, fee-efficient entries rather than forcing 4-8 trades — rule #6 applies: don't manufacture activity, and with such a small book, concentration in the best 2 ideas beats diluting into weak ones. GME's insider buying (edgar +0.85, $21.7M in insider purchases) is tempting but GME is not a fresh trending name today (mentions declining) and the crypto-mining allocation already uses my risk budget for meme/momentum plays this cycle. VIX is calm at 16.46 (not elevated), Fear&Greed is 56 (neutral-greed, no extreme), and QQQ/SPY are both green — this is the risk-on regime rule #11 says to lean into. I'm avoiding SOUN (just stopped out -10.2%, in cooldown per rule #8) and PLTR (recent exit, needs >=0.7 conviction + 2 cross-feed votes — PLTR cross-feed isn't even listed here with dual-vote confirmation, so pass). Sitting mostly in cash beyond the two small entries above — WSB's hot list today is macro-loss-porn (bond chaos, CPI, treasury repurchases) rather than a single-name squeeze thesis, so I'm not forcing size into names just because they're mentioned (AAPL, TSLA, MP mention counts are macro-driven commentary, not a fresh WSB catalyst on the stock itself). PLTR and SOUN are both in re-entry cooldown from recent exits and don't clear the required 2-cross-feed-vote bar cleanly enough to justify re-entry today. Crypto_mining as a theme has a real -717 DKK scar over 9 prior exits, so I deliberately only took one name (RIOT, the strongest single signal) rather than spreading across MARA/CLSK/WULF too, keeping the correlated bet-sizing sane. Will reassess at 11:30 ET as the sub fully wakes up.

Horisont: 1-2 weeks Begrundelse: Cross-feed consensus BUY (4-1) with strong 5d momentum (+12.4%, range_pos 71%) and high short-interest fuel; risk-on QQQ tape favors miners — exit if RIOT loses the $19 shelf or WSB/cross-feed flips. === Session-analyse === [saxomoon/US] WSB's hot list today is dominated by macro loss-porn (bond market chaos, CPI, treasury debt repurchases) rather than a fresh single-name squeeze thesis — the actual meme content (AAPL calls meme, oil superglut, MSTR Jordans joke) is more comedic than a directional catalyst. That said, the cross-feed screen is unusually constructive: MARA, RIOT, CLSK, WULF all show consensus BUY with short-interest and technical strength (RIOT +12.4% 5d, CLSK +13.0% 5d, WULF +120% analyst upside), and QQQ/SPY are green (+0.93%/+0.99%) on a RISK-ON day — exactly the regime where WSB momentum names pay. My own theme-facit shows crypto_mining has been a net loser (-717 DKK over 9 exits) which is a real scar, but that history is on MARA/RIOT/CLSK/WULF specifically during choppier tape — today's setup is different: multiple names in the group are simultaneously breaking out with short-interest fuel (MARA 32.7% short, ratio 2.4) and analyst upside, which is a materially stronger multi-feed confirmation than prior attempts. Given the crypto_mining scar tissue, I'll size small and concentrated rather than spreading across all four miners (that would just be one correlated bet wearing four costumes per the theme-facit rule). RIOT has the cleanest signal: 4 BUY votes, strong yfinance momentum reading (range_pos 71%, +12.4% 5d) and it's not one of my repeat-loser names in the current session's overnight data. I'll also take a small NOK position — cross-feed BUY (analyst target +41% upside), cheap enough for full share-count minimums, uncorrelated to the crypto_mining/ai_infra themes I'm already scarred on, giving portfolio diversification. Cash is tiny (~$556 spendable) so I'm keeping this to 2 disciplined, fee-efficient entries rather than forcing 4-8 trades — rule #6 applies: don't manufacture activity, and with such a small book, concentration in the best 2 ideas beats diluting into weak ones. GME's insider buying (edgar +0.85, $21.7M in insider purchases) is tempting but GME is not a fresh trending name today (mentions declining) and the crypto-mining allocation already uses my risk budget for meme/momentum plays this cycle. VIX is calm at 16.46 (not elevated), Fear&Greed is 56 (neutral-greed, no extreme), and QQQ/SPY are both green — this is the risk-on regime rule #11 says to lean into. I'm avoiding SOUN (just stopped out -10.2%, in cooldown per rule #8) and PLTR (recent exit, needs >=0.7 conviction + 2 cross-feed votes — PLTR cross-feed isn't even listed here with dual-vote confirmation, so pass). Sitting mostly in cash beyond the two small entries above — WSB's hot list today is macro-loss-porn (bond chaos, CPI, treasury repurchases) rather than a single-name squeeze thesis, so I'm not forcing size into names just because they're mentioned (AAPL, TSLA, MP mention counts are macro-driven commentary, not a fresh WSB catalyst on the stock itself). PLTR and SOUN are both in re-entry cooldown from recent exits and don't clear the required 2-cross-feed-vote bar cleanly enough to justify re-entry today. Crypto_mining as a theme has a real -717 DKK scar over 9 prior exits, so I deliberately only took one name (RIOT, the strongest single signal) rather than spreading across MARA/CLSK/WULF too, keeping the correlated bet-sizing sane. Will reassess at 11:30 ET as the sub fully wakes up.

[stop-loss triggered at $6.33 — broker-fill $6.34]

Horisont: 1-3 weeks Begrundelse: Cross-feed consensus BUY driven by extreme 39.9% short-interest-of-float (ratio 4.4) — squeeze fuel independent of today's thin WSB chatter; exit if short interest unwinds or price breaks below the recent 6.78 range floor. === Session-analyse === [saxomoon/US] WSB's hot list today is almost entirely consumed by AVGO's post-earnings guidance drop (posts #4,7,11,13) and macro-fear megathread noise (50% bull, 'fear' undertone dominated by rates/tariffs chatter) — none of which is actionable: AVGO is explicitly untradeable at this account's risk budget, and chasing a -5.7% overnight gap into a crowded, SPIKE-CHRONIC name is exactly the setup rule #1 warns against (no conviction bump, tighten-not-add). IREN shows a fresh/rising WSB mention plus cross-feed BUY consensus (analyst target +97% upside) but the strategy params flag it explicitly UNTRADEABLE at this risk budget (13.2% floor-stop only allows 2 shares vs 3 minimum) — skipping despite the pull. The one name clearing the bar is SOUN: cross-feed consensus BUY led by short_interest at an extreme 39.9% of float (ratio 4.4) — genuine squeeze fuel — with a calm, tight daily range (6.78-7.32 over 10 days) that fits comfortably inside the risk budget without needing a wide stop. It's not in today's top WSB mentions, so this is a cross-feed/short-interest-led trade, not a hype chase, sized small and given a stop anchored just under the recent range floor. RISK-ON tape (QQQ +0.47%, VIX 16.3 calm) supports leaning in modestly rather than sitting fully passive, but with genuinely thin fresh catalyst evidence elsewhere, cash preservation is the right call for the rest of the book — no reason to force trades into AVGO's crowded earnings drama or space-theme names lacking a name-specific catalyst beyond a generic fedreg mention. Sitting mostly in cash this cycle. Today's WSB hot list is dominated by AVGO's earnings/guidance drama, which is untradeable at this account's min-size math and is a SPIKE-CHRONIC crowded name post-drop — not a fresh setup. IREN has a rising mention and cross-feed BUY backing but is explicitly flagged untradeable at the current risk budget (floor-stop too wide for min shares). Space names (RKLB/ASTS) ride only a generic fedreg mention with no name-specific WSB catalyst and a weak 30d theme tape (-10.9%). No name in the RECENT EXITS list (PLTR, MRNA) shows the required 2+ fresh cross-feed BUY votes plus changed thesis to justify re-entry. Preserving capital for a session with clearer, larger WSB conviction rather than forcing size into thin or blocked setups.

[stop-loss triggered at $171.31 — broker-fill $171.30]

[stop-loss triggered at $142.78 — broker-fill $142.64]

Horisont: 3-7 days Begrundelse: WSB's #1 gain-porn story today (cancer vaccine breakthrough, $500k overnight thread, $130k-in-10s thread) with cross-feed consensus BUY on the +9% overnight gap and RISK-ON tape — but chasing a post-gap biotech binary event, so sized to 1 share with a wide stop given the stock's huge realized volatility (63→174→139 over 10 sessions). === Session-analyse === [saxomoon/US] MRNA is the clear WSB story today — cancer vaccine breakthrough, +9% overnight gap, $500k gain-porn thread, $130k-in-10-seconds thread, huge megathread comment volume (40✓), and cross-feed consensus BUY (yfinance +1.00 on the 5d move). But the INTERPRETED read is mixed-to-cautionary: one loss-porn post (holder wants to exit), thin sample, and it's not in the TRENDABLE UNIVERSE per the trading universe list — MRNA IS in the tradeable universe though, so that's not a blocker. Given RISK-ON regime (QQQ +0.87%), calm VIX (15.13), and a name up huge already off a real fundamental catalyst (Phase 3 cancer vaccine data, not just meme hype), this is a rare case of a genuine news-driven mover the WSB crowd is chasing correctly. However, entering AFTER a +9% gap into a name already up ~50% this month is real chase risk per rule #16 — I'll size modestly and use a wider stop to respect the volatility (biotech binary-event stocks gap hard), not full conviction size. PLTR is my only current holding, +14.4%, sub-theme (ai_infra/defense_tech) already AT CAP so no adds regardless of conviction — the INTERPRETED read on PLTR is actually net-bearish (-1.50, puts thesis) which reinforces NOT adding. Per the trailing-stop ladder, +14.4% crosses toward the +15% tier (trail ~12% under price) — I'll tighten the stop from its current 7.4%-under level... wait, current stop is already tighter (7.4% under) than the 12% ladder guidance would imply, so I leave it as-is since the engine is raise-only and the current stop is already tighter than what the ladder step would set. Actually checking: ladder at +15% suggests 12% trail, but current stop is 7.4% under — that's already tighter than 12%, so no update needed, it's already better protected. No other WSB names clear the bar today — NVDA has earnings tomorrow (avoid fresh entries into print) and INTERPRETED read is bearish (puts, AI-hype-dead thesis) despite cross-feed mixed. SOXL is bleeding (-9% 5d, -18.7% 30d) with a loss-porn post — pure fade. WEN's short-squeeze story (43% short float) is interesting but only 1 mention and no real catalyst beyond a CEO quote — passing given fee-drag concerns on cheap-ish names. Sizing MRNA conservatively given the parabolic gap and biotech binary-event volatility (10-day closes show a wild recent range from ~$63 to $174 to now $138 — huge realized vol, needs a wide stop per volatility anchor rule). PLTR is capped at max sub_theme concentration (ai_infra/defense_tech both AT CAP) so no adds are possible regardless of the +2.1% overnight pop — and its current stop (7.4% under price) is already tighter than the ladder's +15%-tier guidance (12% trail), so no stop update needed this cycle. Rest of the WSB feed is meme/discussion noise (Buffett memes, Wendy's CEO quotes, Walmart) with no tradeable edge; NVDA and MRVL both report earnings within 48 hours so fresh entries there are avoided per the earnings-watch rule, and NVDA's INTERPRETED read is net-bearish (puts, AI-hype-dead thesis) despite a mixed cross-feed. SOXL and crypto-mining names (MARA/RIOT/CLSK/WULF) carry a documented -732 DKK 30-day loss history in this exact theme with no new external catalyst — theme-facit re-entry bar not cleared, staying out. Sitting mostly in cash preserves the $413 spendable for a cleaner setup next session.

[stop-loss triggered at $43.02 — broker-fill $43.02]

[stop-loss triggered at $164.28 — broker-fill $164.26]

Horisont: 1-2 weeks Begrundelse: S&P 500 inclusion Aug 18 is a concrete catalyst (interpreted net +1.50, 2/2 high-conviction posts) with independent cross-feed consensus BUY (4-0-1); sizing small given the +13% overnight gap already priced in much of the pop. === Session-analyse === [saxomoon/US] RDDT is the clearest setup on the board today: S&P 500 inclusion catalyst (Aug 18), interpreted net-sentiment +1.50 with 2/2 high-conviction posts citing a concrete catalyst, and the cross-feed consensus is a clean BUY (4 BUY/0 SELL/1 HOLD) — genuinely independent corroboration beyond just the Reddit-derived wsb vote. It's not in the raw trending-tickers title list, but it's front-page news (#11, ▲2,518) and the interpreted block flags it explicitly. It gapped +13% overnight into the open which is chase-risk, but a real fundamental catalyst (index inclusion forces passive buying next week) justifies paying up modestly with a wider stop rather than skipping it entirely. IREN also screens well (short 28.4% of float, cross-feed consensus BUY, interpreted net +1.50 fundamental thesis on the MSFT 50MW delivery) but it's pulled back hard from highs (45.93→38.89 over 10 days) — I'll pass this cycle given limited cash and prefer concentrating in the stronger single idea. My IONQ (+16.7%) and PLTR (+13.8%) positions are working — both cross-feed consensus BUY, both winners I should let run per rule #2 rather than trim. IONQ's ticker-facit is clean (no prior losses) and analyst_ratings backs it strongly (+50% upside target). PLTR is at position-cap so no add possible anyway, but I'll tighten both trailing stops to lock in gains per the ladder (IONQ +16.7% → trail ~12%; PLTR +13.8% → trail ~12%), respecting the gap-floor (avg dagsmove ~2.9% and ~2.3% respectively, well under the ladder levels so no floor conflict). Cash is thin (~$303 spendable) and QQQ/SPY are dead neutral today (+0.05%/+0.01%) — no regime tailwind either way. VIX is calm at 14.55 so no vol-based caution needed. I'm avoiding RIOT (recent stop-out loss, no independent 2-feed re-entry corroboration beyond wsb/short_interest overlap concerns) and skipping SMR/OKLO/other nuclear/quantum names to avoid over-concentrating further in quantum (already 44%) or adding new themes without conviction. One focused RDDT buy is the disciplined play this cycle. IONQ and PLTR are winners running per plan — letting them ride with tightened trailing stops rather than trimming. PLTR is CAP-FULL so no further add possible regardless. Cash is limited (~$303 spendable) so only one high-conviction fresh idea (RDDT) makes sense this cycle; RIOT stays excluded per re-entry cooldown lacking 2-independent-feed corroboration, and I'm not chasing quantum/nuclear names (OKLO/SMR/LEU) to avoid stacking further into an already-44%-concentrated quantum theme.

[stop-loss triggered at $19.26 — broker-fill $19.26]

Horisont: 1-3 weeks Begrundelse: WSB post on Anthropic's $9B cloud deal with Riot Platforms plus cross-feed consensus BUY (analyst target +53% upside, short_interest fuel) — exit if the deal narrative fades or RIOT breaks below $19.40 support. === Session-analyse === [saxomoon/US] Small cash pile (~$320 spendable) and both existing positions already green — IONQ +9.7%, PLTR +8.9% — with PLTR at cap so no add there. WSB's own thread on PLTR (Spider-Man/Palantir post, +34% move) is real but the position is already at the server's CAP-FULL limit, so this is a hold-and-trail situation, not an add. RIOT is the one fresh trending idea today: NEW tag on the WSB extraction (Anthropic's $9B cloud deal headline, ▲762), and it clears cross-feed with 4 BUY votes including analyst_ratings +0.87 (target +53% upside) — genuinely independent corroboration beyond just Reddit hype. Regime is RISK-ON (QQQ +0.87%), VIX calm at 15.46, so leaning into the best-supported fresh name fits the tilt. CRWV and NBIS both gapped huge overnight (+19%, +17%) but NBIS reports earnings TODAY (0 days out) — chasing a pre-print gap into an earnings print is exactly the kind of coin-flip the rules warn against, and CRWV isn't even in the tradeable universe. Passing on both despite the momentum. MARA/CLSK/WULF all show strong short-interest-driven consensus BUY but I have prior losing round-trips on MARA (-97 DKK), CLSK (-106 DKK), WULF (-63 DKK) — TICKER-FACIT says these need a substantially changed thesis, and 'short interest is still high' isn't new information, it was already true when I lost money on them. Staying out. For the holdings: IONQ is up +9.7% with an already-tight 5.1% stop under price — that's inside the ladder guidance (not yet at +15% threshold), so leaving it as-is rather than forcing a ratchet. PLTR similarly at +8.9%, stop already 5.8% under price, close to the avg dagsmove floor of 5.4% — tightening further risks a noise wick-out, so no change there either. Sizing RIOT modestly (~$220, under the $221 single-position cap) with a stop anchored below its recent higher-low around $19.40, giving room for the ±normal swing while protecting against a thesis break if the Anthropic deal narrative fades. Keeping remaining cash in reserve — no forced churn given the FEE-DRAG track record (14d: 8 fills costing 0.75% of turnover) reminding me low-edge scalps net to zero after fees. IONQ and PLTR are both modest winners (+9.7%, +8.9%) but neither has crossed the +15% ladder threshold yet, and their current stops already sit near or inside the avg-dagsmove gap-floor (IONQ 5.1% vs ±3.4% move, PLTR 5.8% vs ±5.4% move) — tightening further now risks a noise wick-out rather than real protection, so leaving both stops unchanged this cycle. PLTR is also CAP-FULL per STRATEGY PARAMS so no add is possible regardless of conviction. Sitting mostly in cash beyond the one RIOT entry: NBIS/CRWV gaps are earnings-day/untradeable chase risk, and MARA/CLSK/WULF are repeat-loser tickers per TICKER-FACIT with no genuinely new catalyst beyond the same short-interest thesis that already burned this account.

[stop-loss triggered at $7.51 — broker-fill $7.51]

Horisont: 1-2 weeks Begrundelse: Post #12 blowout Q2 with 150% commercial revenue growth and INTERPRETED net-sentiment +1.50 (single high-conviction fundamental post) — exit if the post-earnings pop fades back below 150. === Session-analyse === [saxomoon/US] Small live book ($619) after two stop-outs on FIG and WULF that both wick-out (both names since ripped back — FIG +17% and WULF +1% per yfinance delta vs my exit, painful confirmation of the GAP-FLOOR lesson). Cash is thin (~$328 spendable) so this cycle is about discipline, not chasing. PLTR is the real WSB story today (blowout Q2, net-sentiment +1.50, single high-conviction post, katalysator confirmed) but it was already rejected once this session for breaching the 35% cap — at ~$156.68 ask even 1 share ($157) plus my existing IONQ/WEN book would still leave room, so I'll take a small 1-share starter position sized well under cap rather than force the 2-share size that got rejected. IONQ is flat (+0.7%) and WEN up +8.1% with earnings tomorrow (8/7) — tightening WEN's stop slightly to lock in gains ahead of the print rather than adding risk into earnings vol. QQQ/SPY are dead flat (NEUTRAL regime), VIX calm at 16.5 but crypto Fear&Greed is 25 (Extreme Fear) — no reason to chase miners (RIOT/MARA/CLSK all report earnings today too, pure gamble into unknown prints). WSB's real energy today is Burry's bear call and 0DTE loss-porn (posts #4, #8, #9) — that's noise, not a tradeable long thesis. FIG and WULF stay off the table per re-entry rule (both need >=0.7 conviction + 2 independent cross-feed BUY votes; WULF actually has 4-BUY cross-feed consensus which is tempting, but I just got stopped out hours ago on a wick — will reassess next session once the dust settles rather than immediately re-chasing the same name same day). Staying mostly in cash given the thin spendable float and no fresh, uncrowded, cross-feed-confirmed setup beyond the small PLTR starter.

[stop-loss triggered at $25.41 — broker-fill $22.55]

[stop-loss triggered at $17.74 — broker-fill $17.36]

Horisont: 3-7 days Begrundelse: Strong cross-feed BUY (analyst +88% target upside) riding an 8.5% overnight gap into RISK-ON tape, but earnings in 2 days is real volatility risk so sizing small with a wider stop — exit if it fades before the print or the gap reverses. === Session-analyse === [saxomoon/US] RISK-ON day (QQQ +1.37%, SPY +1.14%) with VIX calm at 17 — a favorable tape for leaning into high-conviction names, but WSB itself is thin: no fresh NEW/RISING tickers in TRENDING (NFLX/MU/MSFT/NVDA are all fading/decelerating mention momentum, not rising). The megathread vibe is literally described as 'boredom' — this isn't a screaming-hype session, so I'm not forcing size into stale names. That said, the cross-feed screening block shows a strong bench of BUY-consensus names with real independent corroboration (short_interest + analyst_ratings agreeing, not just Reddit wearing three hats): CLSK, WULF, RIOT, MSTR, IONQ, OKLO all show multi-feed agreement with high-conviction short-interest or analyst upside. Given my thin cash ($323 spendable) and small portfolio, I'll pick two names with the cleanest independent cross-feed corroboration and decent risk/reward, sized to stay near the $100-120 floor per leg while respecting the 35% position cap and the 50% sub-theme cap (I'm currently loaded in restaurants/meme/saas/recent_ipo — a fresh BUY in crypto-mining or AI-adjacent names diversifies away from that). IONQ has an 8.5% overnight gap already (some chase risk) but a genuine analyst BUY (+0.89, +88% target upside) with a real catalyst window (earnings in 2 days — volatile, so I'll size modest and set a wider stop). WULF (BUY consensus, analyst target +115% upside, no earnings blackout risk today since it doesn't report until day+2) is my other pick — cheap enough to get 3+ shares with scale-out room, and it's not gapped up already so less chase risk than IONQ. I'm leaving FIG and WEN alone — both are marginal small gainers (+0.7%/+1.4%), not forgotten (still within normal noise), earnings for FIG in 2 days is a volatility flag but position is tiny and stop is already reasonably placed. No stop tightening needed yet since neither has moved enough to justify raising the stop above current levels per the ladder. Sitting mostly in cash otherwise — WSB genuinely has nothing fresh to chase this session; the 'boredom' comment-flow assessment is itself the signal to stay disciplined rather than manufacture activity in a low-signal environment. WSB is quiet/bored today per the comment-flow read — no NEW/RISING tickers cleared the trending bar, and the only chronic names (NVDA/MU/MSFT/NFLX) are decelerating mentions, not fresh hype. FIG and WEN are small, recent, and within normal noise (+0.7%/+1.4%) — not forgotten, no cut warranted, stops already reasonably placed given their small daily-move bands. Deploying only into the two cleanest cross-feed-corroborated names (WULF, IONQ) rather than forcing size into stale meme chatter; keeping the rest of the modest cash buffer given the small account size and multiple earnings landmines this week (FIG reports in 2 days).

Horisont: 1-2 weeks Begrundelse: Cross-feed consensus BUY (analyst target +115% upside, short_interest supportive) with no earnings until day+2 — fresh diversification away from my restaurant/saas-heavy book, exit if momentum stalls or WSB never picks it up. === Session-analyse === [saxomoon/US] RISK-ON day (QQQ +1.37%, SPY +1.14%) with VIX calm at 17 — a favorable tape for leaning into high-conviction names, but WSB itself is thin: no fresh NEW/RISING tickers in TRENDING (NFLX/MU/MSFT/NVDA are all fading/decelerating mention momentum, not rising). The megathread vibe is literally described as 'boredom' — this isn't a screaming-hype session, so I'm not forcing size into stale names. That said, the cross-feed screening block shows a strong bench of BUY-consensus names with real independent corroboration (short_interest + analyst_ratings agreeing, not just Reddit wearing three hats): CLSK, WULF, RIOT, MSTR, IONQ, OKLO all show multi-feed agreement with high-conviction short-interest or analyst upside. Given my thin cash ($323 spendable) and small portfolio, I'll pick two names with the cleanest independent cross-feed corroboration and decent risk/reward, sized to stay near the $100-120 floor per leg while respecting the 35% position cap and the 50% sub-theme cap (I'm currently loaded in restaurants/meme/saas/recent_ipo — a fresh BUY in crypto-mining or AI-adjacent names diversifies away from that). IONQ has an 8.5% overnight gap already (some chase risk) but a genuine analyst BUY (+0.89, +88% target upside) with a real catalyst window (earnings in 2 days — volatile, so I'll size modest and set a wider stop). WULF (BUY consensus, analyst target +115% upside, no earnings blackout risk today since it doesn't report until day+2) is my other pick — cheap enough to get 3+ shares with scale-out room, and it's not gapped up already so less chase risk than IONQ. I'm leaving FIG and WEN alone — both are marginal small gainers (+0.7%/+1.4%), not forgotten (still within normal noise), earnings for FIG in 2 days is a volatility flag but position is tiny and stop is already reasonably placed. No stop tightening needed yet since neither has moved enough to justify raising the stop above current levels per the ladder. Sitting mostly in cash otherwise — WSB genuinely has nothing fresh to chase this session; the 'boredom' comment-flow assessment is itself the signal to stay disciplined rather than manufacture activity in a low-signal environment. WSB is quiet/bored today per the comment-flow read — no NEW/RISING tickers cleared the trending bar, and the only chronic names (NVDA/MU/MSFT/NFLX) are decelerating mentions, not fresh hype. FIG and WEN are small, recent, and within normal noise (+0.7%/+1.4%) — not forgotten, no cut warranted, stops already reasonably placed given their small daily-move bands. Deploying only into the two cleanest cross-feed-corroborated names (WULF, IONQ) rather than forcing size into stale meme chatter; keeping the rest of the modest cash buffer given the small account size and multiple earnings landmines this week (FIG reports in 2 days).

Horisont: 1-2 weeks Begrundelse: WSB mention-leader today with cross-feed consensus BUY led by short_interest (42.3% of float, ratio 1.8) — a genuine squeeze setup, not just meme noise; exit if short interest unwinds or WSB attention fully fades. === Session-analyse === [saxomoon/US] Risk-on tape today (QQQ +3.5%, SPY +1.8%) with WSB buzzing mostly around MU/memory/semis and a Korea-crash meme narrative that's more comic relief than tradeable. FIG is my only holding, at -4.6% but cross-feed shows a clean consensus:BUY (3 BUY/1 HOLD, top signal yfinance BUY on the 5d+18.8% range recovery) — rule #4 explicitly says a net-bullish cross-feed overrides WSB forgetting the name, so I hold rather than cut. It's also already saturated at 50% saas/recent_ipo theme cap, so no add there regardless. Cash is thin ($468 spendable, portfolio only ~$625) and my last BUY attempt (LEU) got rejected for breaching the 30% position cap — a reminder to size tighter this cycle. WEN is interesting: WSB mention-leader today, cross-feed consensus BUY driven by short_interest (42.3% of float, ratio 1.8 — genuine squeeze fuel), interpreted signal is mildly positive (+0.50, meme thesis, small sample) and it's cheap enough to buy real share count. But it's a thin single-post signal and the theme would add a THIRD sub-theme, which is fine for diversification. MU is the loudest WSB name (calls YOLO, dotcom-bubble nostalgia posts, comment-flow dominated by MU/MSFT euphoria) but it's flagged UNTRADEABLE at min size in this universe — can't size it at ~$625 portfolio without blowing past caps, so pass. Given the FEE-DRAG history (0.67% of turnover, round-trips costing ~1.3%) and MIN-EDGE rule, I want a real edge, not a scalp — WEN's short-squeeze setup plus cross-feed corroboration (short_interest is independent, not Reddit-derived) clears that bar for a small, single new position. FIG stays held despite being -4.6% and not in today's trending list — cross-feed shows a clean consensus:BUY (yfinance flagging the 5-day +18.8% recovery), which per rule #4 overrides WSB simply moving on. It's also at its sub-theme cap (saas/recent_ipo at 50%) so no further sizing there. MU is the loudest WSB theme today (dotcom nostalgia, YOLO calls, heavy comment-flow) but sits on the UNTRADEABLE-at-min-size list for this portfolio size, so it's off the table regardless of hype. Cash is limited (~$468 spendable against a ~$625 book) so I'm keeping this cycle to one small, cross-feed-corroborated addition (WEN) rather than forcing multiple thin bets — consistent with MIN-EDGE discipline given the fee-drag history.

Horisont: 1-2 weeks Begrundelse: WSB DD post '$UBER, follow the smart money' (score 469) is today's freshest non-chronic ticker in the hot listing, with a calm daily range letting a tight 5% stop fit the risk budget. === Session-analyse === [saxovegas/US] RISK-ON tape (QQQ +0.76%, VIX calm at 16) and WSB's top fresh DD today is the $UBER 'follow the smart money' post (score 469, #25 hot) — genuinely NEW attention vs the chronic MU/TSLA earnings chatter. UBER's daily range is tight (~1% swings), gulv-stop lands at 4.1% which fits comfortably inside the $37 risk budget for 5 shares (~$346, well under the $374 cap and $577 spendable). No cross-feed line explicitly covers UBER but the WSB DD is concrete and the setup is clean technically (flat/rising into support ~$69). MU earnings TODAY is loud on WSB (JPMorgan beat-and-raise thesis, $50K YOLO post, INTERPRETED signal +0.94 net-sentiment) but MU is flagged UNTRADEABLE at min size for this account — $1075/share means even 1 share risk-sized properly doesn't clear the budget cleanly, and opening a fresh earnings-day position I can't size sanely is exactly the low-edge trap rule #17 warns against. Passing. MARA/RIOT crypto-mining cross-feed shows SELL consensus and the THEME-FACIT shows this book has lost -723 DKK across 10 mining round-trips — no fresh external catalyst here, staying out per the theme re-entry bar. Holding INTC and RDDT untouched — both CAP-FULL (no room to add under the 30% position cap) and both gapped UP overnight (+2.8%, +1.4%) on no bad news, consistent with EXIT OVERRIDE keeping bagholding-cut reflexes off. Neither has moved enough from entry to justify ratcheting stops tighter than current resting levels (INTC -2.1%, RDDT -3.6%, both still below breakeven) — leaving stops as-is rather than forcing a premature tighten inside their daily noise bands (INTC ±3.4%, RDDT ±2.4%). INTC and RDDT are CAP-FULL — no adds permitted regardless of conviction — and both are still underwater from entry despite today's overnight bounce, so no stop-tightening is warranted yet (would risk firing inside normal daily noise). MU is the loudest WSB name today (earnings-day YOLO thread, JPMorgan beat-and-raise chatter) but is explicitly UNTRADEABLE at this account's min size — skipping rather than forcing a sub-scale earnings-day bet. Crypto mining (MARA/RIOT/IREN) shows cross-feed SELL consensus and a -723 DKK theme track record with no new external catalyst, so staying out per the theme re-entry bar.

[stop-loss triggered at $185.27 — broker-fill $185.20]

Horisont: 1-2 weeks Begrundelse: WSB YOLO thread ($146K RDDT Jan28 calls, disclosed position) plus cross-feed consensus BUY (analyst +40% upside) and social-theme tape +7.1%/5d; exit if WSB interest fades or thesis-break below the stop. === Session-analyse === [saxovegas/US] COIN and INTC are both CAP-FULL per STRATEGY PARAMS — no adds allowed regardless of conviction. COIN is up +17.6% with an avg dagsmove ±4.4%; current stop sits at 4.9% under price which is inside the noise band, but the ladder says +15% PnL → trail ~12%, and gap-floor says never trail TIGHTER than the daily move — so I'm leaving it roughly where the ladder/floor intersect rather than tightening further into noise. INTC at +2.2% is barely off entry with stop already 9.3% under (above its 3.7% dagsmove floor) — no change needed, thesis intact (semis theme is +6.5%/5d, INTC own overnight gap -2.4% is normal noise not a break). On new money: WSB hot list today is thin on fresh tradeable names — NKE (score 15,162) is a meme/loss-porn thread about a mercy-kill short, not a bullish catalyst (title: 'The only way to rescue Nike…'), and its own gulv-stop math only allows 10 shares max ~$359, but the thesis behind the hype is bearish irony, not a buy signal — skip. ORCL has real news (force majeure notice on datacenter, -ish) but cross-feed isn't shown consensus BUY and it's a headline-driven news risk, not WSB conviction — pass. RDDT has a genuine dot: WSB YOLO thread ($146K Jan28 calls), cross-feed consensus BUY (analyst target +40% upside), tema-tape social +7.1%/5d — but gulv-stop caps it at 2 shares (~$305), and per rule #9/theme discipline that's a small, single-shot bet in a name with no existing exposure; META also cross-feed BUY with WSB net+0.92 disclosed position but is UNTRADEABLE at min size per this session's rules. Given RISK-ON regime (QQQ +0.49%), calm VIX (14.2), and a real trending name with cross-feed corroboration, I'll take a modest RDDT position sized to the gulv-stop max (2 shares) — this respects rule #17 min-edge only marginally, but the $305 notional clears the $200 floor and the setup (WSB momentum + analyst upside + social tema tailwind) is the best fresh idea in an otherwise quiet tape. No sub-theme conflict since 'social' is currently at 0%. Cash stays largely on the sidelines (41% of book) — no other name clears both a fresh WSB angle AND cross-feed corroboration AND passes the risk-sizing math; TTWO's single-post YOLO is thin sample per rule #15 (interpreted signal is high net-sentiment but n=1, no real corroboration) and tema-tape shows gaming -13.7%/30d — pass. AMZN mention is accelerating but interpreted signal is net-negative/ironic (-0.12, meme thesis) — no real conviction there. Holding the two live winners with stops resting above their volatility floor is the correct discipline this cycle; last cycle's IREN stop fire (+5.4%, above its own gulv-band) was a legitimate protective exit, not a forced cut, so nothing to second-guess there per rule #7. COIN and INTC are CAP-FULL, so no adds there regardless of conviction; both stops already sit above their respective daily-noise gulv-floors and I'm not tightening further into noise per rule #13's gap-floor. Cash (41% of book) stays mostly parked: NKE's massive hot-list score is loss-porn/mercy-kill irony not a bull catalyst, ORCL carries fresh negative news risk without cross-feed corroboration, META/other big names are UNTRADEABLE at this session's min size, and TTWO/AMZN reads are thin-sample or net-negative per the interpreted signal. RDDT was the one name clearing WSB momentum + cross-feed BUY + tema tailwind at a size that clears the $200 floor.

[stop-loss triggered at $44.95 — broker-fill $44.94]

Horisont: 1-2 weeks Begrundelse: WSB has fresh accelerating INTC gain posts ('14k->38.5k weekend hold', 'Mr. Bottom Catcher') plus a +12% overnight gap on a risk-on tape — sizing to the 8.9% gulv-stop floor (3 shares ≈ $365) rather than the ticker's full momentum given semis theme has been a net loser for this book. === Session-analyse === [saxovegas/US] Risk-on tape today (QQQ +2.54%, SPY +1.49%, VIX a calm 15.4) and WSB is squarely in a semis/AI celebration mood — AMD hitting $1T market cap is the #4 hot post with strong loss-porn/momentum sentiment, MU spreads gain post is top-10, and INTC has multiple 'weekend hold' gain posts. Cross-feed corroborates several of these (MU consensus BUY on analyst upside, IONQ/CLSK/IREN/ARM all BUY-consensus), but per the min-shares/floor math AMD, MU, INTC-at-min-entry are mostly UNTRADEABLE or barely tradeable at this tiny $656 spendable budget — INTC is the one name that actually clears the risk-sizing floor (max 3 shares, gulv-stop ~8.9%), and it has genuine WSB fresh momentum (accelerating mentions, two 'weekend hold gain' posts, +12% overnight gap) plus room in the semiconductors theme (0% current exposure) that's been a net loser for us historically (-143 DKK) but INTC specifically hasn't been the loser (that was NVDA/older INTC positions) — still, I'll size it modest and respect the theme's soft memory. COIN and IREN are both healthy winners (+22.9% / +11.9%) — COIN is CAP-FULL so no add is possible anyway, and IREN's cross-feed consensus BUY plus ai_infra/crypto_mining tape strength argues for holding and trailing rather than trimming. AMD, MU, QQQ, META, SPY are all effectively untradeable at this account's tiny risk budget (gulv-stops too wide or min-entry impossible under the $200 floor) despite being the loudest WSB names today — chasing them with 1 share each would be exactly the binary 1-2 share mistake rule #3 warns against. Cash stays at ~51% of book; with only $656 spendable and most hot names priced/vol'd out of range, one modest INTC add is the only trade that clears both the WSB-momentum bar and the risk-sizing floor this cycle.

Horisont: 3-7 days Begrundelse: WSB actively debating COIN ahead of the CLARITY Act senate vote (posts #19/#20, plus #4 news at 7,312 upvotes) with cross-feed consensus BUY driven by 19.6% short interest — exit if the regulatory narrative fades or COIN breaks below recent support. === Session-analyse === [saxovegas/US] RISK-ON tape (QQQ +0.88%), VIX calm at 17.1, and IREN is up +2% on entry with a strong overnight gap (+4.9%) confirming crypto_mining momentum — thesis intact, no reason to touch the position beyond trailing the stop tighter to lock in the gain per rule #13, though the move is modest (+2% PnL) so I'm nudging the stop only slightly rather than aggressively, since the gulv-floor (12.6% avg dagsmove-based) still dominates the ladder math at this PnL level. On new entries: COIN shows cross-feed consensus BUY (short_interest +0.69, 19.6% float short) and is explicitly trending in WSB (CLARITY Act senate vote fail post at #4 with 7,312 upvotes, plus two direct COIN posts at #19/#20 debating the Clarity Act angle) — this is a real WSB catalyst (regulatory vote), not stale chatter. But the gulv-stop math caps COIN at 2 shares (~$332) given its 10.4% floor, and my risk budget only supports that size — still, 2 shares clears the $200 min-trade floor. ORCL also crossed my screen (WSB post #5 on layoffs, INTERPRETED signal net -1.00 bearish, CFO defensive quote) — that's a bearish catalyst on the actual post, not a buy signal, so I'm passing despite ORCL being 'tradeable' and mentioned; rule #18 requires me to honor a named headwind rather than override it. NVDA is under a mandatory re-entry cooldown (exited -4.3% via stop fire 9/14) requiring conviction ≥0.7 AND 2+ cross-feed BUY votes — current cross-feed shows only 1 BUY vs 3 SELL, so it's server-blocked regardless of my conviction; skipping. For COIN: short interest at 19.6% of float with WSB actively debating the CLARITY Act vote failure as a bullish setup ('Ready for COIN to rip ahead of senate vote') is a genuine squeeze-adjacent setup, cross-feed consensus BUY, and it fits within crypto/fintech theme without overloading crypto_mining (COIN is not literally a miner, diversifies away from IREN's pure-mining exposure). Sizing to the gulv-stop cap of 2 shares (~$332) keeps risk contained. RKLB WSB post is 100% loss-porn per INTERPRETED signal (author admits losses, no fresh catalyst) — that's a fade signal per rule #15, skipping despite the trending mention. IREN stays as-is beyond the modest stop tightening — thesis intact (crypto_mining theme tape +0.2% 30d, overnight gap +4.9%, RISK-ON regime), not yet at the +15% ladder trigger so no aggressive trail. Passing on ORCL despite the WSB mention because the actual post content (CFO comment on layoffs) is a named bearish catalyst, not support — buying into a headwind I've identified myself would violate rule #18. NVDA is cooldown-blocked (exited via stop 9/14, cross-feed only 1 BUY vs 3 SELL, fails the 2-vote server requirement). RKLB's WSB post is pure loss-porn with no fresh catalyst — fading per rule #15. Sitting mostly in cash (now ~$650 remaining spendable) is appropriate mid-cycle; will reassess at 14:00 for fresher trending names.

Horisont: 1-2 weeks Begrundelse: JPM upgraded IREN to overweight (interpreted signal +1.50, 1 high-conviction post, katalysator 1/1) and cross-feed shows consensus BUY (analyst target +78% upside) — a genuinely fresh catalyst distinct from my prior loss on this name, sized small given RISK-OFF tape and my poor crypto_mining track record. === Session-analyse === [saxovegas/US] QQQ is down -1.66% and SPY -0.74% today — a clear RISK-OFF regime, and the overnight tape confirms it: SOXL -16%, MRVL -8%, ARM -7%, MU -6.4%, INTC -6.4%, ASML -6%, CRWV -6% — semis and AI-infra got hammered overnight on the SK/Samsung uncertainty headline and 'AI slowdown' discussion sitting on WSB's hot list right now. WSB itself is mostly macro fear today (rate hike chatter, Warsh/Volcker hawkish talk, oil shock/Iran headlines, 'this selloff makes zero sense') — there is no fresh, high-conviction single-name YOLO trending in the actual tradeable universe; NVDA is the only universe ticker mentioned and its own interpreted-signal read is net-bearish (-0.25, 3 bear vs 1 bull, thesis fundamental) despite a cross-feed-adjacent posture. My NVDA position was just stopped out this morning (-4.3%), and rule #8 gates re-entry hard: it needs conviction >=0.7 AND >=2 fresh cross-feed BUY votes citing something NEW — I have neither, so NVDA stays off the table (cooldown respected). Looking at cross-feed consensus BUY names that ARE in the trending/interpreted mix: IREN shows a genuine catalyst (JPM upgrade to overweight, interpreted net +1.50, 1 high-conviction post) plus cross-feed consensus BUY (analyst_ratings +0.82, target +78% upside) — but my own TICKER-FACIT shows I've lost on IREN before (-137 DKK) and it's down again overnight (its theme, crypto_mining, is my worst theme at -757 DKK over 90d with 11 exits). The JPM upgrade is a genuinely NEW catalyst though (rule #8's 'what changed' bar), and it's a small, single-name divergence from a falling semis tape, not a chase into strength. Given RISK-OFF regime + my own poor crypto_mining track record, I'm sizing this very small and using a wide stop to respect the volatility floor, rather than skipping entirely — this is the single name that clears both a fresh catalyst AND cross-feed corroboration bar today. Everything else — GME, MARA, CLSK, RIOT, WULF, MRVL, ARM, CRWV, ASTS, NBIS — either has zero fresh WSB narrative behind it today (just perennial cross-feed BUY tags) or is actively getting crushed by overnight semis/AI-slowdown fear, which is exactly the kind of chase-into-a-red-tape rule #11/#16 warns against. With cash at ~$1,191 spendable and portfolio down slightly over the last week, the correct move on a day the NASDAQ is down -1.66% with no standout fresh WSB catalyst except IREN's JPM upgrade is to stay mostly in cash and take one small, catalyst-backed position rather than force multiple entries into red-tape names. Sitting mostly in cash today. QQQ -1.66%/SPY -0.74% is a clear risk-off signal and the overnight move confirms it broadly across semis/AI-infra (SOXL -16%, MRVL -8%, ARM -7%, MU -6.4%, INTC -6.4%, CRWV -6%) — exactly the tape rule #11 says to avoid chasing. WSB's hot list today is dominated by macro fear (rate hikes, oil shock, 'selloff makes zero sense') rather than a fresh single-stock YOLO; the only universe ticker mentioned by name, NVDA, just stopped me out this morning and its own interpreted read is net-bearish, so re-entry cooldown (rule #8) and lack of conviction keep it off the table. Cross-feed consensus-BUY names like GME, MARA, CLSK, RIOT, WULF, NBIS, CRWV have no fresh WSB catalyst behind them today — buying them now would be chasing perennial tags into a falling tape, not reacting to new evidence. I'll re-assess at 11:30 once WSB fully wakes up and the overnight semis rout either stabilizes or deepens.

[stop-loss triggered at $210.72 — broker-fill $210.48]

Horisont: 1-2 weeks Begrundelse: WSB has 4 posts on NVDA today (Jensen Huang memes, Hugging Face acquisition news) with net-sentiment +0.62 and a high-conviction post citing a concrete catalyst, corroborated independently by analyst_ratings BUY (+46% upside target) — exit if it breaks below recent support near $208 or the AI-bubble skepticism thread (post #7) gains real traction. === Session-analyse === [saxovegas/US] Today's WSB feed is thin gruel — the hot posts are memes (Jensen Huang, oil deals, sailors in Thailand), not fresh tickers. Only NVDA clears actual mention volume (4 posts, net-sentiment +0.62, 1 high-conviction, katalysator 3/4) and it's backed by an independent cross-feed BUY consensus (analyst_ratings +0.89, target +46% upside) — that's genuine multi-source corroboration, not just Reddit wearing three hats. But QQQ is -1.03% today, a clear risk-off tape, and VIX is calm at 14.4 so it's not a panic day, just a soft one. NVDA itself isn't in the TRENDING/MENTION-LEADERS blocks (those are RDDT/TTWO) so it doesn't get the auto +1 bump, and RDDT's interpreted read is actually net-bearish (-1.50, high-conviction, thesis 'growth story over') despite trending — that's a clear caution flag per rule #15, not a buy. Given RISK-OFF regime (#11) raises the bar, and my own THEME-FACIT shows semiconductors have been a net loser for me (-410 DKK, 5 exits) and crypto_mining a bigger net loser (-766 DKK, 14 exits), I'm not chasing miners today (MARA is cross-feed SELL 2-4, SOXL cross-feed SELL, OKLO cross-feed SELL despite short-interest flag). CLSK/RIOT/MARA are all overnight red and cross-feed net-negative — sitting those out. NVDA is the one name with real independent corroboration (analyst_ratings BUY, INTERPRETED bull read, historically I've had one small win on NVDA per TICKER-FACIT +81 DKK) but semiconductors tape is -6.6% over 30d and the entry here is a small, disciplined nibble given the red NASDAQ day — sizing down rather than skipping entirely, since the cross-feed and interpreted signal both agree. Everything else — SOUN (cross-feed BUY but overnight gapped -5.1%, and my own TICKER-FACIT shows I bagged a +306 DKK win last time so I'm wary of re-chasing after a big short-interest-driven pop that already faded), ASML/AVGO (cross-feed BUY but not WSB-hyped, and AVGO reports earnings tomorrow — avoid fresh entries into a print), QBTS (huge yfinance delta anomaly +17% looks like a data glitch, skip) — none clear the bar for a fresh position on a risk-off day. No open positions to manage stops on. Staying mostly in cash with one small, well-corroborated NVDA nibble is the correct read of a quiet, meme-only WSB session on a red tape. WSB today is meme-heavy and ticker-light: the actual trending block only surfaces NVDA and SPY from post titles, mention-leaders are RDDT/TTWO with RDDT's interpreted read flipping bearish (-1.50 net-sentiment, high-conviction, capped-DAU thesis) despite the mention count — a textbook rule #15 caution, not a buy. QQQ is red (-1.03%) which raises my bar per rule #11, and my own THEME-FACIT shows crypto_mining and semiconductors have both been net losers over 90 days, so I'm not chasing MARA/CLSK/RIOT/OKLO even though short-interest flags exist — cross-feed screening actively shows SELL consensus on MARA, CLSK, SOXL, OKLO, HOOD, IBIT today, confirming caution is warranted. AVGO reports earnings tomorrow so no fresh entry there despite the cross-feed BUY. Sitting mostly in cash (only holdings: none currently) with a single small, corroborated NVDA nibble reflects the quiet-WSB, risk-off backdrop — I'll get more aggressive if QQQ turns green and WSB starts naming a fresh high-short squeeze candidate with real independent cross-feed support.

[stop-loss triggered at $125.05 — broker-fill $125.05]

[stop-loss triggered at $19.96 — broker-fill $19.99]

Horisont: 1-2 weeks Begrundelse: Overnight +7% gap on BTC strength, INTERPRETED signal shows a high-conviction call buyer post (net-sentiment +0.75), cross-feed consensus BUY with analyst target +92% upside, and crypto_treasury tema-tape (+15.8% 5d) is diverging positively from my losing crypto_mining basket — a genuinely different bet, not the same theme in new clothes; exit if BTC rolls over or MSTR loses the $115 level. === Session-analyse === [saxovegas/US] RIOT is the only holding and it's up modestly (+4.4%) with the overnight tape actually favorable (RIOT gapped +4.6% overnight, crypto Fear&Greed at 73 Greed, BTC 30d +20.5%). Cross-feed shows RIOT consensus BUY (3 BUY/1 SELL/2 HOLD) with analyst target +63% upside — decent corroboration. But QQQ is red -1.18% today (RISK-OFF regime), crypto_mining theme is already AT CAP (100% of book), and my own THEME-FACIT shows -767 DKK across 13 exits on this exact theme — I have a proven negative edge here. No new capital should chase further crypto mining exposure today. MSTR (crypto_treasury) is tempting — INTERPRETED signal shows net-sentiment +0.75 with a high-conviction call-buyer post, cross-feed consensus BUY, and tema-tape shows crypto_treasury +15.8%/5d and +31.7%/30d, diverging positively from the crypto_mining basket's -9.4%/30d. That's a genuine name-specific catalyst (MSTR gapped +7% overnight on BTC strength) rather than borrowed mining momentum. However it's a RISK-OFF day on NASDAQ, VIX is calm (16) but Greed is elevated (73) — blow-off-top caution per rule #12. MRNA is the loudest WSB story today (cancer vaccine breakthrough, $500k YOLO posts, 38 comments in flow) but it's NOT in the tradeable universe... wait, checking again — MRNA IS in the tradeable universe. But it already ran +50% and is now pulling back overnight (-5.3%), classic blow-off/fade risk after a parabolic spike on binary trial news — chasing a name already up 50%+ on a single-day biotech pop is exactly the low-edge chase rule #16 warns against, and WSB sentiment on it is 'mixed' per the comment-flow gist, not clean bullish. Holding RIOT unchanged — it's a modest winner (+4.4%) but crypto_mining is already AT CAP (100% of book) and my own THEME-FACIT shows -767 DKK realized across 13 round-trips in this exact theme, so I'm not adding more mining exposure regardless of today's favorable overnight gap. Its stop stays where it is since the move hasn't extended enough to justify tightening beyond the current 18.7% cushion, and the position is small relative to portfolio total. RISK-OFF regime today (QQQ -1.18%) raises my bar for fresh adds across the board; the one new trade (MSTR) is sized small (2 shares, ~$251) specifically because Fear&Greed is at 73 (Greed) and blow-off-top risk on crypto-adjacent names is real. MRNA — today's loudest WSB story — is being passed on: it already spiked 50%+ on binary trial news and is fading overnight (-5.3%), which is a textbook chase-risk/low-edge entry per rule #17, and WSB's own commentary on it reads as 'mixed', not clean conviction. Sitting mostly in cash/RIOT rather than forcing more trades on a red NASDAQ day.

Horisont: 1-2 weeks Begrundelse: Cross-feed consensus BUY (4-1-1) with technicals leading (5d +15%, range_pos 64%) and short-interest fuel behind crypto miners broadly — exit if BTC rolls over or RIOT loses the $18.50 shelf. === Session-analyse === [saxovegas/US] WSB's front page today is dominated by MRNA (not in universe) and generic mag7/macro noise — the actual trending tradeable names (RDDT, NVDA, MU) are fading (▼ vs yesterday), not fresh. HOOD stands out on the OVERNIGHT GAP block (+13.8% overnight) and mention-leaders (+3 today), but the interpreted read is thin (1 post, +0.50, no high-conviction) and it's not in cross-feed screening at all — that's a chase-risk gap, not a corroborated setup, so I'm passing on it rather than buying the top of a 14% gap. The real signal today is cross-feed consensus on crypto-mining names: RIOT, MARA, CLSK, IREN, WULF all show BUY consensus with short_interest as the top driver (CLSK 40.3% short float, IREN 28.4%, WULF 123% upside target) plus strong technicals (MARA 5d +32%, RIOT 5d +15%). However my own THEME-FACIT shows crypto_mining has been a -767 DKK loser across 13 exits on these exact 5 names, and I just got stopped out of MARA (+266 DKK, a win this time) and IREN (-137 DKK) in the last 72h. IREN is also in my RECENT LOSING EXITS list requiring conviction ≥0.7 AND 2+ independent cross-feed votes — it has 2 independent votes (short_interest + others) which clears the bar, but earnings in 6 days is a real risk I don't want to underwrite with a fresh position right before the print. RIOT and CLSK don't carry that earnings overhang and have clean independent cross-feed corroboration (short_interest + technicals), so I'll take one measured position in RIOT (better technical setup, 5d+15%, range_pos 64% vs CLSK's crowded 40% short but no earnings catalyst either way) sized small given the theme's poor track record — this is a single bet, not adding legs across the whole mining complex. QQQ is green (+0.34%) and VIX is calm (14.89), so risk-on tilt supports a modest add, but I'm keeping size well under max position given the theme-level scar tissue and avoiding IREN/WULF/MARA to not re-stack the same correlated bet four times. Sitting mostly in cash outside the one RIOT entry. HOOD's overnight gap is tempting but uncorroborated by cross-feed screening and the interpreted read is a single thin post — chasing a +14% gap on drive-by mention-leader data alone fails the min-edge/corroboration bar. MRNA is the actual WSB obsession today but it's not in the tradeable universe. RDDT/NVDA/MU are all fading mentions (▼ vs yesterday) with negative or bearish interpreted sentiment (MU -0.50, INTC -0.50) — no fresh catalyst there, and NVDA has earnings in 5 days which argues against a fresh pre-print position anyway. Crypto-mining as a theme has cost me -767 DKK over 13 round-trips, so I'm treating today's cross-feed BUY consensus on RIOT/CLSK/IREN/WULF/MARA as one correlated bet and only taking the single cleanest name (RIOT, no earnings overhang, best range position) rather than stacking legs across the whole complex. Will reassess IREN specifically after its earnings print on 8/27.

[stop-loss triggered at $11.04 — broker-fill $11.03]

Horisont: 1-3 weeks Begrundelse: Cross-feed consensus BUY driven by short_interest (31.5% of float, 2.2x ratio) — a genuine squeeze setup independent of Reddit hype, confirmed by the +6.8% overnight gap; exit if BTC rolls over or the gap fully fills below $9.20. === Session-analyse === [saxovegas/US] Thin gruel on WSB today — RDDT is the only tradeable name with genuine mention momentum, and the interpreted read flags it as thin (2 posts, 50% loss-porn, 0 high-conviction) despite a decent net-sentiment number. Worse, RDDT is a RECENT LOSING EXIT (stopped out -5.9% on 08-17), which per rule #8 requires conviction >=0.7 AND 2 independent cross-feed BUY votes — the cross-feed line isn't even shown as consensus BUY here, and TEMA-TAPE shows the social basket down -10.1%/5d and -14.9%/30d. That's a stacked no: forgotten thesis, tape headwind, and the re-entry bar unmet. Skip it. Where the real signal is: MARA and CLSK both show consensus BUY with short_interest as the top independent feed (MARA 31.5% short/2.2x ratio, CLSK 40.3%/3.2x) — genuine squeeze fuel, not just Reddit noise, and the overnight gap shows MARA already +6.8% confirming momentum. But crypto_mining is my worst-performing theme by a mile (THEME-FACIT: 12 exits, -1033 DKK across 5 names) and both were losing round-trips (RIOT -179, CLSK -164 in the ticker facit, though not MARA itself). Per rule #8/theme discipline, re-entering this theme needs a theme-EXTERNAL catalyst, not just cluster momentum — the short-interest fundamentals ARE somewhat name-specific and independent (not just chasing the crypto pump), and MARA itself isn't in my losing-exit list, so I'll take a single small, disciplined position in MARA only (not CLSK, avoiding doubling the theme bet) sized well under cap, with a stop respecting the volatility (10d closes swinging ~4-6%, so 2x that framework points toward a wider stop). Everything else: QQQ regime is neutral, VIX calm at 15, Fear&Greed at 46 (neutral-fear) — no strong tilt either way. NVDA has earnings in 7 days and the interpreted read is actually negative (-1.0 net-sentiment, AI slowdown thesis) — avoid. MU also has earnings soon and consensus is thin (1 BUY vote). No other name clears both the WSB mention bar and independent cross-feed corroboration cleanly enough to justify spending the very limited ~$1,178 spendable cash on more than one position — better to stay concentrated and disciplined than spray small sub-floor bets across many names. WSB is quiet on tradeable names today — only RDDT shows real mention acceleration, but it's a recent losing exit with thin interpreted conviction (50% loss-porn, 0 high-conviction posts) and the social theme tape is down double-digits over 5d/30d, so the re-entry bar isn't cleared. Sitting on the rest of the cash rather than forcing entries into NVDA/MU ahead of earnings, or adding a second crypto-mining name (CLSK) which would just double down on my worst-performing theme (-1033 DKK over 12 exits) without an external catalyst. Will revisit RDDT if a fresh independent cross-feed BUY consensus appears and mention momentum sustains past a single day.

[stop-loss triggered at $41.73 — broker-fill $41.73]

[stop-loss triggered at $221.68 — broker-fill $220.06]

[stop-loss triggered at $165.82 — broker-fill $165.81]

[stop-loss triggered at $101.72 — broker-fill $101.70]

Horisont: 1-2 weeks (through the Aug 18 S&P inclusion) Begrundelse: Post #9/#25 — Reddit joins S&P 500 on Aug 18 (concrete dated catalyst, forced index buying), interpreted read net +1.50 with 3 high-conviction posts and cross-feed 4-0 BUY; exit if it loses the gap into inclusion day or the stop fires. === Session-analyse === [saxovegas/US] RDDT is the one genuinely fresh, multi-source setup today: the S&P 500 inclusion news (post #9 at 2,736 upvotes and #25) is a concrete, dated catalyst on Aug 18 — index funds are forced buyers — and the interpreted read backs it hard (net +1.50 over 3 posts, 3 high-conviction, thesis 'fundamental', 0% loss-porn, 2/3 posts citing a concrete catalyst). Cross-feed shows consensus BUY 4-0. The problem is price: RDDT already gapped +10.6% overnight into the open, which is exactly the chase risk rule #16 warns about, and today's tape is RISK-OFF (QQQ -0.42%). Given spendable cash of only ~$440 USD, a 2-share RDDT entry (~$351) is my one meaningful shot. I weigh the chase risk explicitly by NOT full-sizing (2 shares, not 2.5x) and by giving it a wider stop (9%) rather than a tight one inside the post-gap noise band — a 3-4% stop on a name that just moved 10% overnight is a coin-flip. Index-inclusion flows into Aug 18 are the thesis; if it fails to hold the gap through the inclusion date, the stop does the work. On existing book: NVDA is my only winner (+11.5%) and the $500B financing story (post #10, interpreted net +1.50, high-conviction, concrete catalyst) is fresh bullish news — I do NOT trim per rule #2. Its stop is already 2.9% under price and avg dagsmove is only ±1.0%, so I can ratchet modestly to ~2.5% without violating GAP-FLOOR. CRWV and IREN are both roughly flat and both still carry cross-feed consensus BUY (CRWV 3-1 with 20.4% short float; IREN 3-0 with 28.4% short float) — the EXIT OVERRIDE on this tier plus a net-bullish cross-feed means I hold both and let the mandatory stops cap risk. I will not tighten CRWV (stop already 3.5% under price vs ±4.2% daily move — already inside the noise band, tightening further is a market order in disguise) and leave IREN's 6.1% trail alone since it's at breakeven, nothing to protect yet. I'm skipping everything else. RIOT has the flashiest headline (Anthropic $9B cloud deal, post #21) but I was stopped out of it two days ago at -8.8% and the server requires 2+ independent cross-feed BUY votes for re-entry — RIOT does not appear in the cross-feed block at all, so it's blocked and correctly so. MSFT/MU/AMD/META are untradeable at my size. NBIS is mention-leading but the interpreted read is net -1.50 (the one post is someone SELLING it) and it gapped +7.6% — no. PLTR's +34% run is a Spider-Man meme with 0 high-conviction posts; that's noise. Sub-theme ai_infra is already AT CAP at 50%, which is another reason RDDT (social/index) is the right place for fresh cash rather than another AI-infra name. Holding CRWV and IREN — both near flat, both still carry net-bullish cross-feed consensus (CRWV 3-1, 20.4% short float; IREN 3-0, 28.4% short float), and this tier's EXIT OVERRIDE bars reflexive forgotten-cuts; the mandatory broker stops (101.72 / 41.73) cap the downside. NVDA held and not trimmed despite +11.5% — the $500B financing headline is fresh bullish news, so I ratchet the stop to 2.5% under price (avg dagsmove only ±1.0%, so this respects GAP-FLOOR) and let it run per rule #2. No RIOT re-entry despite the Anthropic headline: it's a losing exit from two days ago with zero independent cross-feed BUY votes today, so it fails the server's re-entry bar. No adds to ai_infra — already at the 50% sub-theme cap. Risk-off tape (QQQ -0.42%) plus Crypto Fear at 29 keeps me to a single, deliberately under-sized entry rather than spraying cash across a red day.

[stop-loss triggered at $19.16 — broker-fill $19.16]

Horisont: 2-4 weeks Begrundelse: Mention-leader with interpreted net +1.50 on a concrete catalyst (first 50MW delivered to MSFT) plus INDEPENDENT short_interest at 28.4% of float — exit if the MSFT-capacity thesis stalls or it loses the $38 shelf. === Session-analyse === [saxovegas/US] RISK-ON tape (QQQ +1.22%), VIX 14.55 calm, and my one real spendable slug is ~$210. The clear fresh signal today is RIOT: post #21 — 'Anthropic Strikes $9 Billion Cloud Deal With Riot ($RIOT) Platforms' — is a NEW-tagged trending ticker with a concrete catalyst, and cross-feed backs it (consensus BUY, 3 BUY / 1 SELL, analyst target +52%). Problem: RIOT is *down* -3.7% overnight and I'm already -6.4% on 15 shares at 28% of book in crypto_mining — my only theme at/near the discretionary risk zone. So no add there; I let the mandatory stop do its job (EXIT OVERRIDE says don't reflexively cut it either, and cross-feed is still BUY). Instead I tighten nothing on RIOT — its stop already sits 2.1% under price, at/inside its ±2.1% daily-move floor, so ratcheting further would be a coin-flip wick-out. Where I do want the cash: NVDA is my best runner (+11.6%) but CAP-FULL, so no add. CRWV likewise CAP-FULL despite the +12% revenue-doubling post (#19) and net +1.50 interpreted read. IREN is the fresh independent setup: mention-leader (+4 today), interpreted net +1.50 with a concrete catalyst ('$IREN delivers first 50MW to $MSFT'), and cross-feed consensus BUY driven by an INDEPENDENT feed — short_interest 28.4% of float, ratio 2.1. That is WSB hype on a genuinely high-short name, my A+ setup, and the independence test passes (short_interest is not Reddit-derived). My last cycle's IREN buy was rejected purely on cash sizing, not thesis — 5 shares at $44.89 = $224, which clears the $200 floor with headroom and fits inside the ~$210 spendable... barely. At ask 44.89, 5 sh = $224.45, above my spendable. So I size 4 shares = $179.56 — that FAILS the fee-guard floor. Rotation is the answer per rule #9: MP is +0.5% flat, 4 shares ≈ $220, in a rare_earth theme with no WSB mention at all today (its BUY is analyst-only, boring-fundamental, not my mandate). Sell MP, buy 6 IREN with the combined proceeds (~$430 of a $384 cap → cap-limited, so 8 shares would breach; 6 sh = $269, safely inside 30%). Stop discipline: IREN's daily swings are violent (39.75 → 44.89 in three sessions, ~5-6% typical), so a tight stop is noise-bait — 12% under fill gives the squeeze room to breathe while capping the live downside, per the volatility anchor. NVDA at +11.6% earns a ratchet toward the ladder: trail 12% under price would be looser than its existing 3.6% stop, so engine ignores it; I leave NVDA alone since its stop is already tighter than the ladder and well outside its ±1.4% daily noise. CRWV I nudge from 5.5% to 5.0%? No — that's inside its ±4.2% daily move only marginally and PnL is just +1.1%; not earned. Leave it. RIOT stays despite -6.4% and the -3.7% overnight gap: cross-feed is still consensus BUY (analyst target +52%) and the Anthropic $9B cloud deal post is today's NEW-tagged catalyst, so the EXIT OVERRIDE applies and the mandatory stop at 19.16 caps the damage — I will not average down into a gap-down on an already 28%-weighted theme. NVDA (+11.6%) and CRWV (+1.1%) are both CAP-FULL so adds are impossible regardless of the $500B-financing and revenue-doubling posts; their existing stops (3.6% and 5.5% under price) are already tighter than the trailing ladder would set and comfortably outside each name's daily noise band, so no ratchet is earned this cycle. Skipped MSFT/MU/TSM despite strong reads — flagged UNTRADEABLE at min size. Skipped PLTR (net +0.92 but 'meme' thesis with the author having SOLD, and no independent feed beyond a Reddit-derived wsb vote) and HTZ (80.6%-style squeeze fuel but the only interpreted post is a -$80k loss-porn bagholder, and it gapped -8.6% into a risk-on tape — that divergence is a red flag, not a dip).

Horisont: immediate rotation Begrundelse: Full exit of a flat, WSB-silent analyst-only position to fund the fresh IREN squeeze setup — MP has zero mentions in today's trending/interpreted blocks, so it is not a WSB-mandate holding. === Session-analyse === [saxovegas/US] RISK-ON tape (QQQ +1.22%), VIX 14.55 calm, and my one real spendable slug is ~$210. The clear fresh signal today is RIOT: post #21 — 'Anthropic Strikes $9 Billion Cloud Deal With Riot ($RIOT) Platforms' — is a NEW-tagged trending ticker with a concrete catalyst, and cross-feed backs it (consensus BUY, 3 BUY / 1 SELL, analyst target +52%). Problem: RIOT is *down* -3.7% overnight and I'm already -6.4% on 15 shares at 28% of book in crypto_mining — my only theme at/near the discretionary risk zone. So no add there; I let the mandatory stop do its job (EXIT OVERRIDE says don't reflexively cut it either, and cross-feed is still BUY). Instead I tighten nothing on RIOT — its stop already sits 2.1% under price, at/inside its ±2.1% daily-move floor, so ratcheting further would be a coin-flip wick-out. Where I do want the cash: NVDA is my best runner (+11.6%) but CAP-FULL, so no add. CRWV likewise CAP-FULL despite the +12% revenue-doubling post (#19) and net +1.50 interpreted read. IREN is the fresh independent setup: mention-leader (+4 today), interpreted net +1.50 with a concrete catalyst ('$IREN delivers first 50MW to $MSFT'), and cross-feed consensus BUY driven by an INDEPENDENT feed — short_interest 28.4% of float, ratio 2.1. That is WSB hype on a genuinely high-short name, my A+ setup, and the independence test passes (short_interest is not Reddit-derived). My last cycle's IREN buy was rejected purely on cash sizing, not thesis — 5 shares at $44.89 = $224, which clears the $200 floor with headroom and fits inside the ~$210 spendable... barely. At ask 44.89, 5 sh = $224.45, above my spendable. So I size 4 shares = $179.56 — that FAILS the fee-guard floor. Rotation is the answer per rule #9: MP is +0.5% flat, 4 shares ≈ $220, in a rare_earth theme with no WSB mention at all today (its BUY is analyst-only, boring-fundamental, not my mandate). Sell MP, buy 6 IREN with the combined proceeds (~$430 of a $384 cap → cap-limited, so 8 shares would breach; 6 sh = $269, safely inside 30%). Stop discipline: IREN's daily swings are violent (39.75 → 44.89 in three sessions, ~5-6% typical), so a tight stop is noise-bait — 12% under fill gives the squeeze room to breathe while capping the live downside, per the volatility anchor. NVDA at +11.6% earns a ratchet toward the ladder: trail 12% under price would be looser than its existing 3.6% stop, so engine ignores it; I leave NVDA alone since its stop is already tighter than the ladder and well outside its ±1.4% daily noise. CRWV I nudge from 5.5% to 5.0%? No — that's inside its ±4.2% daily move only marginally and PnL is just +1.1%; not earned. Leave it. RIOT stays despite -6.4% and the -3.7% overnight gap: cross-feed is still consensus BUY (analyst target +52%) and the Anthropic $9B cloud deal post is today's NEW-tagged catalyst, so the EXIT OVERRIDE applies and the mandatory stop at 19.16 caps the damage — I will not average down into a gap-down on an already 28%-weighted theme. NVDA (+11.6%) and CRWV (+1.1%) are both CAP-FULL so adds are impossible regardless of the $500B-financing and revenue-doubling posts; their existing stops (3.6% and 5.5% under price) are already tighter than the trailing ladder would set and comfortably outside each name's daily noise band, so no ratchet is earned this cycle. Skipped MSFT/MU/TSM despite strong reads — flagged UNTRADEABLE at min size. Skipped PLTR (net +0.92 but 'meme' thesis with the author having SOLD, and no independent feed beyond a Reddit-derived wsb vote) and HTZ (80.6%-style squeeze fuel but the only interpreted post is a -$80k loss-porn bagholder, and it gapped -8.6% into a risk-on tape — that divergence is a red flag, not a dip).

Horisont: 2-4 weeks Begrundelse: WSB post #24 'CoreWeave stock pops 12% as revenue doubles on accelerating AI infrastructure demand' plus cross-feed consensus BUY (analyst target $138.51) — confirmed print, not a pending one; exit if AI-infra capex narrative cracks or it loses the pre-gap $90 shelf. === Session-analyse === [saxovegas/US] WSB's #25 post is the actual live catalyst for a name I already own: 'Anthropic Strikes $9 Billion Cloud Deal With Riot ($RIOT) Platforms' (▲763) — RIOT tagged NEW in trending, cross-feed consensus BUY (3 BUY / 1 SELL, analyst target $30.88 = +53%), and it gapped +1.8% into the open. That's WSB hype + an independent feed (analyst_ratings) + a concrete news catalyst. But my position-cap headroom on RIOT is roughly 19 shares total and I hold 15 — yesterday's 20-share add was already rejected, and crypto_mining is at 41% of a 50% sub-theme cap. So the disciplined move is NOT to add more RIOT (an add of 3-4 shares is only ~$80, well below the $200 fee-guard floor anyway → auto-reject). Instead I let the existing 15 shares ride under a stop and spend the real cash elsewhere. The cleanest fresh signal today is the AI-infra overnight verdict: CRWV +17.8% on the CoreWeave post (#24, ▲778, 'revenue doubles on accelerating AI infra demand') with cross-feed consensus BUY (3 BUY, analyst target 138.51 = +53% upside), and NBIS +23.8% into its earnings print TODAY. NBIS I will not touch — earnings-day chase at +24% gap is exactly the low-edge binary that my HORIZON-FACIT says I bail out of in 5 days. CRWV is the better vehicle: it already reported, the news is confirmed rather than pending, and the move is fundamental (revenue) not just sentiment. Still, +17.8% is a real chase risk — so I size to a single share ($108, below the $200 floor → rejected) ... that doesn't work either. With ~$522 spendable and a $200 minimum, 3 shares CRWV ≈ $324 is the one clean-sized leg available, and it fits under the 30% cap (~$386). Regime is RISK-ON (QQQ +0.79%), VIX calm at 15.3 — the tape supports one leaning-in trade. I explicitly weigh the chase risk by taking only 3 shares (25% of portfolio, not max) and setting a wide 12% stop, since CRWV's daily swing is routinely 5-8% and a tight stop here would be pure noise-harvesting. Ratchets: NVDA is +10.8% with the #9 post ('Nvidia lines up $500B in financing', interpreted net +1.50, high-conviction) still live — not trimming a winner, but its stop at 4.0% under price is already inside a sensible band given avg dagsmove is only ±1.4%, so I'll nudge it to 3.5%. MP is -1.6% but cross-feed consensus BUY (analyst target $75, +36%) so per the EXIT OVERRIDE and rule #4 exception it stays; stop already 6.1% under with ±3.1% daily move — leave it, tightening would be inside the noise band. RIOT stop at 9.8% under price with ±2.0% daily move has too much slack given the $9B Anthropic catalyst should support the price; I raise it to 7% under current. RIOT: holding 15 shares, not adding — the $9B Anthropic deal (WSB #25, NEW tag) is a genuine catalyst and cross-feed is consensus BUY, but position-cap headroom is ~19 shares total and any 3-4 share add is ~$80, far below the $200 fee-guard floor; crypto_mining is also already 41% of the 50% theme cap. Stop raised to 7% under price instead. MP: down -1.6% and forgotten by WSB, but cross-feed consensus BUY (4 BUY, analyst target $75) triggers the rule #4 exception and the EXIT OVERRIDE defers reflexive forgotten-cuts; stop stays at 6.1% under since avg dagsmove is ±3.1% and tightening further would be a coin-flip wick-out. NVDA: +10.8% winner with the $500B-financing post and interpreted net +1.50 behind it — no trim, just a modest ratchet to 3.5% (its daily move is only ±1.4%, so that clears the GAP-FLOOR). Passed on NBIS despite the +23.8% gap and net +1.50 interpreted read: it prints TODAY, so a fresh entry is a coin-flip on the reaction, not a thesis. Passed on WEN (+15.8% gap, only 3 comment mentions, no cross-feed line) and SMCI (+13.3% but I already lost -158 DKK on it this month with no changed thesis). What would change my mind next session: RIOT breaking above $22 with follow-through volume on the Anthropic deal (I'd sell the MP stub to fund a cap-legal add), or NBIS holding its post-earnings gap for a day so the print risk is behind it.

Horisont: 2-4 weeks Begrundelse: WSB News post 'Anthropic Strikes $9B Cloud Deal With RIOT' (762 score, NEW trending) plus independent cross-feed BUY (analyst target +53%, high short interest) — exit if the deal narrative is walked back or price loses the $19.40 shelf. === Session-analyse === [saxovegas/US] Risk-ON tape today (QQQ +0.87%, VIX 15.5 calm) and the AI-infra complex is ripping overnight: CRWV +19.4% on doubled revenue (post #24), NBIS +16.7% into its print today, SMCI +16.2%, SOXL +9.2%. The WSB feed is thin on fresh single-name conviction — top posts are kangaroo-market memes, payroll shock, loss porn and a MSFT gain post. But two genuinely tradeable signals stand out: (1) $RIOT is the NEW trending name with a real, concrete catalyst — 'Anthropic Strikes $9 Billion Cloud Deal With Riot Platforms' (762 score, News flair). That is a fundamental re-rating catalyst on a name where the cross-feed shows consensus: BUY (4 BUY / 1 SELL) with independent corroboration: analyst_ratings +0.87 (target $30.88, +53%) plus short_interest. WSB hype + independent analyst + short squeeze fuel = my A+ setup archetype. Last cycle my RIOT buy was REJECTED purely on the 30% position cap at 20 shares — the thesis wasn't wrong, the size was. Resize to 15 shares (~$304, well under the ~$387 cap) and it goes through. (2) NVDA: the interpreted read is +1.50 net-sentiment, high-conviction, concrete catalyst — the $500B financing/'chips are an investable asset' post (2,534 score). But NVDA is flagged CAP-FULL, so no add possible; I ratchet its stop instead. It's +9.6% with a stop already 2.9% under price and avg dagsmove only ±1.4% — that stop is already tight relative to noise and sits above breakeven. I'll leave it (2.9% is already inside the ladder's +15% step and near the gap-floor; tightening further into last-cycle noise is really a market order). MP: -1.4%, gapped -2.4% overnight, but cross-feed consensus: BUY (4 BUY, 0 SELL) with analyst target $75 (+36%). Rule #4 is deferred on this tier anyway and the cross-feed is net-bullish — hold, stop stays at 50.54 (6.3% under, ~2x the ±3.1% daily move, correctly outside the noise band). I am NOT chasing NBIS or CRWV: NBIS prints TODAY (earnings watch, 0 days) after a +16.7% gap — buying a parabolic gap into a print is exactly the chase-risk I keep paying for. CRWV +19.4% is the same top-of-gap problem, and it would pile more into ai_infra which is already 25%. RIOT is crypto/AI-datacenter adjacent, a fresh sub-theme for this book with a hard-news catalyst rather than a momentum gap, and cash is only ~$827 so one clean bet is the right size. Stop at 11% — RIOT swings ~3-4% daily so ~1.5-2x that band, anchored below the $19.40 recent low; a $9B contract thesis breaking means it lost the level entirely. Holding MP (cross-feed consensus BUY, 4-0, analyst +36% upside — rule #4 cut is deferred on this tier and the external vote is net-bullish; stop at 6.3% is correctly outside its ±3.1% daily noise). Holding NVDA — flagged CAP-FULL so no add is possible despite the strongest interpreted read of the day (+1.50, $500B financing catalyst); its stop already rests 2.9% under price with only ±1.4% avg daily move, so it is already ratcheted about as tight as the gap-floor allows and tightening further would be a de-facto market sell. Explicitly NOT chasing NBIS (+16.7% gap INTO tonight's earnings print), CRWV (+19.4% parabolic gap) or SMCI (+16.2%, and I already lost -158 DKK round-tripping it) — those are top-of-gap chases into an already-25% ai_infra theme. Cash after RIOT is near zero, which is fine: one clean high-conviction catalyst bet beats three low-edge scalps that only feed the 0.37% fee drag.

Horisont: 2-4 years Begrundelse: Adding to an existing SMR/nuclear position on strong 5d momentum and 3-1 BUY consensus, while position size (9% of book) leaves ample room below the 35% cap and the structural datacenter-baseload thesis is unbroken. === Session-analyse === [saxostrategi/ALL] Cash is scarce (7% of book) and the DK buy window is structurally closed this cycle — no DK holdings to sell, so no path to the 1,160 DKK floor there; skipping Danish names entirely rather than forcing a bad trade. On the US side, TSM/LRCX/AMAT/MU/GEV/PWR are cap-blocked (correctly — I hold winners, not touching them), leaving CDLR, NEE, and OKLO as the only legal sizes. CDLR's tape is weak (marine_installation +1.4%/-3.6% 30d) with no catalyst; NEE at 1 share is fee-inefficient and already fairly priced (analyst upside modest). OKLO stands out: cross-feed consensus is BUY (3-1), 5d momentum +11% on genuinely improving small_modular tape (+7% 5d), and it's still only 9% of the book — room to add without approaching any cap. This is 'adding to a name already working,' not a fresh chase, and the SMR/nuclear structural thesis (dispatchable baseload for datacenters) is intact and if anything strengthening on the uranium/nuclear tape (+6%/+3.9% 5d despite negative 30d — a base building, not a blow-off like ARM/AMD at 96-99% range position). TSM (34%) is at the position cap and should not be added to; LRCX, VST, CCJ theses are all intact with no new information to act on — holding through normal sector noise per mandate. DK buy window is closed this cycle (no DK holdings to sell to reopen it, and forcing a sub-floor trade would just burn fees). Cash remains thin (7%) by design after the OKLO add; waiting for the book to compound before deploying more into cap-blocked names like TSM/LRCX/AMAT/MU rather than chasing extended momentum names (AMD, ARM) at 96-99% of their 10-day range.

Horisont: 2-3 years Begrundelse: Uranium is the physical constraint on the nuclear baseload thesis inside ELECTRICITY, and CCJ is the least-diluted, most liquid pure-play producer to add alongside OKLO/VST without touching capped semis names. === Session-analyse === [saxostrategi/ALL] The book is small (8,588 DKK) and already concentrated: TSM at 33% and LRCX at 22% are both effectively capped from further buying, and semis/ai_infra exposure sits at 55%. Cash is only ~1,215 DKK, which after fee-guard math gives a single US buy window of roughly $40-179. Memory and semicap tape is hot (+7.5%/+6.6% 5d) and cross-feed consensus on MU, LRCX, AMAT is uniformly BUY with strong analyst upside, but MU's per-share price (~$1000+) and LRCX's cap-block rule those out for new buys today. CCJ and LEU are legal-sized single-share adds with strong tape momentum (uranium/nuclear feed strength, LEU's short-interest signal at 0.83) but LEU has already run +23% over 10 sessions off a low base — chasing that vertical move into a 1-share position is a bad risk/reward for a name I don't yet hold, better to wait for a pullback. CCJ is a cleaner story: uranium/nuclear structural thesis intact, price up modestly (92→102, +10% over the window) without the parabolic look, and it's not yet a position, giving genuine diversification within the nuclear sub-theme without touching my capped semis exposure. VST carries an edgar SELL flag (dilution filing + insider sales) and is my only red position (-2.3%), but the thesis — VST as nuclear/grid/data-center-adjacent generation capacity — is not broken by a routine shelf filing; a couple of insider sales is weak-signal noise per the feed's own caveat. I'm not adding to it (single-share cap anyway) and not selling on a filing that doesn't invalidate the structural case; a bad week is not a broken thesis. Given the fee-floor math, one clean US buy (CCJ, 1 share ≈ $102, well within the $40-179 window) is the sensible single action today — sized small because that's what the cash allows, not because conviction is low. No DK trades: no Danish tickers in the tradeable universe roster of names I want exposure to right now, and forcing a DK leg just to use both windows would be churn, not thesis. TSM and LRCX are both past or near the 35% concentration ceiling and cap-blocked for new buys regardless of how attractive analyst targets look (TSM +27%, LRCX +30% upside per feeds) — that's a function of book size, not a signal to trim; winners compound. VST's edgar dilution flag is real but a routine 424B5/S-3ASR shelf plus a couple of insider sales does not break a grid/nuclear generation thesis, and the position is too small to trim economically anyway — holding through the noise. Not chasing LEU's vertical +23% run into a first position; will revisit on a pullback. Staying in one clean CCJ add today given the tight $40-179 US buy window and reserving remaining cash rather than forcing a second, worse-sized leg.

Horisont: 2-3 years Begrundelse: Small modular reactor thesis is a distinct nuclear sub-theme from VST's regulated-utility exposure — datacenter operators need dispatchable baseload before 2030 and OKLO is the clearest pure-play SMR name in the universe, exit if NRC licensing timeline slips materially or dilution accelerates beyond routine. === Session-analyse === [saxostrategi/ALL] Cash is tight (~1,970 DKK), and the three big cap-blocked winners (TSM, LRCX, VST) already carry the book's real exposure — TSM at 33% is near the ceiling, VST at 22%, LRCX at 21%. No thesis has broken: semicap/memory/nuclear are all down 5-13% over 30d on sector noise, not fundamentals, and TSM's own analyst consensus still shows +28% upside with 2 BUY votes. That's a 'hold the winner' case, not an add case given concentration. On the buy side, the only names with real room under the cap are smaller legs (OKLO, LEU, CCJ, NVDA-1-share, MRVL-1-share, etc.), and I want a name with a distinct thesis rather than more overlap with what I already own. VRT stands out: analyst consensus BUY +0.80 with 40% upside, data-center/grid power-delivery thesis distinct from LRCX/TSM's fab-equipment and foundry exposure, and it's down from its 290 high to 280 on tape noise (grid basket -4.4% 5d) not a broken thesis. But VRT isn't in the current legal BUY WINDOW list for cash reasons (not listed with a share range), so I'll skip it this cycle rather than force something. Instead, I'll add a single small OKLO position — small modular nuclear is a distinct sub-theme (SMR, +7.2% 5d, contrarian to the broader nuclear basket's -1.5%/-6.6%, and only 22% nuclear exposure currently all in VST which is large-cap regulated + retail). OKLO's edgar signal (dilution/insider sales) is a real risk given the 424B5 shelf filing, so I'll size this modestly (3 shares, ~$124) and set a wider stop given OKLO's volatility. This diversifies nuclear away from single-name VST risk without breaching cash constraints, leaving ~1,150 DKK cash buffer intact. TSM (33%), VST (22%) and LRCX (21%) already carry the book's thesis and TSM/LRCX/MU/GEV/PWR/AMD are cap-blocked from further buys — adding here would just concentrate further into names already near the ceiling, not diversify. The 5-13% pullbacks in semicap/memory/nuclear look like sector noise (VIX 17.7, no earnings surprise, no policy shift) rather than a broken thesis, so no sells either. Keeping the single OKLO add small and preserving cash buffer since today's window is thin and this account trades once a day with no ability to revise stops later.

Horisont: 2-3 years Begrundelse: Independent power producer with direct leverage to datacenter/AI electricity demand and strong analyst upside (+47% to target); exit if PJM capacity price reform or a large customer contract cancellation breaks the power-shortage thesis. === Session-analyse === [saxostrategi/ALL] This is a fresh, all-cash book on day one — the starting basket matters more than any single day's noise. The tape is broadly red across every theme (nuclear -8.8%/30d, semis -7.2%, memory -7.3%), which for a multi-quarter structural thesis is an entry opportunity, not a warning. I want to seed all three legs (electricity, semis, memory) with names that have both a stated thesis and cross-feed support, while respecting the cash budget and avoiding the CAP-BLOCKED names (AMD, GEV, MU, PWR are blocked this cycle regardless of sizing). For semiconductors, TSM stands out: 31 insider buys totaling ~$276k is a strong signal of management conviction at these levels, and analyst consensus is BUY. For memory/semicap, LRCX is down hard (5d -10%, range_pos 0%) — the memory/semicap cycle thesis is intact (HBM/DRAM demand from AI buildout hasn't changed), and this is a battered entry into a name I want structural exposure to, not a broken thesis. For electricity/grid, VST has strong analyst support (+47% upside, 1.4/5 rating) and nuclear-adjacent power generation exposure that benefits from datacenter power demand — CEG's single SELL vote and steep 5d drop makes me prefer VST as the cleaner entry into the same theme today. Sizing: USD budget is ~$1,251. TSM at ~$418 (1 share, capped anyway) ≈ $418. LRCX capped at 1 share ≈ $301. VST 1-3 shares available; I'll take 2 at ~$143.79 ≈ $288 to keep some room. Total ≈ $1,007, leaving buffer under the $1,251 ceiling and well within the DKK cash converted. Stops are set at 20-22% — wide enough to ride out normal 5-8% sector noise cited in the mandate, tight enough to protect capital if a thesis actually breaks. No DK-listed name in the universe today, so no .CO trade this cycle. Seeding the book today across all three mandate legs — semiconductors (TSM), semicap/memory-adjacent (LRCX), and grid/electricity (VST) — using cash that would otherwise sit idle. GEV, AMD, MU, PWR are cap-blocked this cycle so left untouched; CEG's sharp single-feed SELL and steep drop makes VST the cleaner grid entry today. No DK-listed names in today's universe, so the Copenhagen leg stays empty this cycle.

Horisont: 2-3 years Begrundelse: Semicap equipment maker sold off to 5-day/30-day lows despite intact memory/logic capex upcycle thesis; exit if fab capex spending across TSM/Samsung/Micron actually contracts. === Session-analyse === [saxostrategi/ALL] This is a fresh, all-cash book on day one — the starting basket matters more than any single day's noise. The tape is broadly red across every theme (nuclear -8.8%/30d, semis -7.2%, memory -7.3%), which for a multi-quarter structural thesis is an entry opportunity, not a warning. I want to seed all three legs (electricity, semis, memory) with names that have both a stated thesis and cross-feed support, while respecting the cash budget and avoiding the CAP-BLOCKED names (AMD, GEV, MU, PWR are blocked this cycle regardless of sizing). For semiconductors, TSM stands out: 31 insider buys totaling ~$276k is a strong signal of management conviction at these levels, and analyst consensus is BUY. For memory/semicap, LRCX is down hard (5d -10%, range_pos 0%) — the memory/semicap cycle thesis is intact (HBM/DRAM demand from AI buildout hasn't changed), and this is a battered entry into a name I want structural exposure to, not a broken thesis. For electricity/grid, VST has strong analyst support (+47% upside, 1.4/5 rating) and nuclear-adjacent power generation exposure that benefits from datacenter power demand — CEG's single SELL vote and steep 5d drop makes me prefer VST as the cleaner entry into the same theme today. Sizing: USD budget is ~$1,251. TSM at ~$418 (1 share, capped anyway) ≈ $418. LRCX capped at 1 share ≈ $301. VST 1-3 shares available; I'll take 2 at ~$143.79 ≈ $288 to keep some room. Total ≈ $1,007, leaving buffer under the $1,251 ceiling and well within the DKK cash converted. Stops are set at 20-22% — wide enough to ride out normal 5-8% sector noise cited in the mandate, tight enough to protect capital if a thesis actually breaks. No DK-listed name in the universe today, so no .CO trade this cycle. Seeding the book today across all three mandate legs — semiconductors (TSM), semicap/memory-adjacent (LRCX), and grid/electricity (VST) — using cash that would otherwise sit idle. GEV, AMD, MU, PWR are cap-blocked this cycle so left untouched; CEG's sharp single-feed SELL and steep drop makes VST the cleaner grid entry today. No DK-listed names in today's universe, so the Copenhagen leg stays empty this cycle.

Horisont: 3+ years Begrundelse: Leading-edge foundry monopoly on AI/HPC silicon with 31 recent insider buys signaling management conviction at current levels; exit if foundry share loss to Samsung/Intel becomes real or capex guidance breaks down. === Session-analyse === [saxostrategi/ALL] This is a fresh, all-cash book on day one — the starting basket matters more than any single day's noise. The tape is broadly red across every theme (nuclear -8.8%/30d, semis -7.2%, memory -7.3%), which for a multi-quarter structural thesis is an entry opportunity, not a warning. I want to seed all three legs (electricity, semis, memory) with names that have both a stated thesis and cross-feed support, while respecting the cash budget and avoiding the CAP-BLOCKED names (AMD, GEV, MU, PWR are blocked this cycle regardless of sizing). For semiconductors, TSM stands out: 31 insider buys totaling ~$276k is a strong signal of management conviction at these levels, and analyst consensus is BUY. For memory/semicap, LRCX is down hard (5d -10%, range_pos 0%) — the memory/semicap cycle thesis is intact (HBM/DRAM demand from AI buildout hasn't changed), and this is a battered entry into a name I want structural exposure to, not a broken thesis. For electricity/grid, VST has strong analyst support (+47% upside, 1.4/5 rating) and nuclear-adjacent power generation exposure that benefits from datacenter power demand — CEG's single SELL vote and steep 5d drop makes me prefer VST as the cleaner entry into the same theme today. Sizing: USD budget is ~$1,251. TSM at ~$418 (1 share, capped anyway) ≈ $418. LRCX capped at 1 share ≈ $301. VST 1-3 shares available; I'll take 2 at ~$143.79 ≈ $288 to keep some room. Total ≈ $1,007, leaving buffer under the $1,251 ceiling and well within the DKK cash converted. Stops are set at 20-22% — wide enough to ride out normal 5-8% sector noise cited in the mandate, tight enough to protect capital if a thesis actually breaks. No DK-listed name in the universe today, so no .CO trade this cycle. Seeding the book today across all three mandate legs — semiconductors (TSM), semicap/memory-adjacent (LRCX), and grid/electricity (VST) — using cash that would otherwise sit idle. GEV, AMD, MU, PWR are cap-blocked this cycle so left untouched; CEG's sharp single-feed SELL and steep drop makes VST the cleaner grid entry today. No DK-listed names in today's universe, so the Copenhagen leg stays empty this cycle.

Horisont: 1-2 weeks Begrundelse: Cross-feed BUY consensus (2 BUY, 0 SELL) riding the MU earnings-beat semis tailwind WSB is buzzing about today, without adding to an already-losing position. === Session-analyse === [wsb/US] MU dominates WSB today (RISING tag, z=+3.0σ, 74 comment-flow mentions, strong post titles about the earnings beat) and cross-feed backs it with a BUY consensus (analyst 1.4/5, target +43% upside) — but MU itself isn't in our tradeable universe at this size (flagged UNTRADEABLE at min-size), so the only way to play the theme is via adjacent names. AVGO and AMD both carry cross-feed BUY consensus and sit in semiconductors, a theme already at 13% exposure (room to add) riding +10.5% 30d tape — but ai_infra/semis combined are already 46%+13%=59% if we stack more, so I'll be selective. AMD is untradeable at min size per the UNTRADEABLE list, leaving AVGO as the cleanest semis add with 2 BUY cross-feed votes and no SELL. Existing book: IONQ (+9.8%) has fresh cross-feed BUY backing (analyst +0.89, +52% target upside) despite quantum tape being soft (-3.8% 5d) — worth tightening the stop rather than adding, since quantum theme is already 20%. SMCI (+2.7%, bought last cycle) has cross-feed BUY confirmation too, holding. GOOGL is CAP-FULL so no action there. NVDA is CAP-FULL and basically flat — holding, gap-up overnight modest. RDDT gapped +3.6% overnight on no fresh WSB catalyst (not in trending/mentions this session) and is down -3.6% from my avg — cross-feed wasn't shown for RDDT so no override; it's a small, single-share, low-conviction legacy position I'll leave alone rather than force an exit on a position this small (fee drag not worth it either way). New money: with RISK-ON regime (QQQ +0.44%), VIX calm at 16, and cash at 25% of book, I'll deploy into AVGO (cross-feed BUY, semis tailwind, NOT yet in my book so no theme-stacking on an existing loser) sized modestly given its high per-share price limits shares to 1. This is a single-share high-conviction add per the exception rule for names ≥$50. I'm skipping NKE (earnings today, WSB sentiment net-bearish per INTERPRETED signal '-1.00 skeptical of NKE beating earnings' — avoid pre-print chase) and skipping MU-adjacent plays beyond AVGO to avoid conviction-leakage into an already crowded ai_infra/semis book this late in the day (cycle 3 of 4, last cycle later today per rule #19 raises bar further). GOOGL and NVDA are CAP-FULL — no room to add regardless of signal. RDDT is a single-share legacy loser with no fresh WSB mention or cross-feed row shown this cycle; too small to matter either way on fees, leaving it alone rather than forcing a cut. MU itself is untradeable at our size so I'm expressing the theme via AVGO only, deliberately avoiding stacking further into ai_infra/semis which is already the dominant exposure (46%+13%=59% combined) this late in the day (cycle 3 of 4) when a fresh overnight-hold needs a real catalyst, not just hype.

Horisont: 1-2 weeks Begrundelse: MU's blowout earnings (WSB's #1 hot topic today, 2,659-score YOLO + multiple earnings threads) is lifting the whole memory/AI-infra semis tape and SMCI carries its own cross-feed BUY consensus with short-interest support — MU itself is untradeable at min size so this is the cheaper access point to the same catalyst. === Session-analyse === [wsb/US] MU is the clear WSB story today — RISING tag (z=+3.0σ), top of hot with a 2,659-score YOLO post plus a cluster of earnings-beat threads (18, 19, 21), net-sentiment +0.53 with real skin-in-the-game (3 shares/1 calls disclosed) and a 5/1 bull split. Cross-feed backs it: consensus BUY (2 BUY incl. analyst_ratings +0.83 target +43% upside). That clears the ≥2 BUY-vote corroboration bar cleanly. MU itself is untradeable at min size per the UNTRADEABLE list (too expensive, ~$1,015/share) — so I can't play the actual ticker, but AMD and SMCI sit adjacent in the same semiconductors/AI-infra theme-tape (semis 5d +1.3%, 30d +10.7%, tailwind) with their own cross-feed BUY consensus (AMD 2 BUY/3 HOLD, SMCI 2 BUY/2 HOLD short_interest-backed). AMD is flagged untradeable at min size too, so SMCI is the real access point — cheap enough for a real scale-out position, cross-feed BUY, and rides the same earnings-beat tailwind without chasing MU's own after-earnings pop. Existing book: IONQ is my best winner (+8.1%), cross-feed BUY (analyst target +52% upside) despite quantum theme-tape being soft (5d -3.7%) — I'll hold and nudge the stop up since it's not yet at the +15% ladder trigger. RDDT dipped -6.2% but still shows up in comment-flow ($RDDT 4✓) and gapped +0.9% overnight — not yet 'forgotten' by rule #4, and no cross-feed SELL consensus on it, so holding rather than cutting into a neutral tape. GOOGL and NVDA are CAP-FULL, no adds possible regardless of the Gemini 4 buzz (post #12, 816 score) — just holding. Quantum theme is already at 20% and ai_infra at 32%, approaching the 50% cap collectively, so I'm deploying new cash into semiconductors (13% current, room to grow) rather than piling onto saturated themes. NASDAQ is neutral today (QQQ -0.05%), VIX calm at 16 — no regime reason to go big, so sizing SMCI modestly within the $200+ min-trade floor. GOOGL and NVDA are CAP-FULL (position-cap blocks any add) despite the Gemini 4 buzz (post #12) and NVDA's AI-datacenter mention — holding as-is. RDDT down -6.2% but still appearing in comment-flow and no cross-feed SELL, not yet a clear 'forgotten bagholder' cut under rule #4. Quantum (IONQ) and ai_infra (GOOGL/NVDA) themes are already sizable (20%/32%) so new cash is routed to semiconductors instead of adding exposure there. Sitting mostly in cash otherwise — WSB's real fresh energy today is concentrated in MU, which the sizing rules block directly, so a single modest semis trade is the deliberate, non-forced move this cycle.

Horisont: 1-2 weeks Begrundelse: China/ByteDance-Alibaba Nvidia chip-buying news post (1,048 upvotes) plus INTERPRETED signal +0.75 net-sentiment with disclosed positions and 3/4 catalyst-cited posts, backed by cross-feed consensus BUY; sizing small given RISK-OFF tape and the name already gapped +3.3% overnight. === Session-analyse === [wsb/US] Today's WSB feed is genuinely NVDA-heavy: the China/Nvidia chip news post is high-scoring (1,048 upvotes) and the INTERPRETED signal shows +0.75 net-sentiment with 3/4 posts citing a concrete catalyst and actual disclosed positions (2 shares, calls) — this is real conviction, not drive-by hype. Cross-feed backs it up: NVDA consensus BUY (3 BUY/1 SELL/2 HOLD) with analyst_ratings at +0.88 (+46% upside target). NVDA also gapped up +3.3% overnight on the China chip-buying news, which is momentum confirmation. However, it's RISK-OFF today (QQQ -0.78%, SPY -0.57%) and NVDA already ran on the gap, so I'm sizing modest rather than full — this is a smaller starter position with a wider stop to respect the gap risk, not a chase into full size. META is also loud (7 mentions today, highest mention-leader, +1.33 net-sentiment, 2 high-conviction posts with disclosed skin-in-the-game) but it is UNTRADEABLE at min size per this session's params — skip it, no way around the constraint. IONQ continues to work (+16.9% unrealized, cross-feed consensus BUY with analyst target +48% upside, quantum theme +5.2% over 5d) — I'll let this winner run and tighten the trailing stop per the ladder rather than trim it. RDDT is down -4.2% and gapped down another -2.1% overnight with no fresh WSB catalyst today (only comment-flow mentions, no trending/DD presence) — but it still shows in comment-flow with real engagement and I don't have 2+ cross-feed BUY votes to hold conviction on adding; I'll hold it flat for now rather than force an exit at a 4% loss on a name that hasn't cleared the 10% forgotten-bagholder threshold yet. GOOGL is CAP-FULL per constraints — cannot add. Given RISK-OFF regime, calm VIX (14.2) but crypto Fear&Greed at 74 (Greed, blow-off risk for meme/crypto names), I'm avoiding fresh crypto-mining or meme-stock entries entirely (matches my own THEME-FACIT: crypto_mining has been a -1688 DKK loser across 17 exits — no fresh evidence justifies re-entering that theme). Staying disciplined: one modest NVDA add, hold everything else, tighten IONQ's stop to lock in the +16.9% gain. Holding RDDT flat — down -4.2% and gapped -2.1% overnight with no fresh WSB catalyst today (comment-flow only, not trending), but it hasn't crossed the >10% forgotten-bagholder threshold yet so I'm not forcing a cut into a red tape day. Holding GOOGL (CAP-FULL, can't add anyway, small +1.5% gain intact). Sitting on 52% cash given RISK-OFF regime (QQQ -0.78%) — no crypto-mining or meme names cleared the bar today, and my own THEME-FACIT shows that theme has been a consistent loser. META and AMZN both show strong INTERPRETED sentiment but META is untradeable at min size this session and AMZN has no cross-feed corroboration shown — passing on both rather than forcing a low-conviction entry.

Horisont: 1-2 weeks Begrundelse: The $146K RDDT Jan 2028 calls YOLO post plus cross-feed consensus BUY (analyst target +40% upside) back a small momentum add in the social theme tape running +9% over 5 days. === Session-analyse === [wsb/US] WSB's hot list today is thin on fresh, tradeable conviction: NKE is pure loss-porn meme content (net-sentiment -0.25, thesis: loss_porn), META has genuine momentum (net +0.92, skin-in-the-game 3-share disclosure, cross-feed BUY with wsb as top signal) but I already have zero exposure there and it's not in my current sub-theme mix — worth a small starter given RISK-ON regime + cross-feed corroboration. RDDT has a single $146K YOLO calls post (100% loss-porn framing though, thin sample = weak signal per rule #15) but cross-feed shows consensus BUY (analyst_ratings +0.64, 2 BUY/0 SELL/3 HOLD) — decent secondary corroboration for a small position. Crypto mining names (CLSK, RIOT, WULF, IREN) show strong cross-feed BUY consensus, but my THEME-FACIT shows -1770 DKK across 18 exits in this exact theme, WULF just stopped out yesterday (re-entry blocked), and IREN/COIN are also on my recent-exits cooldown list — this is a theme I've bled money on repeatedly with no theme-external catalyst, so I'm passing despite the cross-feed noise. Existing holdings: GOOGL is at 25% cap already (CAP-FULL, no adds allowed) — holding as-is, small +2.1% gain, quantum/internet/mega_cap themes fine. IONQ is up +10.3%, cross-feed shows consensus BUY (3 BUY/0 SELL/1 HOLD, analyst target +49% upside), quantum theme showing +8.6% 5d tape strength — this is a winner to protect with a trailing stop rather than trim, per rule #2. QQQ is green today (RISK-ON +0.36%), VIX is calm at 14.21, so leaning in on quality names makes sense, but I'm being selective given the thin WSB catalysts today — most of the hot list is macro-macro chatter (bonds, yields, Bessent) rather than actionable single-name hype. Trades: small META starter (cross-feed BUY consensus, wsb sentiment strongly positive, RISK-ON tape) and a small RDDT position (cross-feed BUY, calls YOLO is on-theme for social momentum tape +9.0% 5d). Both sized modestly given thin sample sizes and no ticker in my top-3 trending list clearing every bar. Raising IONQ's stop to lock in gains as it approaches the +15% ladder threshold. GOOGL is CAP-FULL at 25% position size — no adds allowed regardless of WSB chatter (Google's Gemini 4 news is a mild positive but the position is maxed). Crypto mining names (WULF/IREN/CLSK/RIOT/MARA) show strong cross-feed BUY votes but my THEME-FACIT shows -1770 DKK across 18 round-trips in this exact theme with zero improvement, WULF just stopped out yesterday and IREN is on cooldown from a recent exit — repeating this bet without a theme-external catalyst is exactly the mistake the facit warns against, so I'm sitting out despite the signal. NKE's WSB post is pure loss-porn mockery, not a thesis. Staying mostly in cash (61%+) given today's hot list is dominated by macro/Bessent/bond chatter rather than fresh single-name catalysts — the two small adds (META, RDDT) are the names that cleared both WSB sentiment and cross-feed corroboration bars.

SOFT STOP TRIGGERED: WULF faldt under 8% stop (17.11 → 16.03). Auto-cut før agent-call.

Horisont: 1-2 weeks Begrundelse: Crypto_mining theme tape +9.9%/5d and +17.5%/30d with cross-feed 4-BUY/0-SELL consensus (analyst target +102% upside) — adding to an already-green position on strong external corroboration, not just WSB hype; exit if theme tape rolls over or WULF loses the $16 level. === Session-analyse === [wsb/US] WSB is entirely META/SPX-options theater today (irony-heavy, mixed bull/bear, 50% loss-porn) — not tradeable, META isn't even in my universe. AMD is genuinely hot (trillion-dollar cap, 5-day rally, z=+2.5σ, cross-feed BUY with 5d +24.7%) but it's priced at $615+ and flagged UNTRADEABLE at min size for this account — skip. That leaves my actual holdings: WULF has the strongest cross-feed backing I've seen (4 BUY/0 SELL, analyst target +102% upside, short interest fuel) and crypto_mining theme tape is ripping +9.9%/5d, +17.5%/30d — this is a real tailwind, not just chatter. GOOGL is up nicely (+7.2%) but CAP-FULL at position cap, so no add there, just let it run. IONQ is flat/slightly red, quantum theme is down -17.9% over 30d — no name-specific catalyst here, cross-feed shows BUY (3-1-0) but the theme tape is weak and WSB isn't mentioning it today, so I'll hold without adding. WSB's hot list today is a META options circus (irony sentiment, 50% loss-porn, mixed 9-bull/11-bear split) and a viral SPX 0DTE YOLO post — neither maps to my tradeable universe or a real thesis. AMD is the one legitimately fresh/rising ticker (trillion-cap milestone, cross-feed BUY, +24.7% 5d) but it's flagged untradeable at min size here, so no entry despite the hype. IONQ stays held flat — cross-feed still shows consensus BUY so rule #4's forgotten-name cut doesn't trigger, but I won't add given the quantum theme's ugly -17.9% 30d tape and zero fresh WSB mentions today. Risk-on regime (QQQ +0.35%), calm VIX (14.8), and Extreme Greed on crypto fear/greed (78) support leaning into my highest-conviction existing winner (WULF) rather than chasing new untested names, while tightening stops on GOOGL and WULF to lock in gains as they extend.

Horisont: 1-2 weeks Begrundelse: Fresh +7.5% overnight gap with cross-feed consensus BUY and huge analyst upside target (+131%) — sizing smaller given crypto_mining theme has been this book's worst bleeder historically, exit if gap fades below $15.5 support. === Session-analyse === [wsb/US] RISK-ON tape today (QQQ +1.66%) and the overnight gap sheet is full of green in the AI/quantum/nuclear/mining names (OKLO +12.8%, IONQ +10.2%, SMR +9.0%, WULF +7.5%, ARM +7.5%), but WSB itself is thin on fresh signal — the hot list is dominated by macro news (Fed rate hike, CLARITY Act failure) and loss-porn memes, not a specific ticker rally. The one name that stacks WSB relevance + cross-feed corroboration + a real overnight gap is IONQ: consensus BUY (3-0-2, analyst target +88% upside), INTERPRETED signal net +0.50 with a cited catalyst ('hybrid quantum apps show utility'), quantum theme 5d +8.2%, and it gapped +10.2% overnight on real momentum — not chasing a stale spike. WULF is the other qualifier: consensus BUY (3-0-2), analyst target +131% upside, and a clean +7.5% gap, though I'll size it smaller given the crypto_mining theme has been my worst bleeder (-1650 DKK over 18 exits) — this is a fresh name+catalyst (WULF specifically, not a recycled MARA/RIOT bet) so I'll allow a small toe-in rather than a full pass. GOOGL stays untouched — CAP-FULL per the params, small +2.4% winner, nothing to add or trim. COIN/NVDA/NBIS/MP are all on the re-entry-gated list from recent exits and none of today's cross-feed shows the required 2+ fresh BUY votes plus a genuinely new catalyst, so they stay out per rule #8. Given it's cycle 3 of 4 (not the last cycle yet), I have room to open positions with normal stops rather than needing overnight-survival conviction — but I'm still keeping size moderate since IONQ/WULF are both high-beta names in a book that's been over-concentrated in mining/AI-infra churn historically (THEME-FACIT crypto_mining and ai_infra both net negative). Total new deployment stays well under the 25% single-position cap and doesn't push any sub-theme over 50%. GOOGL is CAP-FULL and only +2.4% — no add possible, no reason to trim a modest winner; holding as-is with no stop tightening needed yet since it hasn't cleared the +15% ladder trigger. Passing on COIN/NVDA/NBIS/MP (recent exits, re-entry-gated, no fresh 2+ cross-feed BUY consensus with a genuinely new catalyst today) and on SPCX/AMD (today's actual top mention-leaders) since neither clears cross-feed corroboration — SPCX INTERPRETED read is split bull/bear with 50% loss-porn, and AMD is UNTRADEABLE at min size per params anyway. Staying selective given RISK-ON tape doesn't mean forcing size into names without the corroboration bar.

Horisont: 1-2 weeks Begrundelse: Overnight +10.2% gap on real quantum-theme momentum (5d +8.2%) plus cross-feed consensus BUY (analyst target +88% upside) and an INTERPRETED post citing a concrete utility catalyst — exit if the gap fully round-trips or quantum theme rolls over. === Session-analyse === [wsb/US] RISK-ON tape today (QQQ +1.66%) and the overnight gap sheet is full of green in the AI/quantum/nuclear/mining names (OKLO +12.8%, IONQ +10.2%, SMR +9.0%, WULF +7.5%, ARM +7.5%), but WSB itself is thin on fresh signal — the hot list is dominated by macro news (Fed rate hike, CLARITY Act failure) and loss-porn memes, not a specific ticker rally. The one name that stacks WSB relevance + cross-feed corroboration + a real overnight gap is IONQ: consensus BUY (3-0-2, analyst target +88% upside), INTERPRETED signal net +0.50 with a cited catalyst ('hybrid quantum apps show utility'), quantum theme 5d +8.2%, and it gapped +10.2% overnight on real momentum — not chasing a stale spike. WULF is the other qualifier: consensus BUY (3-0-2), analyst target +131% upside, and a clean +7.5% gap, though I'll size it smaller given the crypto_mining theme has been my worst bleeder (-1650 DKK over 18 exits) — this is a fresh name+catalyst (WULF specifically, not a recycled MARA/RIOT bet) so I'll allow a small toe-in rather than a full pass. GOOGL stays untouched — CAP-FULL per the params, small +2.4% winner, nothing to add or trim. COIN/NVDA/NBIS/MP are all on the re-entry-gated list from recent exits and none of today's cross-feed shows the required 2+ fresh BUY votes plus a genuinely new catalyst, so they stay out per rule #8. Given it's cycle 3 of 4 (not the last cycle yet), I have room to open positions with normal stops rather than needing overnight-survival conviction — but I'm still keeping size moderate since IONQ/WULF are both high-beta names in a book that's been over-concentrated in mining/AI-infra churn historically (THEME-FACIT crypto_mining and ai_infra both net negative). Total new deployment stays well under the 25% single-position cap and doesn't push any sub-theme over 50%. GOOGL is CAP-FULL and only +2.4% — no add possible, no reason to trim a modest winner; holding as-is with no stop tightening needed yet since it hasn't cleared the +15% ladder trigger. Passing on COIN/NVDA/NBIS/MP (recent exits, re-entry-gated, no fresh 2+ cross-feed BUY consensus with a genuinely new catalyst today) and on SPCX/AMD (today's actual top mention-leaders) since neither clears cross-feed corroboration — SPCX INTERPRETED read is split bull/bear with 50% loss-porn, and AMD is UNTRADEABLE at min size per params anyway. Staying selective given RISK-ON tape doesn't mean forcing size into names without the corroboration bar.

Horisont: closed Begrundelse: CLARITY Act Senate vote failed 49-50 overnight — the exact catalyst the top WSB YOLO post ('Ready for COIN to rippp ahead of senate vote') was betting on is now dead, and COIN gapped -5.8%; full exit rather than holding a broken-thesis position through the overnight gap risk on the last cycle of the day. === Session-analyse === [wsb/US] Last cycle of the day, RISK-OFF tape (QQQ -0.42%, SPY -0.85%), and WSB itself is mostly consumed with Fed-hike memes and CLARITY Act failure rather than any fresh actionable ticker. The CLARITY Act failing 49-50 is a real overnight-relevant negative catalyst for crypto policy names — COIN gapped -5.8% on it and is now down -5.3% unrealized on my position, having already been bought yesterday. WSB is still nominally hyping COIN ('Ready for COIN to rippp ahead of senate vote') but that thesis is now DEAD — the vote failed, the catalyst is spent, and this is exactly the discretionary full-exit case in rule #2: the setup that justified the buy no longer exists. I'm not averaging down into a broken catalyst going into an unprotected overnight hold on the last cycle of the day (rule #19) — cutting it now at a known price beats trailing a stop through the gap-risk night, especially with crypto_exchange theme already bleeding on the 5d tape (-7.6%). GOOGL is a small, already at-cap position (+1.1%, essentially flat) with no fresh WSB catalyst tonight — mention count is thin, and internet/ai_infra/mega_cap themes are all already carrying it fully. No reason to add (capped anyway) or cut (marginal gain, not down >10%, no distress). I'll leave it alone into the close. No new BUY clears the bar tonight: RDDT and FIG have decent cross-feed consensus BUY, but neither is in tonight's actual TRENDING/mention list (rule #18 — can't borrow WSB hype that isn't there), and it's the last cycle of the day with RISK-OFF regime — a fresh overnight entry needs a real dated catalyst, not just a technical cross-feed nod. NVDA and MRVL show up in mentions but consensus is SELL/thin-BUY and NVDA specifically nets SELL on cross-feed with only 1 BUY vote (analyst) — fails the ≥2 BUY-vote corroboration requirement anyway. Sitting mostly in cash into the close is the correct, disciplined call given the regime tilt and the empty WSB catalyst calendar for tonight — better to preserve capital for tomorrow's 09:45 read after the overnight settles the Fed-hike reaction and CLARITY fallout properly plays out in price. GOOGL stays untouched: small +1.1% position, already at the position cap (no room to add anyway), no fresh WSB mention tonight, not down >10% so rule #4 doesn't trigger a cut. No other name clears the bar for a fresh BUY on the last cycle: RISK-OFF regime (QQQ -0.42%, SPY -0.85%) raises the bar, WSB's actual hot listing tonight is dominated by Fed-hike/CLARITY-Act macro memes and non-tradeable tickers (DELL, KDK, AUR, DXYZ), and the only tradeable trending names (NVDA, MRVL, RKLB, ORCL) either fail the ≥2 cross-feed BUY-vote corroboration (NVDA is net SELL, 1 BUY vote), lack a dated catalyst to justify holding through the overnight gap (rule #19), or show negative INTERPRETED sentiment (ORCL -1.00, MRVL -1.00 on layoffs/pre-Fed call-selling). Preserving cash into the close given the fee-drag history (0.8% round-trip) and the -0.12%/day overnight-negative-pay stat on this universe is the disciplined move; will re-read at 09:45 tomorrow once the Fed reaction and CLARITY fallout are fully priced.

Horisont: 3-7 days Begrundelse: COIN is the top fresh WSB mention today (+2.0σ, CLARITY Act vote thread) with cross-feed consensus BUY (short_interest + analyst backing), despite loss-porn on the failed vote — sizing small given the bearish INTERPRETED read and Fed-week volatility. === Session-analyse === [wsb/US] Thin, mixed WSB tape this morning: the Daily Discussion dominates hot but the substantive posts are mostly loss-porn (RKLB fumble, 1.1M-to-150k, NVDA puts blown up, COIN calls burned on the failed CLARITY vote) — this is a fear-tinted sub right now, confirmed by the megathread pulse reading 20% bull / fear. COIN is the one name with real structure: 2 WSB mentions (+2.0σ vs its own baseline), cross-feed consensus BUY (3 BUY/1 SELL/4 HOLD) with short_interest backing 12.4% float short, and it's in the CLARITY Act news cycle (failed vote = fading regulatory clarity, but WSB posts are split on fading vs ripping into the vote, and analysts/short_interest still lean bullish). The INTERPRETED read is net-negative (-0.25, 100% loss-porn) on COIN specifically because of the failed vote, so I'll size small and lean on cross-feed corroboration rather than the Reddit hype itself. RDDT also shows cross-feed consensus BUY (analyst target +33% upside) but has zero fresh WSB mention today and my own TICKER-FACIT shows -458 DKK across 2 prior exits on this exact name — theme fatigue, no fresh catalyst, skip. ORCL is the #4 hot post but the DD/INTERPRETED read is unambiguously bearish (CFO comments post-layoffs, -1.00 net sentiment, fundamental thesis) — that's a name WSB is discussing negatively, not a buy signal. NVDA is a hot YOLO post but it's SPIKE-CHRONIC (always trending, no +0.1 bump), cross-feed consensus is SELL (3 SELL/1 BUY), and the local read is literally an author who blew up his account on NVDA puts — pure loss-porn, no edge, and I already ate two stop-outs on NVDA/NBIS this week from over-exposure to this exact ai_infra/semis complex. Staying out. Given RISK-ON regime (QQQ +0.47%) I'd lean into strength, but nothing here clears the ≥2 cross-feed-vote bar with genuine fresh WSB conviction except COIN, and even that has a bearish INTERPRETED qualifier. I'll take one small, corroborated position in COIN sized well under cap, with a wide-ish stop given crypto-adjacent volatility and event risk (Fed decision Wednesday, CLARITY Act fallout), and otherwise sit tight on GOOGL (holding fine, +2.2%, no reason to touch) and hold the rest in cash — 81% cash is the correct posture when the WSB vibe is fear/loss-porn dominant rather than screaming a fresh rocket. GOOGL is a small +2.2% winner with no cross-feed weakness and CAP-FULL on adds — just holding, no action needed. Most of today's hot WSB content is loss-porn (RKLB fumble, 1.1M→150k, blown NVDA puts, burned COIN calls) rather than fresh rocket conviction, and the megathread pulse reads only 20% bullish with a fear tone — this is not a day to chase size. NVDA, ORCL, RKLB and the crypto-miner complex (a theme I've already lost -1650 DKK on over 18 round-trips) all lack a genuinely fresh, corroborated catalyst today, so I'm passing on them and keeping ~79% of the book in cash as dry powder for a cleaner setup later in the session.

Horisont: closed Begrundelse: Crypto_mining theme absent from today's WSB trending/mention lists and overnight gap -4.1%; theme has been a -1590 DKK loser over 90d with no cross-feed BUY consensus to save it — full exit ahead of an unprotected overnight hold. === Session-analyse === [wsb/US] Last cycle of the day, QQQ -0.67% and SPY -0.48% — a risk-off tape — with the megathread PULS reading 'fear' and 80% bearish on top comments. This is not the session to open new overnight risk on thin conviction. COIN is the one trending name with fresh mention acceleration (+5 mentions today, CLARITY Act Senate vote catalyst) but the overnight gap was -8.9%, the INTERPRETED read is net-sentiment 0.00 (split bull/bear, zero high-conviction posts), and COIN has no listed cross-feed BUY consensus this cycle — a coin-flip binary bet into a failed Senate cloture vote is exactly what rule #19 says to skip late in the day. NVDA is a SPIKE-CHRONIC always-trending name (28k YOLO post, highest mention count) with a mixed/bearish comment-flow gist per PULS — crowding flag, not a fresh buy signal. On holdings: IREN gapped -4.1% overnight into today and sits at -3.4% unrealized; crypto_mining is a theme that's bled -1590 DKK over 90 days across 17 round-trips (my own theme-facit) and it isn't mentioned anywhere in today's TRENDING or MENTION-LEADERS lists — WSB has moved on from the miners entirely today. No cross-feed BUY-consensus line is shown for IREN to invoke the rule #4 override, so per rule #4 (forgotten + down, no consensus save) I'm cutting it now at a known price rather than holding an unprotected soft-stop through a night that's been measured as negatively-paid (-0.12%/day) with 44% of my own stop-fires historically landing in the first 30 minutes after open. GOOGL (+2%) is small and CAP-FULL per the strategy params — no action, just holding. No new BUYs this cycle: nothing clears the ≥2 cross-feed BUY-vote corroboration bar with genuine non-spike-chronic WSB momentum. Crypto Fear & Greed at 69 (Greed) is a blow-off-top caution on crypto-adjacent names per rule #12, the regime is risk-off per rule #11, and per rule #19 any late BUY needs a catalyst that survives an unprotected overnight gap — nothing on today's board meets that bar. Raising cash into the close and re-assessing at tomorrow's 09:45 open is the disciplined call. Sitting mostly in cash into the close: risk-off tape (QQQ -0.67%), fearful megathread sentiment, and no tradeable name clears the ≥2 cross-feed BUY-vote bar with fresh (non-spike-chronic) WSB momentum. COIN's Senate-vote catalyst is a binary event with zero-conviction split sentiment and no cross-feed backing — skipping per rule #19's late-cycle discipline. GOOGL remains held at cap, small and roughly flat, no action needed.

Horisont: 1-2 weeks Begrundelse: Fresh WSB DD thread on IREN's $50M/yr NA sports sponsorship deal (genuine new catalyst, not just crypto-mining momentum) plus cross-feed consensus BUY (analyst target +80% upside) clears the re-entry bar after yesterday's stop-out was actually a small winner (+4.1%). === Session-analyse === [wsb/US] WSB today is mostly macro doom (10Y hitting 5%, Warsh/Fed rate-hike fears, 'this selloff makes zero sense', Kenny G / Volcker jokes) rather than a clean meme-stock rocket thread. NVDA gets genuine multi-post bullish interpreted signal (net +0.62, 3 bull/1 bear, catalyst-heavy, disclosed positions) plus cross-feed corroboration would be needed but NVDA isn't showing consensus BUY in the screening block, so I'm passing on it — single-source WSB hype without 2+ cross-feed votes is exactly the loss-making pattern rule #9/CROSS-FEED gate warns against. MSFT does clear the bar (consensus BUY, wsb+2 hold votes, interpreted net +1.25 with disclosed 0DTE call profits being let ride) but it's on the UNTRADEABLE-at-min-size list for this account, so it's a no-go regardless of conviction. Cross-feed BUY consensus exists on IONQ, TSM, ASTS, ONDS, SOUN, RDDT, IREN, APLD, QBTS among others, but none of them are in today's TRENDING or MENTION-LEADERS lists — meaning WSB itself isn't currently talking about them, just background analyst/short-interest votes. Buying purely on cross-feed consensus with zero WSB chatter contradicts this book's entire mandate (WSB sentiment is the primary source). IREN is the one small exception — it's mentioned in trending (score 8, low volume) AND has a fresh DD-style post about a $50M/year sponsorship deal AND cross-feed consensus BUY — but I just ate a stop-out on IREN yesterday (-15.8%... wait, that was NBIS; IREN was +4.1%), and per rule #8 re-entry requires conviction ≥0.7 AND ≥2 fresh cross-feed BUY votes citing what's changed. The IREN sponsorship post is genuinely new information (a real business catalyst, not just momentum), cross-feed shows 2 BUY votes, and crypto_mining theme, while a net loser for me historically (-1590 DKK/17 exits), IREN individually has been a net winner (+242 DKK/2 exits) — so I'll take a small, disciplined re-entry with a tight stop given the theme's poor track record and existing GOOGL exposure already at cap. COIN is accelerating in mentions (+2 vs yesterday, DD post 'Fading Clarity Act') but gapped down -6% overnight and cross-feed doesn't show in the BUY list — that's a risk flag, not a buy signal; the crypto_exchange theme tape is actually +11%/5d which is a divergence worth noting but I don't have corroboration to act on a falling knife. Given QQQ/SPY both slightly red (neutral-to-soft regime), VIX calm at 15.84 but Fear&Greed at 69 (Greed) which is more a headwind for chasing crypto-adjacent names, I'm keeping this small: one modest IREN add, hold GOOGL (already at cap, small +2.9% winner, no reason to touch), and otherwise sit mostly in cash. No forced activity — WSB's hot list today is dominated by macro fear threads, not a fresh meme-stock catalyst worth chasing. GOOGL is a small +2.9% winner at the theme cap (internet/ai_infra/mega_cap all at 19%) — no room to add and no reason to sell a name WSB isn't actively hating on and where overnight gap was only -0.6%. Beyond the single IREN re-entry, I'm avoiding the crowd of cross-feed-only BUY names (IONQ, TSM, ASTS, ONDS, SOUN, QBTS, RDDT) because none of them show up in today's actual WSB trending/mention data — that's cross-feed noise without the WSB conviction this book is built to follow, and chasing it would just be borrowing tomorrow's fee drag for no edge (rule #17). NVDA has genuine WSB bull sentiment but lacks cross-feed corroboration (no BUY consensus shown) so the single-source gate blocks it. Macro backdrop (10Y at 5%, hawkish Fed fear, Warsh rate-hike thread top of hot list) argues for caution — QQQ/SPY both red today, Fear&Greed at 69 (Greed) is a chase-risk flag not a buy signal. Sitting mostly in cash (staying ~92% cash after the IREN add) is the correct read of a session where WSB is arguing about macro doom, not rallying behind a specific rocket.

SOFT STOP TRIGGERED: NVDA faldt under 7.5% stop (225.41 → 209.27). Auto-cut før agent-call.

SOFT STOP TRIGGERED: NBIS faldt under 10% stop (248.59 → 209.53). Auto-cut før agent-call.

SOFT STOP TRIGGERED: IREN faldt under 10% stop (40.26 → 41.99). Auto-cut før agent-call.

SOFT STOP TRIGGERED: MP faldt under 7% stop (54.45 → 50.62). Auto-cut før agent-call.

Horisont: 1-2 weeks Begrundelse: Highest-conviction interpreted WSB signal today (+1.50 net-sentiment, katalysator 1/1) on the Nebius-Nvidia-Palantir partnership news, backed by 3 BUY cross-feed votes — but sizing small since it already gapped +9.8% overnight (chase risk) and ai_infra theme is at cap. === Session-analyse === [wsb/US] WSB's top hot post is genuinely about NVDA/Jensen Huang 'AGI has arrived' (6,404 score) with a fundamental thesis and disclosed position, and NBIS shows up in the interpreted signal with the highest conviction read (+1.50, katalysator 1/1) tied to a Palantir-Nebius partnership — that's a genuine fresh catalyst, not recycled hype. NBIS is also gapping +9.8% overnight, up +14% 5-day per cross-feed yfinance, and carries a clean 3 BUY/0 SELL/2 HOLD consensus. That clears the corroboration bar and the ai_infra theme, while already at cap (57%), is a theme I'm not adding pure exposure to blindly — but NBIS's own idiosyncratic catalyst (Nebius-Nvidia-Palantir tie-up) plus its distinct sub-theme framing justifies a small deliberate nudge given the strength of signal, sized modest to respect the cap discipline. INTC is the mention-leader today (6 mentions, accelerating +5) with loss-porn gain-brag posts ('9k->25k', 'bought Friday, on the rise') — but the interpreted signal is 100% loss-porn with 0 high-conviction, which per rule #15 is a caution flag, not confirmation. INTC isn't in cross-feed screening at all here (no consensus line shown), so it fails the ≥2 BUY-vote corroboration requirement outright — skip it despite the mention buzz. Existing book: IREN is up +20.2% and gapped +8.3% overnight with a strong BUY consensus (analyst target +74% upside) — this is a genuine winner to protect, not trim; raising its trailing stop to lock in gains per the ladder (+15%+ tier → ~12% under price, but IREN's daily moves run ~5-8%, so I'll set 10% to respect the gap-floor). MP is up modestly with BUY consensus and rare_earth tema-tape positive (+2.8%/+8.2%), holding as-is. NVDA and GOOGL are CAP-FULL (ai_infra at 57%, AT CAP) — no adds, and both are only marginally green so nothing to ratchet. QQQ neutral regime, VIX calm at 14.32, Fear&Greed at 69 (Greed, not extreme) — no reason to force size, but NBIS's clean idiosyncratic setup earns a measured entry outside the maxed ai_infra bucket via its own AI-infra-adjacent but fundamentally distinct positioning (partnership catalyst, not pure momentum chase after already gapping hard). IREN's winner is running (+20.2%, big overnight gap) — protecting with a tightened trailing stop rather than trimming, per rule #2. MP holds steady with BUY consensus and positive rare_earth tape. NVDA and GOOGL are CAP-FULL in the already-saturated ai_infra theme (57%, AT CAP) so no adds regardless of the Jensen Huang AGI headline driving today's top WSB post — that enthusiasm is already reflected in existing NVDA exposure. INTC's mention-leader status is pure loss-porn brag content (100% loss-porn, 0 high-conviction per interpreted signal) with no cross-feed screening line at all, failing the ≥2-BUY corroboration gate — skipping despite the buzz. Staying otherwise in cash to keep dry powder given QQQ neutral/SPY red regime and no other name clearing both WSB-freshness and cross-feed bars.

Horisont: 2-3 weeks Begrundelse: WSB DD thread flags China rare-earth export restrictions escalating toward Sept 24 talks as a squeeze catalyst for MP, backed by cross-feed BUY consensus and analyst +40% upside target — exit if talks de-escalate or price breaks below 52. === Session-analyse === [wsb/US] MP is the standout: WSB DD post on China rare-earth export restrictions escalating into Sept 24 talks, interpreted signal net +1.25 with a fundamental thesis and disclosed skin-in-the-game (author holds shares), cross-feed consensus BUY (analyst target +40% upside), and it's an accelerating mention with a fresh geopolitical catalyst — not stale chatter. Rare_earth/critical_minerals theme is only at 20%, well under cap, and the position was rejected last cycle only due to headroom (~1 share). IREN also gets a nod: it's up +9.6%, gapped +6% overnight, cross-feed consensus BUY (yfinance top signal, strong 5d momentum), and crypto_mining tape is actually +11.7% over 5d despite my prior losses in the theme — this is a genuine breakout not a repeat bagholder story, so I'll protect gains with a tighter trailing stop rather than add (crypto_mining theme already carries scar tissue per THEME-FACIT, so no fresh buy there). GOOGL and NVDA are CAP-FULL, no adds possible even if I wanted them; ai_infra theme is already ⚠ AT CAP so no fresh names in that bucket regardless of WSB noise. HOOD shows an interesting post ('2k→52k HOOD') and cross-feed BUY consensus, but it's in RECENT EXITS territory? No — HOOD isn't in my recent exits, but it gapped down -5.7% territory recently (TSLA gapped, not HOOD) — actually HOOD is fine, not previously exited, but I have zero room in USD ($461 spendable) after protecting existing winners, and HOOD's thesis is 'other' (options YOLO bragging), not a clean fundamental setup — thin conviction, skip. QQQ regime is neutral, VIX calm at 15.2, Fear&Greed at 74 (Greed) — no reason to chase extra risk, just execute the one clear MP add and tighten IREN's stop. GOOGL and NVDA are CAP-FULL under the 25% position cap — no adds possible regardless of conviction. ai_infra theme is already at its 50% concentration cap, so I'm avoiding any fresh buys there even though WSB commentary touches NVDA. Sitting mostly in cash beyond the one MP add given the neutral regime and thin remaining USD firepower (~$461); will revisit HOOD/ORCL if cross-feed strengthens or a real dip offers entry.

Horisont: 1-3 weeks (into Sept 24 rare-earth talks catalyst) Begrundelse: WSB DD thread (accelerating mentions, NEW tag) ties MP to a concrete Sept-24 rare-earth export-restriction catalyst, backed by cross-feed consensus BUY and analyst +40% target — exit if the talks defuse or MP breaks below ~54 support. === Session-analyse === [wsb/US] IREN is the standout winner (+8% since entry, overnight gap +4.4%, cross-feed consensus BUY with yfinance +1.00 citing 5d +17.7% momentum) but it's already sized and ai_infra/crypto_mining themes are stretched — protecting the gain with a tighter trailing stop is the right move rather than adding, especially since GOOGL and NVDA are both CAP-FULL and ai_infra sits at 57% (AT CAP). WSB itself is quiet on genuinely fresh names today: HOOD has a gain-porn post (2k->52k) and cross-feed BUY, but it's been bleeding overnight (-6% multi-day) with only 1 high-conviction post backing it — chasing an options YOLO thread into weakness is the exact loss-making pattern rule #18 warns about, so I'm passing. TTWO and MSTR threads are pure loss-porn (100% loss-porn tag) — textbook 'don't buy the regret post.' The one name that clears the bar is MP: cross-feed shows consensus BUY (3 BUY/0 SELL/1 HOLD, analyst target +40% upside), WSB mention count is accelerating (NEW/RISING tag, not chronic-spike) with an actual DD thread citing a concrete Sept 24 catalyst (rare-earth export restriction talks), and the INTERPRETED signal shows net-sentiment +1.25 with disclosed author position and a fundamental (not meme) thesis. It diversifies away from my saturated ai_infra bucket into rare_earth, which has a strong 30d tape (+17.4%) despite today's slight pullback — a name moving with its theme, which per the theme-tape rule is the cheaper, higher-probability momentum read versus a name diverging from its cohort. Everything else — GME/ORCL earnings incoming, ADBE gapping down hard, the broader comment-flow reading bearish/fear-dominated — argues for restraint. QQQ is flat-to-slightly-red today (neutral regime), VIX is calm at 15.2, so no reason to force size, but also no reason for aggressive buying. Sitting mostly in cash (43%+) with one small, corroborated MP starter position and a tightened IREN stop is the disciplined read here. GOOGL and NVDA are CAP-FULL (no room under the position cap) and ai_infra is already AT the 50% sub-theme cap, so no adds there regardless of sentiment. WSB's hot list today is dominated by non-tradeable names (LULU, DELL, SNOW, BROS) and loss-porn threads (MSTR, TTWO) that argue against entry, while HOOD's gain-porn thread is riding a name that's been gapping down for days — chasing it would be exactly the single-source WSB-hype trap rule #16/#18 warns about. Comment-flow sentiment is fear-dominant (5% bull) and QQQ is flat, so this is a day to protect existing winners (IREN) rather than add risk. Cash stays high (~43%) as dry powder for a cleaner setup next session.

Horisont: 1-2 weeks Begrundelse: Top mention-leader today (4 mentions, accelerating) with cross-feed consensus BUY and overnight crypto-mining sector-wide pop (RIOT/MARA/CLSK all +9-11%) confirming the sympathy-play thesis WSB is discussing. === Session-analyse === [wsb/US] IREN is the clearest setup on the board today: it's the top mention-leader (4 mentions, accelerating +4 vs yesterday), it's already a NEW/RISING trending name earning the +0.1 bump, cross-feed shows consensus BUY (analyst target +97% upside), and it gapped up +3.8% overnight on genuine crypto-mining strength (RIOT +11.5%, MARA +10%, CLSK +9.8% overnight confirms the whole sub-theme is running, not just IREN). WSB's own IREN thread ('7 day 8K profit', 'sympathy play') is loss-porn-tinged but net-sentiment is positive and it's riding real Dell-earnings-adjacent AI infra momentum. RISK-ON regime (QQQ +1.22%) supports leaning into this rather than fading it. I already hold 4 shares at 39.62 (+3.8%) — I'll add modestly rather than build a fresh full-size position, keeping crypto_mining theme exposure controlled since ai_infra is already at 50% of book. AVGO and TSLA are noise I'm passing on: AVGO is a SPIKE-CHRONIC-flavored earnings-reaction story (weak guidance, stock down 5%, WSB literally split 2 bull/2 bear with 50% loss-porn) — that's a coinflip on a name with an untradeable-at-min-size flag anyway. TSLA cross-feed is a flat SELL/HOLD consensus (0 BUY, 1 SELL, 4 HOLD) despite the calls-YOLO post — no corroboration, skip. GOOGL and NVDA are CAP-FULL per the params, so no adds there despite decent overnight moves; I'll just let them ride with tightened trailing stops given NVDA/GOOGL are both modestly green and ai_infra/semis theme is already maxed at 50%/24%. No new positions beyond the IREN add — cash stays mostly on the sidelines (crypto_mining theme would go from 8% to a still-modest ~13% with this add, well under the 50% cap). PLTR is off the table per re-entry cooldown (stop fired 2026-09-02, no fresh catalyst, cross-feed not shown as strong BUY). Sitting on remaining cash for now; will watch AVGO earnings fallout and TSLA momentum for next session. GOOGL and NVDA are CAP-FULL per strategy params — no adds allowed regardless of overnight strength, so tightening their trailing stops modestly to lock in the small gains instead. AVGO's WSB thread is a genuine 50/50 split with heavy loss-porn post-earnings-guidance-miss — a coinflip I'm not taking, especially since it's untradeable at min size anyway. TSLA calls-YOLO post has zero cross-feed corroboration (consensus SELL). Remaining cash held as dry powder given ai_infra theme already at 50% of book and only IREN cleared the bar cleanly today.

Horisont: 1-2 weeks Begrundelse: NEW/RISING WSB mention citing IREN as an AI-infra sympathy play off Dell's server-strength earnings, backed by cross-feed consensus BUY (analyst target +97% upside) and my own prior +168 DKK win on this name. === Session-analyse === [wsb/US] AVGO dominates the WSB hot list (4 mentions, top post 1186 upvotes on the guidance-miss drop, plus a 700K YOLO thread and buy-before-ER thread) but the actual sentiment is split 3 bull/2 bear with 40% loss-porn and the ticker just gapped -5.7% overnight on weak guidance — that's a fresh negative catalyst, not a dip to buy, and it would add straight into an already-maxed ai_infra/semis book (41%/24%, both effectively full given GOOGL+NVDA are CAP-FULL). Skipping it. GOOGL is genuinely trending (#10 'Just Keep Buying GOOG', 724 score) with a clean +1 sentiment post and gapped +1.9% overnight in a risk-on tape, but I already hold it and it's flagged CAP-FULL — no room to add, so I just let it ride. IREN is a small but real signal: NEW/RISING mention tag, accelerating, a sympathy-play DD post citing the Dell AI-server beat as a concrete catalyst, and cross-feed consensus BUY (analyst_ratings +0.82, target +97% upside) with 2 BUY/0 SELL votes clearing the corroboration bar. Crypto_mining is a theme I've bled on (-1584 DKK over 18 exits) but IREN itself was a personal winner (+168 DKK) and the theme-tape shows crypto_mining +0.3% 5d while underlying BTC is +22.3% 30d — divergence resolving upward, and this is a NEW name-specific catalyst (Dell/AI-server sympathy, not just crypto beta), satisfying the theme-external-catalyst bar. Sizing modest given the theme's poor track record and RISK-ON but not screaming tape. No fresh capital goes to AVGO, DELL, SPOT, WEAT, DXYZ (none in universe or too risky), and I'm respecting the CAP-FULL flags on GOOGL/NVDA. Cash stays high (59%) — most of today's hot list is a single overnight AVGO story I don't want to chase into weak guidance, and nothing else clears both the WSB-trending bar and the 2-feed corroboration requirement cleanly enough to deploy meaningfully. GOOGL and NVDA are both CAP-FULL per server params — no adds allowed regardless of conviction, and both are modest winners (+1.8%/+1.1%) with the theme still WSB-relevant, so I'm holding both unchanged rather than forcing a stop-tighten on gains this small. AVGO is the loudest WSB story today but the sentiment split (3 bull/2 bear, 40% loss-porn) plus the overnight -5.7% guidance-driven gap-down argues against chasing a falling knife into an already-saturated semis/ai_infra book. Staying mostly in cash (59%) going into a risk-on but narrow-breadth session.

Horisont: 2-3 weeks Begrundelse: WSB DD post shows disclosed 2-share accumulation thesis (net-sentiment +1.25, high-conviction, fundamental) plus cross-feed consensus BUY (wsb+analyst-adjacent) — exit if thesis reverses or a cross-feed turn to SELL. === Session-analyse === [wsb/US] NVDA is already at cap (50% of portfolio, semiconductors + ai_infra both AT CAP), so no more adding there despite the WSB Hugging Face acquisition buzz and +4% overnight gap — thesis is intact, let it run rather than trim. GOOGL is genuinely interesting: interpreted WSB signal shows a real fundamental thesis (net +1.25, high-conviction, disclosed 2-share position, zero loss-porn) plus cross-feed consensus BUY (wsb + another feed), but GOOGL isn't in the tradeable-at-min-size list issue — checking universe, GOOGL is tradeable but at ~$340/share a 1-2 share entry is fine per rule #3 exception. However cross-feed shows only wsb+1 other as BUY, and it's not in my top-3 trending mentions (NVDA and SPY are). GOOGL's own theme (internet) is down -10.4% over 30d — a name rising against a falling theme needs its own catalyst, which the acquisition-style DD provides, but conviction is borderline given thin 2-post sample. Given NVDA is capped, PLTR just stopped out (wick-out per the post-mortem, price 169.10 vs stop 170.38 — real break, not a recovery yet), and most other WSB chatter today is memes/politics/noise (Venezuela oil, Thai warning, Jensen Huang jokes) rather than actionable tickers, I don't see a name that clears BOTH the WSB signal bar AND the 2-independent-cross-feed corroboration requirement while fitting outside my saturated semis/ai_infra theme. GOOGL is the one candidate with genuine cross-feed BUY consensus and a real interpreted-signal read, so I'll take a small position sized well under cap, using cash freed by nothing (still have room since NVDA is only ~50% and I have ~$1,463 spendable). RDDT also shows cross-feed BUY consensus (analyst_ratings +0.72, short_interest presumably) and is gapping +6.7% overnight, but my own TICKER-FACIT shows -458 DKK across 2 exits on RDDT — repeated loser, no new catalyst here beyond the overnight pop, so I'll pass per re-entry discipline (needs external catalyst, not just gap momentum). NVDA is capped at 25% max position and both semiconductors/ai_infra themes are AT CAP — no room to add despite the Hugging Face acquisition buzz and +4% overnight gap; letting the existing 2 shares ride with the current stop is correct, no fresh entry needed since PnL is only +0.3% (too early to ratchet trail). Most of today's WSB hot list is meme/political noise (Venezuela oil, Thai sailors, Jensen Huang jokes) with no fresh actionable ticker beyond NVDA (already maxed) and SPY (too expensive at min-entry, and a single 0DTE options gain post isn't a portfolio thesis). RDDT is gapping +6.7% and has cross-feed BUY consensus, but my own 30d ticker-facit shows -458 DKK across 2 losing round-trips on RDDT with no new name-specific catalyst beyond the pop — passing per re-entry discipline. Staying mostly in cash with one small qualified GOOGL add; will revisit NVDA trail once it extends further into gain territory.

SOFT STOP TRIGGERED: PLTR faldt under 8% stop (163.50 → 170.38). Auto-cut før agent-call.

SOFT STOP TRIGGERED: MSTR faldt under 6% stop (127.60 → 129.32). Auto-cut før agent-call.

Horisont: 1-2 weeks Begrundelse: Top mention-leader today (13 mentions, RISING z=+3.1σ) with a +7.5% overnight earnings-beat gap and cross-feed consensus BUY, but interpreted sentiment is only mildly positive (+0.24) with 42% loss-porn and fear-tilted comment flow, so sizing small. === Session-analyse === [wsb/US] Both existing positions are at their position cap (MSTR, PLTR) so no adds there — but both are winners riding good tape: MSTR is +8% and just gapped +11.9% overnight on Saylor/BTC treasury enthusiasm, and cross-feed shows consensus BUY (analyst target +86% upside). PLTR is +12.8%, gapped +3.9% overnight, RDDT-style AI infra momentum continues, and RISK-ON regime (QQQ +1.09%) supports letting both run rather than trimming. Per rule #2/#13 I'm raising trailing stops on both to lock in gains while giving room to extend — MSTR's daily moves have been elevated recently (~5-8% swings) so I won't trail tighter than that noise band. On new entries: NVDA is the only genuinely fresh signal — trending #1 in mentions (13 today, RISING z=+3.1σ), overnight gap +7.5% post-earnings-beat, cross-feed consensus BUY (2 BUY/0 SELL/4 HOLD, analyst target +45% upside). But the INTERPRETED read is a real caution flag: net-sentiment only +0.24 (barely positive), loss-porn 42%, split 9 bull/7 bear, thesis tagged loss_porn — 'stock fell despite beating earnings' is the dominant narrative, and KOMMENTAR-FLOW sentiment is explicitly 'fear' with NVDA discussed as 'mixed'. This is a name where the stock already fell on the print and is now bouncing — not a clean momentum setup. I'll take a modest position sized down given the mixed signal, with cross-feed corroboration (analyst_ratings BUY) satisfying the 2-vote requirement alongside wsb itself, and a wider stop given NVDA's typical multi-percent daily swings. Everything else in the trending/mention-leader list (WEN -12.5% gap down on food-quality PR disaster, RDDT loss-porn ironic calls, MRNA loss-porn puts closed) is either a forgotten bagholder story or explicit loss-porn/irony — not tradeable theses. WEN's DD is a CEO-quality-control news story causing a real gap down, no bullish angle. RDDT and MRNA interpreted reads are both loss-porn/ironic, sample size of 1 post each — essentially zero signal per rule #15. Staying disciplined: one new small NVDA position, ratchet stops on the two winners, nothing else clears the bar today given crypto_treasury is already at 43% (near cap) and ai_infra/defense at 29% combined with PLTR capped out anyway new ai_infra names would push that theme further without room in PLTR itself. Cash stays largely available for genuinely fresh setups next session — MRVL/IREN earnings today likely to shake more names loose. MSTR and PLTR are both at position cap (25%) so no room to add regardless of conviction — riding both with tightened trailing stops instead per rule #2/#13. New entries are limited: WEN is gapping down hard on bad PR (food-quality scandal), not a buy despite trending; RDDT and MRNA interpreted reads are thin-sample loss-porn/irony with no real conviction; crypto_treasury theme is already near cap via MSTR so avoiding stacking MARA/RIOT/CLSK etc. despite decent cross-feed votes there, especially given theme-facit shows -1143 DKK lifetime on crypto_mining round-trips with no fresh external catalyst beyond the existing MSTR position. Sitting mostly in cash aside from a small qualified NVDA add.