Signal: Wrong at the horizon traded — ASML closed next day at 1643.42, below the 1656.46 entry, though it did reach +1.72% at some point while open, so the directional thesis wasn't dead on arrival. Timing: The fill is the problem: decision-time price was ~1642 (prev close 1650 minus 0.47%), yet the fill printed at 1656.46, roughly 0.9% higher — the position paid up before it even existed, and that gap alone exceeds the final loss (at the decision price the trade would have been roughly flat). The journal doesn't say whether that was stale pricing or a late fill, so the cause isn't settled here. Exit: The mechanical next-close exit landed near the worst unrealized (-0.93%) and gave back the entire +1.72% peak — but the dominant factor is the entry fill ~0.9% above the quoted price, which converted a marginally positive drift into a -13.04 USD loss.

Signal: the long call was wrong at the chosen horizon — TSLA closed next day at 355.10, down 2.27% from the 363.36 entry, so the overnight-drift / merger-speculation thesis did not deliver a positive close-to-close move. Timing: the stock was already down 1.79% on the day at decision time, so the entry was not chasing a move that had already happened; nothing here suggests the rationale's catalyst had been priced in, but there is also no evidence the catalyst mattered at all. Exit: the position reached +5.16% unrealized while open (about +$37 on 2 shares) and the mechanical next-close exit gave back all of it to finish at -2.27% (-$16.52), near the -3.80% worst — the dominant factor is the fixed next-close exit discarding a large intraday gain on a signal that was ultimately wrong for the holding period.

Signal was wrong: the rationale bet on positive overnight drift from a favorable regulatory headline (DOJ probe halted), but META fell -4.13% from entry to next close (737.45 → 706.98), so direction was the dominant factor. Timing is unsettled by this data — the pre-entry move was tiny (+0.33% on the day, +0.27% last 30m), so the "move already happened" question doesn't apply; the loss came from something after entry that the journal doesn't show. The exit was mechanical (next close) and captured nearly the worst of the range: the trade saw at most +0.81% unrealized gain versus -4.68% at the low, and the -4.13% realized is close to the trough, so exit timing added little but also wasn't the cause — the -30.47 USD loss is a signal failure.

Signal: Right — MSFT fell 2.31% from entry (541.68 → 529.16) and the worst unrealized drawdown was only -0.13%, so the short was almost never underwater. Whether the CMA/cloud-choice news actually drove the drop isn't settled by this data; only the direction is.

Timing: Not stale — at decision time MSFT was +0.39% on the day and flat over the prior 30 minutes, so the bearish move had not yet occurred; entry was near the high of the holding period (worst unrealized -0.13%).

Exit: Captured ~73% of the best available (+2.31% realized vs +3.17% peak), leaving roughly $4.60 on the table — a reasonable but not optimal exit. Dominant factor: a correct directional call with clean timing; the exit modestly underperformed the peak.

Signal was wrong: the bullish call on TSM's AI-capacity leverage never produced upside — the best unrealized was -0.03%, so price was never above entry while open, and it closed at 413.82 for -0.91%. Timing is the dominant factor in a specific sense: the stock was already soft (prev close 422.23, -0.38% at decision, then filled at 417.62, roughly 1.1% below prev close), meaning the "FRESH" verdict was being faded by the tape before entry even completed, though this data can't settle whether the article was actually stale news. Exit was reasonable given what was available — -0.91% landed between the -0.03% best and -1.66% worst, so the exit neither captured gains (none existed) nor rode the full drawdown.

Signal: right. AAPL fell after the short was placed, so the bearish call paid off, though the article (TSMC capacity constraints across NVIDIA/Apple/AMD) is a thin, indirect link to Apple specifically and nothing here shows the move was driven by that story. Timing: the predicted move had not already happened — at decision time AAPL was only +0.36% vs. prior close (~343.2), and the fill at 345.67 (+1.08% vs. 341.98) was actually ~0.7% higher than the decision-time price, a favorable fill for a short; the data doesn't explain that gap between decision and fill. Exit is the dominant factor: the realized +0.72% (+5.00 on 2 shares) captured only about half of the +1.49% best unrealized, against a minor -0.49% worst, so the trade was correct but left roughly half the available gain on the table.

Signal: Right — NVDA fell after the short, from 216.25 to 214.60 (-0.76%), consistent with the bearish read, though whether the "OpenAI valuation / CoreWeave sell-off" headline actually drove the move versus general tape can't be determined from this data. Timing: The move had not already happened — at decision time NVDA was only -0.43% from prior close and flat over the last 30m, and the entry at 216.25 was essentially at that level, so the entire captured decline occurred after entry; the -1.03% worst unrealized shows the trade also ran against by about a percent at some point, but the ordering of best vs. worst isn't given. Exit: The exit captured roughly 60% of the +1.26% best unrealized (-0.76% realized vs. -1.26% available), a reasonable but not full harvest; the dominant factor here is simply a correct directional call on a small 4-share position, with the exit leaving about 0.5% on the table.

Signal was wrong: the rationale called for a bullish move on a positive-sounding demand article, but ASML fell about 1.74% from entry (1671.89 → 1642.79), never rising more than +0.49% before sliding toward the -1.85% low. Timing is partly implicated — the stock was already +0.80% above the prior close at decision time with the last 30m flat, suggesting any article-driven pop had largely occurred before entry — though the data can't confirm whether the article itself caused that earlier move. Exit was near the worst point (-1.74% realized vs -1.85% worst), but with only +0.49% ever available on the upside, there was little to capture; the dominant factor is the wrong directional call, not exit mechanics.

Signal: Right — the bearish call played out, with TSLA falling from the 364.27 fill to 355.39 and the position never more than -0.26% against it, so the direction was correct and the entry was essentially at the local high.

Timing: Not stale — the fill was above the prior close (362.47) despite the -0.86% last-30m dip seen at decision time, so the move had not already happened; the short got in before the decline rather than chasing it.

Exit: Partial capture — the +17.76 USD exit equals about +2.4% on 2 shares, versus +4.05% best unrealized, so roughly 40% of the available move was given back before closing. Dominant factor: a correct signal with a good entry, undercut only by an exit that left a meaningful slice of the move on the table; the journal doesn't say what triggered the exit, so whether that giveback was rule-driven or reversal-driven can't be determined from this data.