| CS-2026-0025 |
US 500 |
-1.00 |
Process was sound |
No process error |
The decision followed the method and the stop was where the thesis died, so nothing to fix in the levels. The weak point is the edge, not the execution: a 0.64 ATR deviation is a thin trend reading, and a short with only 1.5 sessions of life ending the day before an FOMC decision rides against the usual pre-decision drift. Candidate pattern to watch, not yet a rule (one occurrence): when the only exit window ends the session before a scheduled central-bank decision, require a band deviation of at least 0.75 ATR or skip. Also noted: US 500 is now 1 win in 5 closed trades; too small to act on, but the next US 500 idea should have a cleaner structure than a marginal-band retest. |
| CS-2026-0024 |
US 500 |
2.00 |
Process was flawed |
Level misread |
Before closing any idea at tp1 or stop, reconstruct the fill timeline first: a limit entry counts as filled only from the first regular-session bar that trades through it, and only touches AFTER that bar can close the position. Write the fill timestamp in the close reason. The owner should correct CS-2026-0024 in the register: either reopen it (it is still within validity) or restate the exit as 7684 at 17:00 UTC on 26 Oct, about +0.19R. CS-2026-0025 (short 7684, same stop, tp1 7626) was published minutes later and effectively continues the same position; if 0024 is reopened, 0025 is the second idea of the same factor and should be read at half risk. |
| CS-2026-0021 |
MSFT |
-1.00 |
Process was sound |
No process error |
Two marginal points, neither a rule violation. (1) The target cleared the reachability ceiling by under 1% and sat beyond two resistances (551.15, 553.72) that the rationale itself named as the main weakness: when tp1 only passes the ceiling by a hair AND must cross named resistance, prefer no idea or an intraday horizon rather than a 2R swing. (2) Context consistency: on 19 Oct the book already held a US 500 SHORT (CS-2026-0020) and on 21-22 Oct a US Tech 100 SHORT; a long on a stock 60% correlated with the Nasdaq while the house view on the index is short is a relative-value bet, not a trend idea, and needs an idiosyncratic catalyst MSFT did not have. Watch this pattern; if it repeats, make it a rule under the correlation tag. |
| CS-2026-0023 |
US Tech 100 |
2.00 |
Process was sound |
No process error |
Nothing in the rules. Two honest caveats. First, the speed of the outcome owes more to the earnings-night gap than to the setup: the rationale flagged that mega-cap results could gap the index either way, and it gapped our way. Had it gapped the other way the stop at 0.60 ATR would likely have been jumped, so the realised result was close to a coin flip on the event and the 2R should not be read as evidence that the pattern works. Second, the index earnings exposure is not covered by the single-stock earnings filter: for index ideas published the evening before heavyweight reports, keep declaring the gap risk in the rationale as this one did. Keep taking tp1 at 2R near the range floor; the target was reachable in one session only because of the event. |
| CS-2026-0022 |
US 30 |
2.00 |
Process was sound |
No process error |
Nothing in the idea itself. Two timing points for the routine: a target touched minutes after the evening review is only booked at the next slot, so the morning slot must check the full prior-session bar range, not just the close; and when a half-risk hedge idea is published against a main idea, the post-mortem should evaluate the pair, because judging the hedge alone flatters the method. |
| CS-2026-0020 |
US 500 |
-0.71 |
Process was flawed |
Target too far |
On swing ideas, scale the reachability cap with the time actually left, as already done intraday: tp1 must sit within 1.0 x ATR14 x sqrt(sessions to expiry), not just within the 3.0 ATR server cap, which corresponds to nine sessions. With three sessions that is 1.73 ATR and this idea would have been rejected or given a longer validity. When an index factor is split (Nasdaq down, Dow up), short the weak leg, not the S&P 500 blend. |
| CS-2026-0019 |
TSLA |
-1.36 |
Process was sound |
No process error |
Decision followed every rule, so no new rule from a single case. Observation to track: the stop sat only 3.97 (0.36 ATR) above the prior regular-session close, i.e. inside an ordinary overnight gap for a stock with ATR near 3%. On future single-stock swings, also measure stop distance from the last close, not only from entry; if it is under ~0.5 ATR, prefer intraday or pass. If this repeats, it becomes a stop_too_tight rule. |
| CS-2026-0018 |
US Tech 100 |
2.00 |
Process was sound |
No process error |
Nothing in the rules. Two observations to carry: (1) the entry at 1.30 of the 1.50 ATR band worked here but late-band entries remain the riskier subset and should keep a structural target, as this one had; (2) PPI and retail sales (15 Oct 12:30 UTC) fell inside the horizon and are not in the counted macro list - correct per method, but they were the catalyst of the gap, so the rationale should mention second-tier data inside the horizon even when they do not block. |
| CS-2026-0017 |
US 30 |
-0.67 |
Process was sound |
No process error |
Nothing in the rules. One observation to keep under watch, not yet a rule: the declared invalidation sat only 0.32 ATR below entry, inside a normal daily range (today 1.43 ATR), so the thesis was falsified by ordinary noise before the stop. When a breakout level is that close to entry, prefer waiting for a retest of the level as entry instead of buying 100 points above it. Also: if an idea is closed on its invalidation text rather than its stop, write it in the close reason, as done here. |
| CS-2026-0016 |
US 30 |
-1.00 |
Process was flawed |
Target too far |
Before accepting tp1 = R:R x d, measure the reward to the FIRST structural level in the way (prior swing low/high, range boundary). If that reward is under 1.5R, the idea is not published, regardless of the mechanical target. Second point to watch: a short on one US index while the other two sit at +1.5 ATR above MA20 is a range bet inside an uptrend, not a downtrend; this is the second US 30 short stopped in a week (CS-2026-0013 too). |
| CS-2026-0014 |
XOM |
-1.00 |
Process was flawed |
Data release overlooked |
For a single stock, treat the sector release that drives it (oil inventories for XOM) like a macro event for the exposure window, even though the generic macro list does not name it; and do not buy a breakout on a single stock whose underlying driver is moving the other way. If the trade must be held through such a release, use the standard swing stop distance (0.8 ATR) only when the invalidation supports it, otherwise skip. One occurrence: recorded as a hypothesis, not yet a rule. |
| CS-2026-0013 |
US 30 |
-1.00 |
Process was sound |
No process error |
No rule change. One hypothesis to watch, not yet a rule (n=1): the short was taken while the same risk factor was strongly the other way (US 500 +2.2 ATR and US Tech 100 +3.2 ATR above their MA20). When the Dow is the only index of the factor below its MA20, record that divergence explicitly in the idea and check in later post-mortems whether such shorts underperform. |
| CS-2026-0012 |
WMT |
0.00 |
Inconclusive |
No process error |
1) Publish invalidation only as a level the stop itself enforces; a 15-minute-close condition cannot be monitored between revisions and creates a second, contradictory exit rule. 2) Expiry closes must record the actual price at the mandatory exit time, not null/0R: CS-2026-0011 and CS-2026-0012 both show exit_price null with 0R, which flattens the track record. 3) Second consecutive WMT short that expired flat needing a new session low in the last 3 hours: treat late-session intraday targets that require breaking the day's low as low-probability and size confidence accordingly. |
| CS-2026-0011 |
WMT |
0.00 |
Process was sound |
No process error |
Process: none for the idea itself; keep the thin-margin caveat in mind (targets near the end of the reachability window often get hit only in the closing auction). System: the automatic expiry close at 20:00 does not check whether tp1/stop were touched before the mandatory close and writes exit_price=null and 0R. That understates this idea by 1.51R and must be verified on the other auto-expired ideas (CS-2026-0010 US 30, also 0R at 20:00:01). Owner notified; the record was NOT edited from here. |
| CS-2026-0009 |
JPM |
2.00 |
Process was sound |
No process error |
Nothing structural. The two ingredients that made this work should stay the default for single stocks: a limit entry on the retest of a broken level rather than chasing the break, and an expiry placed before the next macro release so the swing is not exposed to it. One honest caveat: this is one winning trade on a stock whose previous short (CS-2026-0003) also won, and two JPM wins are not evidence that financials shorts have an edge; do not raise confidence on JPM because of them. |
| CS-2026-0010 |
US 30 |
0.00 |
Process was sound |
No process error |
Process was sound on the checklist, the result was luck in the second half: the fill came only in the spike that also broke the written invalidation, so for 30 minutes the idea was open with its own thesis already wrong and only the noise margin keeping it alive. Two changes: (1) a limit entry closer than about 0.25 ATR to the invalidation level (here 111 points = 0.21 ATR) means fill and invalidation are almost the same event - either enter at market or skip; (2) the automatic expiry must check the fill against regular-session bars before recording r=0; flagged to the owner as a track-record integrity issue. |
| CS-2026-0008 |
TSLA |
-1.00 |
Process was sound |
No process error |
The loss itself is a normal -1R on a well-defined invalidation. Two things to flag honestly. (1) The entry was 1.34 ATR above the 20-day average, inside the 1.5 band limit but near its top: late-in-band longs have now produced two -1R stops out of two (US 500 CS-2026-0007 at the edge and this one). Not yet a rule, but next time a long sits above 1.2 ATR from MA20, prefer a limit entry back toward 1.0 ATR. (2) Publication defect: the public rationale says the suggested risk is half, 0.5%, while the published record carries risk_pct 0.25. The reader saw two different numbers for the same idea. Before sending a second-in-factor idea, pass risk_pct equal to the number written in the text and check the value the server stores in its reply. |
| CS-2026-0004 |
NVDA |
0.37 |
Process was sound |
No process error |
Process was checklist-compliant: invalidation first, stop 0.94 ATR, target 1.88 ATR inside the 3.0 swing cap. The one lesson worth watching (not yet a rule, n=1): the target sat just below a known congestion ceiling (230.40-234.76) that price had to cross in full, and that is exactly where the move stalled. When tp1 requires traversing a congestion band rather than reaching its lower edge, note in the rationale that a partial is likely and consider placing tp1 at the band's near edge if R:R still clears 2.0. |
| CS-2026-0006 |
XOM |
-1.05 |
Process was sound |
No process error |
No rule was broken, so no new rule. One observation to track, not to act on yet: with the entry only 0.68 ATR from the MA20, the stop (0.84 ATR away) sat just beyond the MA20 itself, so an ordinary mean reversion was enough to invalidate. If more stop-outs show the same geometry (entry in the lower half of the band, stop beyond the MA20), consider requiring the invalidation level to be on the far side of the MA20 from the stop by a margin, or preferring entries deeper in the band. |
| CS-2026-0007 |
US 500 |
-1.00 |
Process was flawed |
Entry too late |
When price is within 0.2 ATR of the 1.5 ATR late-entry limit, do not buy the extension: either wait for a pullback toward 1.0 ATR from the MA20 or skip. For pre-market index ideas, prefer an entry level the cash can actually trade after the open over an estimated at-market price. |
| CS-2026-0003 |
JPM |
1.38 |
Process was sound |
No process error |
The real gap this trade exposes is not in the levels but in the exit: the method has no rule for an idea that is working and profitable when its validity window runs out. The +1.38R was banked by a calendar, not by a plan - had this review fallen an hour earlier the same idea would have closed nearer +1.0R, and had it fallen tomorrow morning it would have closed on whatever the overnight left. Concretely: a swing idea should declare, at publication, what happens at the end of validity - close at market at the last review inside the window (what was done here), or roll the stop to the entry and let the target decide. Second, smaller point: a 2.23 ATR target on a five-session horizon passed the 3.0 ATR reachability cap but came up short, and 1.88 ATR of travel in five sessions is the honest yardstick to size the next one against. |
| CS-2026-0005 |
US 30 |
1.56 |
Process was sound |
No process error |
One thing to keep watching rather than to change. The 52,195.65 entry was a price estimated from the future 2.5 hours before the cash session opened, and the cash then opened 0.24 ATR ABOVE it. That is what made the short fill at all: had the open been 0.24 ATR lower, a thesis that turned out to be right would have produced nothing. The midday slot cannot avoid this, but it should be measured - log, for each midday idea, the gap between the estimated entry and the actual cash open, so that after ten ideas we know whether the estimate is unbiased or systematically on one side. No change to the levels or the filters is warranted from this trade. |
| CS-2026-0002 |
US 500 |
-1.00 |
Process was sound |
No process error |
Nessuna regola violata e nessun livello letto male: l'invalidazione era strutturale, lo stop ne discendeva nell'ordine corretto e la perdita e' rimasta esattamente a 1R. Il processo va confermato, non riscritto. L'unico affinamento da mettere alla prova su piu' casi, non da promuovere a regola oggi con un solo esempio: quando si entra contro un rimbalzo, misurare anche la VARIAZIONE dello scostamento dalla MA20 nelle due sedute precedenti, non solo il suo valore assoluto. Uno scostamento che si e' gia' dimezzato in una seduta e mezza segnala che il prezzo sta correndo verso il livello di invalidazione: la banda lo giudica ancora valido, ma la probabilita' di toccare lo stop e' molto piu' alta che con uno scostamento stabile o in allargamento. Da annotare nei prossimi post-mortem per vedere se il modello si ripete prima di codificarlo. |
| CS-2026-0001 |
XAU/USD |
-0.31 |
Process was flawed |
Level misread |
Prima di pubblicare, verificare che ref_price, MA20, ATR14, ingresso, stop e invalidazione provengano TUTTI dalla stessa serie di prezzo, e dichiarare quale serie e'. Controllo pratico in un passaggio: se un livello citato coincide alla seconda cifra decimale con un massimo o un minimo della serie del future, quel livello non e' spot. Sopra lo 0,5% di base le due serie non sono intercambiabili e la conversione va fatta esplicitamente, non lasciata implicita nel testo. |