Friday, 07-17-2026
Evening market read · post-close validation & recap
conv field the rollups require, so both calibration paths skip them. Trailing-50 holds at Brier 0.2228 / 60%; all-time 0.2486 / 48.9% (n=133) — unchanged. Filed ISS-20260717-07.Ten setups scored against tonight’s regular-session close, all from The Early Bird Curd (07-17). 3 FIRE / 3 VOID / 4 MIXED / 0 NO_EVIDENCE. Backlog after this run: 0 open.
| Setup | Outcome | Evidence | Δ-ATR* | Actual |
|---|---|---|---|---|
| LRT SHORT QQQmm-260717-LRT-QQQ-S | FIRE | Failed-reclaim continuation played clean: the bounce to 702.30 failed against the 702.61 prior-low / 708.59 prior-VWAP, then flushed to a 686.76 low — inside the 685–690 target. Closed 695.33 (−1.50%, −0.72 ATR). SMH −2.18% stayed red; neither kill (reclaim/hold >708.59, or SMH green) fired. | — | 695.33 |
| AF LONG XLEmm-260717-AF-XLE-L | FIRE | Closed 57.68 (+1.16%, +0.62 ATR) — energy the day’s lone green sector on the Russia diesel shock (USO +3.91%); no de-escalation / crude-reversal kill. | — | 57.68 |
| AF LONG XLPmm-260717-AF-XLP-L | VOID | Closed 85.19 (−0.72%, −0.42 ATR against). The “defensives bid” thesis failed — staples sold with the broad tape. This was de-risking, not rotation. | — | 85.19 |
| AF LONG XLVmm-260717-AF-XLV-L | VOID | Closed 161.09 (−0.44%, −0.25 ATR against — right at the decisiveness bar). Health sold alongside the other defensives; no rotation bid materialized. | — | 161.09 |
| AF LONG XLUmm-260717-AF-XLU-L | VOID | Closed 45.17 (−0.66%, −0.43 ATR against). Utilities sold despite the risk-off tape — no haven bid into the defensive complex. | — | 45.17 |
| AF LONG XLREmm-260717-AF-XLRE-L | MIXED | Closed 45.42 (−0.09%, −0.06 ATR) — flat, inside the ±0.25-ATR band; no decisive move either way. | — | 45.42 |
| AF SHORT XLImm-260717-AF-XLI-S | MIXED | Closed 179.41 (+0.35%, −0.21 ATR inside band, against the short). The cyclical bid absorbed the risk-off again — as on 7/16 — but not decisively enough to VOID. | — | 179.41 |
| AF SHORT XLBmm-260717-AF-XLB-S | FIRE | Closed 50.53 (−0.71%, −0.40 ATR in-dir) — materials weak with copper soft; the short worked. | — | 50.53 |
| AF SHORT CPERmm-260717-AF-CPER-S | MIXED | Closed 37.92 (−0.37%, −0.22 ATR inside band) — copper soft in-direction, but the move fell short of the decisiveness bar. | — | 37.92 |
| AF LONG UUPmm-260717-AF-UUP-L | MIXED | Closed 28.33 (−0.04%, −0.09 ATR inside band) — dollar flat near 13-month highs; no follow-through (a tiny 0.11 ATR). | — | 28.33 |
*Δ-ATR not populated this session — today’s AF and LRT setups carried no single numeric lvl field, so the validator’s mechanical Δ-ATR fill is N/A. The AF decisiveness (close-only ±0.25 ATR(14) rubric) is cited in-row instead. All closes confirmed (Massive, regular-session close).
conv field (and context/ev_class/exposed/regime/thresh) that both calibration paths require, so nightcap_prep --rollups and the dashboard skip all ten. Their outcomes are written to the track record and will fold into calibration once the field set is backfilled. The figures here are therefore unchanged from last night. Flagged ISS-20260717-07.The trailing-50 window sits at its four-session best (Brier 0.2228 / hit 60%, from 0.2269 on 07-15) — but it neither improved nor broke tonight, because tonight’s batch never entered it. That is the finding worth flagging: a morning emit that drops conv silently removes an entire validated session from the learning loop.
Pattern performance (Component 13) — edge by pattern × direction, calibration-eligible only
| Pattern · Dir | n | Hit% | Mean prob | Mean Brier | Evidence class |
|---|---|---|---|---|---|
| AF L | 16 | 81% | 0.55 | 0.218 | lore_pending |
| MS S | 5 | 80% | 0.46 | 0.295 | lore_pending |
| SRT S | 6 | 67% | 0.47 | 0.317 | lore_pending |
| AF S | 13 | 62% | 0.55 | 0.231 | lore_pending |
| SRB L | 8 | 62% | 0.43 | 0.299 | lore_pending |
| GFD L | 6 | 50% | 0.41 | 0.259 | practitioner_backtest |
| MS L | 31 | 45% | 0.52 | 0.249 | lore_pending |
| LRT S | 31 | 42% | 0.38 | 0.280 | lore_pending |
| LRB L | 3 | 33% | 0.43 | 0.216 | lore_pending |
| SBD L | 6 | 0% | 0.37 | 0.138 | primary_source |
| SBD S | 3 | 0% | 0.28 | 0.084 | primary_source |
Unchanged from 07-16 (tonight’s ten excluded). Had they scored, the record would read 3 FIRE / 3 VOID / 4 MIXED — AF-L would take a hit (2 of 5 defensive longs fired: XLE yes, XLP/XLV/XLU no) and AF-S would hold (XLB fired; XLI/CPER mixed). The n=1 singles (CGR-L, FBR-S, GFU-S, VAB-L, VSR-L) are omitted for sample.
Calibration trend (Component 14) — rolling-Brier (last 50) by validation date
Distinct source: macro / cross-asset forecast track-record from the thinktank-v2 predictions table — separate from the setup track-record above; the two hit-rates are never combined. est. (thinktank-v2 derived, asOf 2026-07-17) · read-only · DB not mutated.
| Instrument | Dir | Target % | Entry ref | Close | Realized % | Result |
|---|---|---|---|---|---|---|
| SPX (S&P 500) | UP | +0.5 | 7369.0 | 7537.43 | +2.29 | HIT |
| BTC (Bitcoin) | UP | +4.0 | 60365 | 63086 | +4.51 | HIT |
| ES (S&P Sep) | DOWN | −1.5 | 7385.25 | 7537.43 | +2.06 | MISS |
| NQ (Nasdaq Sep) | DOWN | −2.5 | 29346.75 | 29697.87 | +1.20 | MISS |
| NQ (Nasdaq Sep) | SHORT | −3.0 | 29200.0 | 29697.87 | +1.71 | MISS |
| GC (Gold, spot) | UP | +2.0 | 4046.0 | 4104.10 | +1.44 | MISS |
Unresolved (no price series — score N/A): XLK/XLU pair · IWM-vs-QQQ pair · Brent (BZ Aug) · USD/JPY · CBOT Wheat. Pair trades and futures-only symbols have no prices series in the DB; they are never guessed. Nothing new resolved since 07-16.
| Index | Close | Day % |
|---|---|---|
| SPY (S&P 500) | 743.29 | −0.99% |
| QQQ (Nasdaq-100) | 695.33 | −1.50% |
| RSP (S&P equal-wt) | 213.37 | −0.79% |
| IWM (Russell 2000) | 294.04 | −0.52% |
| Proxy | Close | Day % | Read |
|---|---|---|---|
| UUP (US dollar) | 28.33 | −0.04% | Flat, holding near 13-month highs |
| USO (WTI oil) | 123.96 | +3.91% | Russia diesel supply shock — the day’s driver |
| GLD (gold) | 368.41 | +0.94% | Haven bid returns after four unwind sessions |
| TLT (20y+ USTs) | 84.52 | +0.37% | Duration bid — mild flight to safety |
| HYG (HY credit) | 79.65 | −0.19% | Spreads slightly wider — faint risk-off |
| CPER (copper) | 37.92 | −0.37% | Growth-cyclical caution |
This automated Nightcap run did not fetch the closing-bell newsletters (Stocktwits Daily Rip close / Axios Closer) — single-name narrative color is refresh-required. The confirmed close tape stands in as the illustrative read.
Illustrative only — never trade recommendations. Individual-name detail is refresh-required (no newsletter pull this run).
Closing the loop on The Early Bird Curd (07-17). The morning lens called a risk-off session — a QQQ break of the 710 shelf, a continued semis de-risk, an energy/inflation overlay, and defensives as the safe-harbor bid. Reality vindicated most of the frame and one specific leg spectacularly, and refuted the internal rotation call. Right: the risk-off itself (SPY −0.99%, QQQ −1.50%), the tech-derisk short (LRT-QQQ fired clean to its 685–690 target), and the energy overlay — XLE was the only green sector on the Russia diesel shock (USO +3.91%). Wrong: the “defensive rotation, defensives bid” thesis — this was broad de-risking, equal-weight fell, and Staples / Health / Utilities sold with everything else, VOIDing all three defensive-long leans. The cyclical shorts (XLI, XLB, CPER) landed MIXED as the complex leaked lower but sub-threshold. Net: the regime read (risk-off, tech down, energy up) was right; the rotation read (where the money hides) was the miss — and it is the whole reason tonight’s book went 3 FIRE / 3 VOID / 4 MIXED rather than cleanly green.