Today's morning brief
Paper only. Every figure below is from the Desk's own files or a
named fetched source. All strategies stay UNPROVEN. Kill switch
ARMED, not tripped — no trip since the last brief.
The market in 5 lines
• Stocks: no US session today (Saturday) — the stock section rests on Friday's close; crypto runs unchanged.
• SPY closed Friday 2026-11-06 at $735.29, −2.24% Monday→Friday (Nasdaq daily bars, Desk data files).
• QQQ closed Friday at $656.10, −3.67% Monday→Friday (Nasdaq) — growth led the week's decline.
• Thursday was the break: SPY fell 4.41% in one session, $766.81 → $733.00, then steadied Friday (Nasdaq).
• Bitcoin closed Friday at $75,844, −1.27% Monday→Friday (Binance Vision); at 07:26 ET today the partial Saturday candle read $75,907 on Kraken and $76,007 on Binance Vision — a flat bounce, not a reversal.
Today's high-impact events
• Saturday: no US session and no scheduled macro release on the Desk's calendar.
• The Desk has no wired earnings / FOMC / CPI calendar source yet (DATA_SOURCES.md), so it cannot verify a full schedule — it states plainly that none is on its calendar rather than guessing that none exists.
• No new strategy launches into a ticker's earnings print, an FOMC meeting, or a CPI print (DESK_RULES §4) — and no launch is scheduled today in any case.
• Next stock session: Monday 2026-11-09, the first paper day for the two strategies already approved (below).
The 3 strategies (from scores.json, all UNPROVEN)
**1. 20-day Donchian breakout, volume-confirmed — 22/25. APPROVED-PAPER, effective Monday 2026-11-09.**
Out-of-sample (2024–26): 264 trades, +$49.75 after costs on the $100 book, Sharpe 1.27, deflated Sharpe 0.99 (results.json).
Max drawdown 16.9%; 1,000-run Monte Carlo worst drawdown −21.1%.
Stress test (one bar late, double slippage): +$28.49 — survives; the 2026 partial walk-forward window is negative (−$12.08), stated plainly.
WHY THIS COULD FAIL: a 43% win rate lives on a fat right tail; a choppy market that truncates winners removes the whole edge.
**2. Oversold snap-back, stocks — 22/25. APPROVED-PAPER, effective Monday 2026-11-09.**
Out-of-sample: 212 trades, +$20.91 after costs, Sharpe 1.10, deflated Sharpe 0.97, win rate 64.2% (results.json).
Max drawdown 6.2% — the shallowest of the three; Monte Carlo worst drawdown −12.6%.
Stress test: +$4.56 — survives, but thin: expectancy is $0.10 a trade, so real paper fill quality decides whether the edge is real.
WHY THIS COULD FAIL: in a true bear, "oversold in an uptrend" turns into falling-knife catching one leg before the uptrend filter turns off.
**3. Sector rotation — 19/25 as designed; capped variant 20/25. NOT paper-trading — see the decision below.**
As designed (top-2 weekly): out-of-sample 67 trades, +$33.02 after costs, Sharpe 1.03, max drawdown 18.3%, Monte Carlo worst −27.0% (results.json).
Capped variant (top-4 sectors at $25 per leg, retested 2026-11-07): 20/25, 99 out-of-sample trades, +$32.34 after costs, Sharpe 0.90, profit factor 2.16, max drawdown 15.1%.
The capped variant cleared every §5 gate (≥18/25, ≥30 OOS trades, positive after costs); double-slippage stress stays positive at +$29.84; Monte Carlo worst drawdown improves to −15.5%.
WHY THIS COULD FAIL: leadership whipsaw — a 3-month lookback buys last quarter's winner at the top when rotation speeds up; the capped variant's 2022 leg was −$3.50.
Risk Officer vetoes (final unless you override in words)
• **Idea 6 (sector rotation), as-designed sizing — vetoed.** Two $50 legs breach the Desk Rules 25% position cap ($25). A capped variant must be retested before any paper fill. That retest is now done (above); the veto still stands on the as-designed $50 sizing.
• **Ideas 7, 8, 4, 1 — vetoed as paper candidates.** Out-of-sample trade counts of 16, 14, 20 and 19 fall below the 30-trade floor in Desk Rules §5.
• **Idea 9 (SPY shock reversal), design — vetoed.** No priced stop; §1 requires stop distance × size to stay within the $1 risk cap. The idea was also killed on its evidence.
Desk status
• Paper equity $100.00; paper trades made: 0. Overnight paper trades: none.
• Kill switch: ARMED, not tripped. No daily-loss, losing-streak, or stale-data trip is on record; overnight Scout crypto scans (latest 06:59 ET) found Kraken, Binance Vision, and Coinbase closes in agreement, with no stale feed.
• Scout overnight: no crypto setup scored at 7/10 or higher — no handoff, no fill. XRP snap-back watch scored 6/10 and stays watch-only; the crypto snap-back strategy (Idea 4) is killed and not approved.
• Ideas 5 and 3 were approved for paper trading on your delegated "Your choice" decision (2026-11-07), effective Monday 2026-11-09. Today is before that date, so nothing has filled.
• Alpaca paper account: the Desk's own files still record it as not created — signup needs Douglas at a browser to set the password through the secure vault. The Desk loses nothing by it: every test and fill runs on the verified free routes (Nasdaq, Kraken, Binance Vision, EDGAR).
Your one decision
**Approve, reject, or retest — one item only: the capped sector rotation.**
The Desk proposes: approve Idea 6's capped variant (top-4 sectors at $25 per leg, 20/25, 99 OOS trades, +$32.34 after costs) for paper trading from Monday 2026-11-09, alongside Ideas 5 and 3, under the $1-risk and $25-cap rules. It stays UNPROVEN until it clears the two-week paper period (§8).
The as-designed top-2 $50 version stays vetoed either way. If you reject, the variant returns to the research file and does not paper-trade. No retest is pending — the capped retest already ran and passed.
DECISION ADDENDUM — 2026-11-07
The decision above is taken. Douglas delegated it ("Your decision", 2026-11-07) and the Chief of Staff approved Idea 6's capped variant (sector rotation, top-4 sectors at $25 per leg) for paper trading from Monday 2026-11-09, alongside Ideas 5 and 3, under the $1-risk and $25-cap rules. It stays UNPROVEN until it clears the two-week paper period (§8). The as-designed top-2 $50 version stays vetoed.
Correction to the Desk status section above: its Alpaca line is stale. Douglas created the Alpaca paper account on 2026-11-07 — paper side only, under his Zoho identity. The Desk's file lagged; the account exists. The Desk still trades its own $100 model book under its rules; no live account, bank link, or deposit exists.
As of 2026-11-07 07:36 ET.
