February 2026 Monthly Recap: 24 Trades, 62.5 Percent, +6.64R Net

Restated: Gold (XAUUSD) was part of SkyAnalyst's coverage from inception (Jan 12, 2026) through May 2026. We've since narrowed coverage to six instruments — EURUSD, GBPUSD, USDJPY, US30, NAS100, US500 — and these numbers are restated for the current lineup. The original publish date is preserved; cumulative figures have been recomputed.

February 2026 was the first full calendar month of the published record and the first month every published entry passed the full four-agent gate. The system took twenty-four trades across three active equity indices, closed at +6.64R net on the TP1 baseline, and produced a 62.5 percent win rate. Fifteen winners, nine losers. The simulated $100,000 account at 2 percent risk per trade closed February at $113,279. Through Mar 1, 2026, the system has banked +9.66R YTD across 27 trades from Jan 12 inception. The simulated $100,000 account at 2 percent risk per trade sits at $119,324.61 on a static basis. One methodology note up front: every R-multiple in this recap is computed on a TP1 exit for every winner. That is our conservative baseline. Six of February's fifteen winners ran past TP1 to TP3 in the live broker fills, and the dollar lift on the credited line under-counts the realized broker P&L on those runners. The +6.64R is the floor of the credited projection, not the realized ceiling. We point the next reader to the March monthly recap for the second full month on the record and to the January monthly recap for the soft-launch context that precedes this window.

Act 1: Feb 4-13, the opening swing and the first dense session

The month opened on Feb 4 with a US30 long that stopped at minus 1R, the only entry of the opening week and an unhelpful first print. Feb 9 to Feb 10 produced three US30 trades: a pullback-to-support long that paid +1.31R on TP2, a VWAP/EMA confluence long that stopped, and a pullback to VWAP / 61.8 long that paid +2.30R on TP1. The Feb 10 print was the largest single TP1 credit of the opening swing. Cumulative MTD reached +1.61R into the weekend.

Feb 11 produced the standout trade of the month: a US30 responsive long off intraday support that ran +2.64R on TP1, the highest credited R-multiple of February and a standalone case study. By the Feb 11 close, cumulative MTD sat at +4.24R.

Feb 13 produced the densest single session of the month: five trades inside forty-two minutes. A US30 short paid +0.57R on TP1, a NAS100 short stopped, a US500 short paid +0.74R on TP1, a US30 short stopped, a US500 long paid +0.55R on TP2. Three winners, two losers across one session. The simulated account exited the first ten sessions at $108,203.

Act 2: Feb 17-20, the cleanup week and the first TP3 prints

The third week opened steady. Feb 17 produced two TP1 winners inside the same minute: a NAS100 pullback-to-go long for +0.62R credited and a US30 pro-trend intraday long for +0.43R credited. Cumulative climbed to +5.15R. Feb 19 brought a NAS100 continuation short that stopped at minus 1R, then a US30 SHORT VWAP/EMA fade that ran the full ladder to TP3 for +1.08R credited. That Feb 19 trade became a standalone case study and the first TP3 print of the month.

Feb 20 added a US30 short pullback-to-supply that also ran to TP3 for +0.61R credited, the second case study of the week. Cumulative reached +5.84R into the weekend. The simulated account closed the week at $111,670.

The Feb 19 to 20 sequence is the cleanest example in February of the same trade architecture working twice on the same instrument on consecutive sessions: structural fade against an intraday counter-trend, Macro gate not flipped, Trend confluence cleared, Risk sized off independent confluence reads. Two TP3 hits in two days.

Act 3: Feb 24-27, the closing bumper and the densest TP3 cluster

The closing week is the densest TP3 cluster of the published record so far. Feb 24 produced three trades: a NAS100 pullback long to TP3 for +0.60R credited, a US30 mean-revert short that stopped at minus 1R, and a NAS100 buy-the-dip into reclaimed VWAP/EMAs that ran to TP3 for +0.69R credited. Both NAS100 winners became case studies, the pullback buy and the reclaim trade. Cumulative climbed to +6.13R.

Feb 25 added a NAS100 breakout-and-retest long to TP2 for +0.60R credited. Feb 26 produced a NAS100 trend-continuation short on a weak retest that stopped and a US30 buy-the-dip that also stopped: two minus 1R losers on the same session. Cumulative dipped to +4.72R, the only meaningful giveback of the month.

Feb 27 was the closing payoff. A US500 intraday fade into resistance ran to TP3 for +1.19R credited (the highest single TP3 credit of the month), then a US30 primary fade ran to TP3 for +1.73R credited (the largest single credited R of the closing week), then a US500 pullback buy stopped at minus 1R. Two TP3 prints and a stop inside twenty-six minutes. Both winners became case studies, the US500 fade and the US30 primary fade. Cumulative closed at +6.64R for a final simulated balance of $113,279.

Pattern of the week

The pattern that defined February was the fade against an intraday counter-trend, applied across US30, US500, and NAS100, and concentrated in the back half of the month. Five of the six TP3 winners fit this template: Feb 19 US30 short on a VWAP/EMA fade, Feb 20 US30 short on a pullback-to-supply, Feb 27 US500 short on an intraday fade into resistance, Feb 27 US30 short on a primary fade, and the Feb 24 NAS100 buy-the-dip into reclaimed VWAP/EMAs, which is the long-side mirror of the same logic.

Why the fade kept paying through Feb 19 to 27

The cleanest signal in February is that the same Trend Agent confluence read kept clearing on counter-trend fades inside sessions that had set up like continuation prints. The Macro Agent did not flip its gate. The Trend threshold did not change. Cross-Asset confirmed the correlated instruments were inside tolerance, and Risk sized each entry off independent confluence math. The fade kept paying because the regime kept setting up the same way: a morning continuation impulse that exhausted into a structural level, then a reversal back through VWAP, then a trend-day move in the opposite direction. The system caught the reversal six times across the back half of the month, with three of those prints clustered inside Feb 19 to Feb 27.

Decision highlights

The Feb 13 dense session is the cleanest example in February of the system holding posture through a fast tape. Inside forty-two minutes the system fired five trades across three instruments: a US30 short, a NAS100 short, a US500 short, a US30 short, and a US500 long. Three winners, two losers, all sized off independent confluence reads. A discretionary trader watching five setups clear inside the same window would have rationed entries by gut feel; the system sized each off its own threshold and let the broker decide which ran and which stopped.

The Feb 19 to 20 back-to-back TP3 sequence on US30 demonstrated what the same architecture looks like when the regime stays cooperative for forty-eight hours. Feb 19 ran the VWAP/EMA fade to TP3 for +1.08R credited, and Feb 20 ran the pullback-to-supply to TP3 for +0.61R credited. Same instrument, same trade direction, same structural fade premise, two consecutive sessions. Neither entry consulted the other. The Trend Agent's confluence cleared on each independently, the Macro gate did not flip between them, and Risk sized them off their own R math.

The Feb 27 closing-bumper sequence is the densest TP3 print of the month: a US500 short to TP3 at 16:13 UTC for +1.19R credited, then a US30 short to TP3 at 16:33 UTC for +1.73R credited, then a US500 long stop at 16:39 UTC. Two TP3 hits and a stop across correlated equity indices inside twenty-six minutes. Cross-Asset did not veto the back-to-back fades because the structural premises were independent at the bar level. The third entry, the US500 long, was a counter-direction setup the playbook also approves; it took the minus 1R on the same logic that paid +1.73R on the prior US30 short.

Loss worth learning from

The cleanest loss the month produced was the Feb 26 cluster. Two trades, both minus 1R, inside a single session. A NAS100 trend-continuation short on a weak retest stopped at 15:49 UTC, and a US30 buy-the-dip stopped at 16:11 UTC. Two correlated equity-index entries, both fired off independent confluence reads, both stopped against the next session-leg move.

What the system saw that was right

A back-and-fill regime that had paid on Feb 24 (NAS100 longs to TP3) and Feb 25 (NAS100 breakout-retest to TP2). The Trend Agent's confluence cleared threshold on both Feb 26 entries. The Macro Agent's gate had not flipped. Cross-Asset confirmed the correlations were inside tolerance. The setups looked like continuations of the same tape that had paid for two prior sessions.

What the system got wrong

The Feb 26 tape reset intraday in a way the evaluation rhythm could not detect ahead of time. The NAS100 short and the US30 long were directionally opposite, which means the system was reading both sides of the equity-index book as setups inside the same window. Cross-Asset did not veto either because the structural premises were independent at the bar level. Both stopped within a twenty-two-minute window when the session reversed against each entry's primary thesis.

What we would do the same

The entry rule is the entry rule. The system does not consult the recent trade record to size or decline a new entry. If the inputs cleared threshold, the entries were correct under the same logic that produced the Feb 19, Feb 20, Feb 24, and Feb 27 TP3 winners. Streak-aware filters would lower realized expectancy across calendar windows, not raise it. The Feb 26 cluster was the cost of holding posture; the Feb 27 closing sequence was the payoff the next session. Both came from the same playbook.

From the desk

Through Mar 1, 2026, the cumulative ledger reads +9.66R YTD across 27 trades from Jan 12 inception. The same $100,000 account at 2 percent risk per trade sits at $119,324.61 on the static line and $120,466.28 on the compounded line. The compounded line runs $1,141.67 above the static line because February's winners clustered inside a positive-expectancy stretch, and compounding through a positive run multiplies each R-print against a slightly larger base. That spread is the structural argument for disciplined sizing: identical R, traded through 2 percent of equity rather than 2 percent of the starting balance, produces a measurably different dollar outcome over even a two-month window.

The honest reading of February is that the system delivered the kind of month a positive-expectancy playbook is supposed to deliver across a 24-trade sample: a clean opening swing, a steady cleanup week, and a closing bumper that concentrated the TP3 prints. The Feb 27 close added +2.92R on two TP3 prints inside twenty-six minutes, which is the densest credit-per-time print of the month. Subtract Feb 27 and February still closes at +3.72R net, comfortably above flat.

The simulated $100,000 account closed February at $113,279. That is up $11,715.78 from the January-end balance of $101,563.94, on twenty-three more trades than the January window produced and with the four-agent gate at full operational tempo for the first time. Six of February's fifteen winners ran to TP3 in the live broker fills. A subscriber on the published scale-out plan would have closed February meaningfully above $113,279 on the same trades, because the recap underweights those runners by design.

What carries into March is the same playbook and the same four-agent gate. The cross-model panel goes live in March, which is the first month with attributable Claude-side entries running alongside the GPT-5 baseline running alongside each other. February is the baseline. March is the first month that begins to read as a comparison.

Weekly views: Feb 2-8, Feb 9-15, Feb 16-22, Feb 23-Mar 1. Prior monthly: January 2026 monthly recap. Next monthly: March 2026 monthly recap.

What we're tuning

February did not produce a tuning signal. A 62.5 percent win rate at +6.64R net across twenty-four trades is inside the playbook's expected distribution for a regime where equity-index continuations and intraday fades both pay through the back half of the month. The instrument-level dispersion (US30 plus 5.66R, US500 plus 1.47R, NAS100 essentially flat) is a sample-size artifact across the smaller cells, not a structural fit-issue with any single instrument.

What we are tracking forward into March: whether NAS100 stabilizes or continues to drift flat, whether the fade-into-counter-trend setup library keeps converting at the Feb 19 to 27 cadence, and whether the FX book (EURUSD, GBPUSD, USDJPY) produces a qualifying setup once the cross-model panel goes live in March. None are immediate tuning items. They are the variables we watch through the rolling-100 window before any threshold gets moved.