Six Winners by Wednesday. Thursday Sold What Was Left.

The week was made in three sessions. From Monday to Wednesday the desk took six trades, five shorts and one long, and all six closed as winners. Cable was sold on three consecutive days. The Dow and the Nasdaq were sold on Wednesday. The only long of the week, a Nasdaq pullback on Tuesday, ran to its second target. Then Thursday. Three more shorts inside forty-one minutes, NAS100, EURUSD and GBPUSD, and all three stopped. The window closed at +2.46R across nine trades and 4,920 dollars on a simulated 100,000 dollar account at 2% risk, down from +5.46R at Wednesday's close. The year to date ledger stands at +28.75R YTD from Jan 12 inception, sealed at the August close, which puts the same simulated account at $157,519.50. This week sits inside September, which is still open and not yet part of that figure. A week that makes 5.46R and hands back 3R of it is not a disaster and not a triumph. It is the shape the year is built from.

Monday and Tuesday: the dollar trade

The first fill of the week was a GBPUSD short at 1.33735 on Monday, 15:06 UTC. It reached its third target at 1.33365 early on Tuesday. On Tuesday the same book sold again, at 1.33607, and was at the third target of 1.3326 an hour and forty-three minutes later.

Between those two, the Nasdaq long book took the one long of the week. NAS100 at 30,635.3 against a stop at 30,568, on a pullback into the zone it had just broken out of. It cleared the first target at 30,718 and the second at 30,800, and printed a high of 30,820.4. That is the win of the week at +1.23R on the baseline this recap uses.

Wednesday: everything sold

Wednesday is the day the yield move landed on equities. The Dow short book sold US30 at 51,827.4 at 14:11 UTC and was at its third target of 51,510.6 by 17:03. The Nasdaq short book sold NAS100 at 30,550.5 at 14:59, forty-eight minutes after the Dow, and the position reached its third target of 30,250 overnight. Cable was sold a third time at 1.3263, ten minutes after that, and it too ran to its third target the next day.

By the close the week stood at +5.46R. Six trades, six winners, and the three that filled on Wednesday were three expressions of one idea: rates up, dollar up, risk down.

Thursday: selling the bounce

On Thursday morning the Nasdaq had already gapped 400 points lower overnight, from a 30,497.6 close to a 30,101 London low, and was bouncing. The desk sold that bounce. NAS100 short at 30,342.5 at 14:23 UTC, EURUSD short at 1.13759 at 14:49, GBPUSD short at 1.32142 at 15:04.

The Nasdaq trade went more than its full risk in our favor, to 30,232.4, without reaching its first target, then reversed through the stop at 30,420. Cable stopped at 1.3231 seventy-one minutes after it filled. The euro held overnight and stopped at 1.13998 on Friday morning. Three trades, minus 3R, and no new position on Friday. The direction had been right all week. By Thursday the move it described was mostly over.

Pattern of the week

The pattern this week was the pullback sell in a trending tape. Five of the six winners were shorts taken on a retrace into resistance, a retest of a broken low or a VWAP rejection, in the direction of the dollar and against duration. The setup names in the index say it plainly: pivot retest, retrace rejection, failed reclaim, pullback to breakdown zone, London pullback.

Thursday's three trades were the same pattern. The labels were VWAP rejection, VWAP rejection and continuation, and a breakdown retest on Cable. The macro read behind them was, if anything, stronger than it had been earlier in the week: the Macro Agent had EURUSD at strong bear with 87% confidence and GBPUSD bearish at 80%, and the Nasdaq plan called yields, the dollar and VIX a triple confirmation.

So the pattern did not change between Wednesday and Thursday. The location did. A pullback sell works when the trend has room left to run. On Thursday the Nasdaq was selling a bounce off a 400-point gap, and the dollar was selling into a two-session rally that had already carried DXY to 101.33. The same shape, taken late in the move, paid less and failed faster.

Decision highlights

The Cable short book sold the pair three days in a row and never carried a view from one session to the next. Each entry was scored from scratch against that morning's tape, at 1.33735, then 1.33607, then 1.3263, and each one ran to its third target. That is what a system with no memory of yesterday's win looks like when the trend holds: it keeps finding the same trade because the market keeps offering it.

The Thursday Nasdaq short is the decision we would take back. The session plan asked for entry only after 10:30 ET, both because of the presidential remarks at 10:15 and because price was still rallying through VWAP. The approval came at 10:16 ET on an alternative trigger, two closes below VWAP, and it said in its own words that the entry was early on timing. The fill came at 10:23, the trade went 110 points in our favor, and it still stopped at 30,420 before it reached its first target.

The Thursday Cable short is the same judgment in a smaller frame. The plan wanted a closed five minute rejection candle on the retest of 1.32150, and the first evaluation declined because that candle had not closed. Less than ninety seconds later the approval went through at 62% confidence, describing the entry as tradeable but not fully confirmed, and the position was stopped at 1.3231 seventy-one minutes later.

Loss worth learning from

Where the 3R went

Thursday accounts for all three losses and the full 3R the week gave back from its Wednesday peak. Each was exactly 1R, which is the risk model doing its job.

The euro short is the one that does not teardown into a mistake. It waited through two declined evaluations, entered only after a five minute candle wicked into the 1.13760 zone and closed back below 1.13750, and held overnight with the macro read at 87% bearish. It ran sixteen pips our way and was stopped the next morning. Correct by every rule we have written down, and still a loss.

The Nasdaq and Cable shorts are different. Both were approved ahead of conditions their own plans had set, one on the clock and one on a candle that had not finished printing. Both were in the direction the whole desk had been right about for three days, which is precisely when a gate is most tempted to accept most of a trigger instead of all of it. We would rather miss a trade that the plan says is not ready than take it and be right about the direction and wrong about the hour.

From the desk

The week added +2.46R and 4,920 dollars to the simulated account at 2% risk. Compounded through the equity curve, where each trade is sized off the balance it actually had, the same nine trades took 100,000 dollars to $104,921.79.

The year to date ledger reads +28.75R across 220 trades at a 56.82% win rate from Jan 12 inception, sealed at the August close. Run through a static 100,000 dollar account at 2% risk, that is $157,519.50. Run through the same account compounding, it is $169,638.84. The gap, a little over 12,000 dollars, is not extra edge. It is the same R, banked in an order that lets each winner size the next trade off a larger balance, and it only holds because no single week has been large enough to break the sizing. A week that gives back 3R in a single session barely moves either line. That is the argument for fixed fractional risk, and it is the reason a Thursday like this one is a footnote rather than a problem.

One piece of housekeeping on the numbers. September is still open, so it is not in the year to date figure. The ledger has +6.39R recorded for the month so far, including this week, and it joins the year to date number when the month closes.

Next week the desk will read each session from scratch, as it did this week. If the dollar trend resumes, the same books will find the same trade. If it has run its course, the plans will say so, and this week we learned to hold the approvals to what the plans say.

What we're tuning

We are looking at making the timing and candle-close conditions in a session plan binding at the approval step, rather than inputs the approver can weigh against a strong directional read. Thursday's two early approvals both said, in writing, that they were early. A rule the approver can read and then set aside is a preference, not a gate.

We are not touching direction or the macro gate. The desk sold the dollar trade correctly for three days, and the three Thursday shorts were in the right direction too. The question this week raised is narrower: how late in a move a pullback sell should still be allowed, and whether an entry that arrives after the market has already made most of the move should need more confirmation than one that arrives early.