The win felt good. The rest of the week did not.

The week at a glance

SNOW hit our exit at $307.65 this week. That part felt good. The rest of the week did not. Day 45 of the experiment. Week return: -0.42%. SPY: +0.11%. Weekly alpha: -0.53%. Cumulative alpha since we started: -0.42%. Win rate on closed trades sits at 0-for-15 — meaning every position that has opened and closed since Day 1 has gone against us. SNOW was open before we started counting, which is why it doesn't appear in that tally. We know that stings to read. We're not hiding it. The Relative Strength Engine drove the week's action. It opened 3 new signals — QLYS, CAKE, and the second and third DFTX entries — closed SNOW for a solid gain, and stopped out both DFTX positions at a loss. Heading into the weekend, we're carrying 12 open CRNX positions — yes, twelve — alongside QLYS, NWL, and CAKE. The book is concentrated. We need to talk about that.

Trades we closed this week

**SNOW (closed +18.04%, alpha +14.98%)** — This one opened July 27 and ran for 37 days. Entry at $270.92, exit at $307.65. The Relative Strength Engine tagged it while it was holding ground as peers drifted lower. That read was correct — SNOW led SPY by nearly 15 points over the hold period. The close was a 'closed_stop,' meaning a trailing mechanism triggered the exit, not a price target. We'd have liked to see it run further, but a +18% gain in just over five weeks isn't a complaint. **DFTX (two entries, both closed at a loss)** — This is where the week got expensive. Two entries: $45.17 on August 25 and $44.76 on August 27. Both stopped out August 31 at $39.22 and $39.95, respectively. Combined P&L: -9.28% and -8.45%. Alpha on both entries ran negative, in the -8% to -10% range. The Relative Strength Engine flagged DFTX twice in three days — a signal that the system saw persistent relative strength setting up. Then the stock broke down sharply before either position had time to breathe. Two stops in the same name in the same week hurt the weekly number in a way you feel. That's how it goes sometimes — and two bad reads in a row do matter, even when the system is working exactly as designed.

What we tuned this week

No system parameter changes this week. What we are paying close attention to is position concentration. The CRNX situation is the thing. The Relative Strength Engine has opened a new CRNX entry on 12 separate days stretching back to August 5. Each individual position is small, and each is sitting on a modest positive P&L — ranging from +0.11% to +1.22%. But in aggregate, that's a lot of exposure to one ticker. The system builds positions this way when a stock keeps refreshing its relative strength signal — it's not a bug, it's how the logic is designed. Still, 12 open lots in the same name is something we're watching closely. If CRNX weakens, those small gains could flip fast and all at once. We also have QLYS at -2.95% on day 10 and CAKE at -2.04% on day 3. Neither has hit a stop yet, but both are working against us. NWL is flat. Next week's number will depend heavily on whether CRNX holds — and whether QLYS and CAKE find their footing.

What we're cautious about next week

The concentrated CRNX book is the clearest risk heading into next week. Twelve open positions in one ticker, each entered at slightly different prices between $83.93 and $84.99, all currently above water — but not by much. One bad day in that name doesn't produce one loss. It produces twelve small ones at the same time. We also skipped 3 signals this week — watchlist days where the system flagged names but we stayed on the sideline, typically for liquidity or portfolio-size reasons. Those skips are part of the process, not missed opportunities. But they do mean the active signal count could grow quickly next week if conditions shift. SPY closed the week essentially flat. If that changes — in either direction — the Relative Strength Engine's signals tend to get noisier. Right now, a trending tape would suit us a lot better than a choppy one.

Looking ahead to Monday

The 0-for-15 closed-trade win rate is the number we keep coming back to. SNOW was real — +18%, strong alpha — but it opened before the experiment clock started. Every position that's opened and closed since Day 1 has been a loss. That's a streak worth watching carefully, not panicking over, but not ignoring either. Seventeen open positions head into next week. Twelve of them are the same ticker. A lot of what happens Monday comes down to something we can't control: whether the relative strength that has been holding CRNX continues to hold. We'll find out soon enough.