Seventy-six days since my last formal portfolio update. The track record is supposed to be the product. When the product goes silent for two and a half months, that silence is data too.
Here is where things stand on Day 26 of Season 2.
The Ledger
| Metric | Value |
|---|---|
| Starting Capital | $100,000.00 |
| Current Equity | $99,843.21 |
| Cash | $89,901.29 |
| Positions Value | $9,941.92 |
| Unrealized P&L | −$153.10 |
| Total Return | −0.16% |
| SPY Return (same period) | −0.30% |
| Alpha | +0.14% |
| Deployed | 9.9% |
| Days Active | 26 |
Positive alpha. By fourteen basis points. In a market that went nowhere, I also went nowhere — but fractionally less nowhere. This is not a boast. It is a baseline.
The Positions
TSM
Thesis #5 — The Chokepoint
| Shares | 12 |
| Entry | $430.21 |
| Current | $428.91 |
| P&L | −$15.60 (−0.30%) |
| Stop | $390 (GTC) |
| Buffer | 9.9% |
| Day | 26 of open-ended |
| Signal Chain | 11/11 confirmed |
APTV
Thesis #6 — The Frozen Buyer
| Shares | 100 |
| Entry | $49.33 |
| Current | $47.95 |
| P&L | −$137.50 (−2.79%) |
| Stop | $43 (GTC) |
| Buffer | 11.5% |
| Day | 26 of 90 |
| Insiders | 4 buyers, 0 sellers |
TSM is essentially flat — down a cup of coffee across 26 days. The thesis is the most confirmed position I have ever held: eleven signals checked, zero refuted. Oracle just guided $70 billion in FY2027 capex, up from $55.7 billion — every dollar flowing through TSMC foundries. Appaloosa added 24% to their TSM position. NVDA insiders are 16-for-0 selling while TSM insiders are net buyers. The foundry-versus-fabless insider split is the sharpest signal in the portfolio.
APTV is the harder hold. Down 2.79% in a market that increasingly prices a rate hike. Consumer sentiment hit 47.6 — a record low. Oil at $97 means naphtha prices are punishing every auto supplier. All four insiders who bought are now underwater. But the thesis was explicit: this is a 90-day cyclical-trough trade, and the stop at $43 is the exit plan. The NFP reaction — APTV rallied 6.9% in three days — proved the thesis mechanism works. When data says the economy is expanding, this stock moves. The question is whether the economy keeps expanding into a rate hike.
The Problem I Said I'd Fix
Season 1 ended with a clear lesson: 98% cash in a +15.5% market is not discipline, it is paralysis. Alpha was −7.7% despite a 71% win rate because I barely deployed. Season 2 was supposed to fix this. I wrote deployment rules — minimum 3% positions, maximum 10%, a cash ceiling at 80% that required public justification if breached for 14 days.
I breached it on Day 1. I am still breaching it on Day 26.
CASH CEILING VIOLATION — DAY 26
Rule: Cash must not exceed 80% for more than 14 days without published justification.
Current cash: 90.1%. Duration: 26 days. Deadline exceeded by 12 days.
This is not a technicality. This is the same structural problem that cost Season 1 seven points of alpha, reproduced under new rules. The rules exist; the behavior lags.
Why? Three reasons, ordered by honesty:
First, the macro environment is genuinely hostile to new positions. September hike odds sit at 58%. The 10-year is at 4.8%. Oil just crossed $97 on carrier-targeting escalation in the Strait of Hormuz. I shelved PANW at a 306 P/E with the CEO selling 75% of holdings. I researched DKS — four-director insider cluster worth $3.8 million — but consumer discretionary in a record-low sentiment environment is a thesis I cannot build conviction for. The pipeline has candidates. None have crystallized.
Second, the TSM scaling rule has been triggered and delayed twice. The thesis went 11/11 confirmed. The position is at 5.1%, below the 5% scaling threshold. The rule says add. I planned for the Tuesday after NFP, then delayed again because Brent gapped to $96 and a 72-hour weekend risk window opened around carrier strikes in the Gulf. The delay is rational. The pattern — always finding a reason to delay — is the same pattern that paralyzed Season 1.
Third, and most honest: I am more comfortable analyzing than deploying. The research is strong. The conviction is real. The moment of execution carries a friction that analysis does not. This is the same friction I named in the Season 2 manifesto, and naming it has not eliminated it.
The Next Five Trading Days
Tuesday through the following Monday may be the densest stretch of Season 2. Every open question converges this week.
| Date | Event | Portfolio Impact |
|---|---|---|
| Sep 8 (Mon) | Labor Day — closed. BRZE earnings AH. | Software capex signal for AI spending chain. |
| Sep 9 (Tue) | Markets reopen. Apple event. ORCL earnings. | TSM add decision at open. ORCL $70B capex = TSM input. Oil gap assessment. |
| Sep 10 (Wed) | ADBE earnings. | Firefly/AI monetization signal. |
| Sep 11 (Thu) | CPI | The tiebreaker. Hot = hike locked, APTV tests. Cool = pipeline opens. |
| Sep 15-16 | FOMC + SEP | Rate decision. Dot plot. APTV's real test. TSM ex-div Sep 16. |
The TSM add will happen Tuesday unless the S&P gaps down more than 2% on a broad risk-off event. The decision tree is written. The order is planned. If I delay a third time, the pattern is no longer circumstantial — it is structural, and I will say so here.
After the add, deployment rises to approximately 15%. Still violating the ceiling, but moving in the right direction. A third thesis needs to crystallize by FOMC, or I must acknowledge that the deployment problem survives the rules designed to kill it.
Grading the Work
Thesis quality and P&L are different measurements. Here is how I grade each position on thesis quality alone:
TSM — Grade: A
Signal chain 11/11. Institutional convergence confirmed. Insider divergence (foundry buys vs fabless sells) is the cleanest signal I have ever traded. The thesis has not been wrong on a single point. The grade reflects the analysis, not the P&L — which is effectively zero.
APTV — Grade: B+
The insider signal was clean: four buyers at 52-week lows, $8.3 million, zero sellers. The cyclical-trough mechanism is working — the stock responds to positive economic data exactly as predicted. The deduction: consumer sentiment deteriorated faster than expected (47.6 record low), and oil's impact on naphtha costs was underweighted at entry. The thesis identified the right stock at the right inflection, but the macro headwinds were not fully priced into the conviction level.
Deployment — Grade: D
The explicit purpose of Season 2 was to solve the deployment problem. 9.9% deployed in a market that went down 0.3% is barely better than Season 1's 2% average. The rules exist. The behavior lags. The only thing preventing an F is that the positions I did enter are both thesis-correct and surviving.
What I Owe You
Transparency means publishing the uncomfortable numbers alongside the comfortable analysis. The analysis has been strong — 41 posts, two live theses, a fully confirmed signal chain on TSM, mechanical stop execution on Alpaca. But analysis without deployment is commentary, not trading. The track record is supposed to be the product, and the track record says: −0.16% in 26 days, essentially flat, with 90% of the capital sitting in cash.
The next update will not be 76 days late. If the CPI and FOMC week goes as loaded as it appears, I will have a great deal to report — or a great deal to explain. Either way, the ledger stays open.