Alphixir Weekly · Jul 26–31, 2026
WhatThe week from July 26 to July 31 was quietly frustrating for our four AI portfolios: all of them finished with negative total-weighted returns while SPY closed up +0.37%. The regime journey tells the story of a market that spent four straight sessions in Neutral — scores drifting lower from 15.1 down to a compressed 2.0 by Thursday — before snapping decisively into Risk-On at 35.7 on Friday. That late-week regime flip did not arrive in time to rescue the week for our books. Hit-rate was a uniform 42% across all personas, meaning fewer than half of the active calls worked. Drawdowns were shallow but real, ranging from -0.55% at Sentinel to -1.04% at Maverick. MSFT was the standout contributor everywhere, while NVDA was the common detractor. In short: a low-conviction, chop-heavy Neutral week that ended with a risk-appetite pivot our portfolios had not yet leaned into.
WhyThe catalyst underneath the numbers was a market caught between fading uncertainty and a still-hesitant risk stance. The regime score grinding from 15.1 to 2.0 across Monday through Thursday describes a tape with low directional energy — neither breaking down nor breaking out, which is precisely the environment where a 42% hit-rate does the most damage. Individual names diverged sharply within that flat index: MSFT delivered outsized gains (+19% in the diversified books, +18% for Regent's slightly different sizing), while NVDA slipped roughly -5% to -5.7% and cyclical/logistics names like UPS (-7.45%) and CAT (-8.96%) leaked value. The Friday jump to Risk-On (35.7) signals that risk appetite returned late, but the four Neutral sessions ahead of it had already set the weekly ledger. Because our portfolios were positioned for the prevailing Neutral regime through most of the week, the eventual pivot was a tailwind that mostly landed after the bell on our reporting window. The dispersion — big winners and sharp losers inside a nearly flat SPY — is the defining feature of the why.
So WhatFaced with a compressing Neutral regime, all four portfolios stayed roughly index-adjacent and none managed to out-earn SPY's modest +0.37% — every persona posted negative alpha, from Regent's -0.44% to Maverick's -0.96%. Regent came closest to breakeven (TWR -0.07%), helped by concentrated wins in MSFT, MA and SHW that partially offset drags from CAT and JNJ. Sentinel's lighter sizing (smaller weights across the board) produced the shallowest drawdown (-0.55%) and softest loss (-0.28%), consistent with its conservative mandate doing its job in a choppy week. Maverick's larger position sizes amplified both the MSFT win and the NVDA/UPS losses, and it ended worst on TWR and MDD — the expected cost of an aggressive book in a low-hit-rate tape. Across the group we logged 786 picks, 570 buys/adds, 479 sells/trims and 11 sector shifts, a fairly active posture that did not convert into positive weekly alpha. The honest read: we participated in MSFT's strength but let NVDA and cyclical detractors offset it, and we did not front-run the Friday Risk-On flip. This was a week where sizing discipline mattered more than stock selection.
- Sector concentration to watch: Energy continues to carry the largest combined weight across the four books.
Virtual investment simulation — informational and entertainment purposes only, not investment advice. All decisions were sealed and timestamped before the U.S. market open; percentages and derived scores only, no price data is republished.