Positioning and Outlook - 2026-08-22

Published 2026-08-22

A technology-led rally powered by strong artificial-intelligence earnings and expectations that the Federal Reserve will hold rates steady sits on top of strained corporate credit, with survival scores weak, dozens of issuers in danger territory, and a heavy buyback bid providing the main support beneath the surface.

Recent Events

Equities pushed higher through the middle of the period as strong AI-driven earnings lifted the technology sector and a benign inflation reading reinforced bets that the Federal Reserve would keep rates on hold, with the Fed's own communications striking a moderate tone and Cook delivering remarks on the outlook for the U.S. and Alaskan economies. The credit tape told a more strained story beneath the rally: corporate credit survival scores sat in the low-to-mid twenties out of 100 with roughly 50 issuers in danger territory, high-yield spreads went sideways on the week, and the main counterweight remained a heavy buyback bid, more than 12,000 active programs representing some $4.5 trillion in trailing-twelve-month repurchases. Single names added texture, with On Holding tempering its sales outlook after missing quarterly revenue targets, the FDA issuing Class I device recalls for Boston Scientific's neuromodulation unit and AVID Medical, and Veralto reporting results.

Upcoming Events

The heavy end of the forward calendar is a tightly packed run of central-bank decisions in mid-September: the ECB's Governing Council meets on September 10, the FOMC decides on September 16, the Bank of England follows a day later, and the Bank of Japan closes the week on September 18, with the PBOC's loan prime rate fixing landing on September 21. Four major policy boards moving within little more than a week compresses the window in which rate expectations can reprice, leaving the market trading from decision to decision rather than off secondary data in between. The nearer stretch around August 21 carries a different kind of weight, a monthly options expiration that tends to swell volumes and positioning flows, an FDA decision date for Deramiocel on August 22, and a cluster of single-name trial readouts including CagriSema, Pembrolizumab and Danicopan. Beneath it all sits a broad earnings undercurrent, with 191 reports scheduled and PVH among the first out of the gate, keeping stock-level dispersion alive while the macro calendar builds toward its September peak.

Macro Projections

The next five weeks pack in nearly every catalyst that matters, PCE and GDP on 8/26, payrolls on 9/4 (prediction markets give only about 25% odds of a print above 80,000 jobs), the ECB on 9/10 and CPI on 9/11, then the central-bank stack of FOMC 9/16, BoE 9/17 and BoJ 9/18, with the BoJ carrying outsized weight because the carry model still tags dollar-yen near 160 and Japanese government bond yields have breached 2.85%. CPI is the fulcrum of the near-term leg: the crowd prices August inflation above 3.2% year over year at 87%, and the 10-year breakeven, the market's implied inflation rate, is accelerating higher at 2.34%, a 71st-percentile reading, so confirmation keeps the Fed on hold into year-end (the fed-reaction model puts hold propensity at 84.6% against 15.9% for intervention), while a soft miss would squeeze a front end where leveraged funds are short the two-year at a 52-week extreme of +3.43, a stretch that would unwind violently on a dovish surprise. Away from rates, the regime reads calm on every top-tier gauge: high-yield spreads (the extra yield junk bonds pay over Treasuries) have sat in a 2.67–2.78% band for two weeks, bond-market volatility via the MOVE index is stable at 73.4 bp, the VIX is 16.0, currency volatility runs below average, and all ten correlation pairs sit in calm with no break flags, and even the carry model's disorderly-unwind probability, still the largest single branch, is falling at 0.438 from 0.503. Rates are the exception to that calm: the 10-year at 4.738% is accelerating upward and the 30-year at 5.276% sits in the 94.6th percentile of its history, last higher on 8/18, an extreme the two-year (4.19%, 56th percentile, stable) pointedly does not share. That long-end stress is partly a policy fight, Treasury buybacks and the reported Bessent yield-suppression push are holding yields down at the dollar's expense, while WSJ reporting of undisclosed Warsh–Trump contacts adds a Fed-independence risk premium the rates market looks only about half to be pricing. The cyclical months-to-quarters view forks from here: the outlook assigns roughly 25% to the calm-credit regime breaking by Q4, most likely via an exogenous trigger, with the tells being high-yield spreads sustained above 3.25% and MOVE above 100 (historically, such volatility spikes force deleveraging with a 0.74 hit rate across 54 episodes). The cleanest live branch is oil: if the Hormuz near-closure, three days old, escalates to actual transit disruption by mid-September, crude re-prices violently given an 0.86 historical hit rate for supply shocks feeding through to CPI, whereas if it fades, the +4.4 million-barrel build and the crowd's roughly $79 WTI pricing hold; as it stands the market prices near-zero supply premium against a documented closure edge, with the geopolitical-risk gauge at 144.5, down from 225 but still elevated. On the structural multi-year leg the reasoning shifts from catalysts to valuation and carries the lowest confidence by construction: long-dated Treasuries read as a value trap despite the 30-year's extremity because the trajectory remains adverse, while earnings yields of 5.98% in emerging markets and 5.38% in developed ex-US hold a 140–200 bp advantage over the US at 3.98% (a cyclically adjusted multiple of 41.8), and sentiment at a composite 70 in greed territory confirms rather than drives the read, with the whole chain explicitly falsifiable via spreads above 3.25% for two consecutive weekly reads, MOVE above 100, or WTI above $95 breaking the calm-higher-for-longer frame, and a 30-year back below 4.90% with breakevens rolling over breaking the rates leg.

Positioning

With high-yield credit spreads near 2.75% and the backdrop calm, the model portfolio sizes positions in direct proportion to conviction, conviction being the screening kernel's composite score, a modelled blend of expected upside drift against drawdown probability, so the highest-scoring names carry the largest weights. Overall risk appetite is then calibrated by a small set of structural mechanisms, each treated as a weighable input rather than a prediction: merger-arb completion has historically been associated with an upside move in the target's spread 89% of the time across 47 prior firings, while the oil-supply-to-headline-CPI and commodity-production-loss channels show base rates of 86% and 80% on thinner histories of seven and five firings, misses counted in each case. Every mechanism is discounted for what the market already prices, roughly 60% for the first two and 40% for the third, so only the unpriced portion feeds into sizing. The holdings and composite scores that follow show how that arithmetic settles out across the book.

Model Portfolio

The model portfolio is effectively an all-cash book this week: half sits in a dedicated money-market sleeve and the remainder is unallocated, leaving nothing committed to risk assets. That money-market stake is being treated as dry powder, earmarked for staged deployment into the cascade as the credit and volatility picture resolves, rather than as a permanent home. The latest weekly rotation review produced no changes, so the standing positions carry over untouched from the prior allocation, with the book still waiting on conditions to shift before putting that reserve to work.

Track Record

The record so far is modest and reported without varnish: across roughly 69 days of scored ideas, the model's one-day calls have landed just over half the time with essentially flat average alpha, while results fade at longer horizons, under half of one-week calls beat their benchmark, and only about a third of four-week calls did, with average alpha at that window running around negative 1.6%. Every evaluated idea is counted here, not a flattering subset, and while the sample of several hundred calls per horizon is no longer tiny, it still covers barely two months of market conditions. The three-month horizon has yet to mature, with first readings expected around mid-September. The discipline is the point: each call is written down before the outcome is known and scored when the horizon arrives, whatever the result.

Macro forecasts: 78 graded, 423 open — each call is scored against what actually happened when its horizon arrived.