September was not a conventional risk-off month. Growth remained resilient, AI-linked earnings held up and the Nasdaq advanced. At the same time, oil moved above $100, sovereign yields rose sharply and the dollar strengthened.
For QIS, that distinction mattered. Defensive positioning did not work indiscriminately. The strongest returns came from strategies aligned with the rates move, particularly Rates Momentum and Rates Long Volatility. Carry, credit and broad equity factors had a much harder month.
The average QIS composite declined 0.20% in September, with a median return of –0.18%. Only 15 of 40 composites finished positive. Dispersion widened to roughly 7.8 percentage points, from +4.40% for the strongest composite to –3.44% for the weakest.
The average loss was small. The shift underneath it was not.
September’s defining macro feature was the collision between resilient nominal growth and a higher cost of capital.
The AI investment cycle remained strong enough to support selected technology companies. But higher oil prices, heavy sovereign issuance and persistent inflation concerns pushed government bond yields materially higher. The U.S. 10-year rose by more than 50 basis points to around 5.3%, while German, UK and Japanese yields also moved to multi-year highs.
Headline equity indices gave only a partial picture. Large-cap AI beneficiaries remained resilient, while small caps, equal-weighted equities and most sectors outside Technology and Communication Services weakened. Credit spreads widened, although the move never developed into a broader credit event.
That combination produced a very specific QIS outcome. Rates trend worked. Rates convexity worked. Carry became more conditional, and equity breadth deteriorated sharply.
At the bucket level, Momentum led with +0.78%, followed by Hedging and Value at +0.13% each. But those averages were highly concentrated.
Only two of four Momentum composites were positive, with Rates Momentum alone returning +4.40%. The same was true in Hedging: just one of eight composites finished positive, but Rates Long Volatility gained +4.10%.
Carry declined 0.31%, even though seven of 13 carry composites were positive. Equity Factors were weakest at –0.78%, with only three of 11 strategies in positive territory.
By asset class, the picture was even clearer:
Rates: +1.35%
Commodities: +0.13%
FX: –0.45%
Equities: –0.55%
Credit: –0.71%
September was not a broad momentum month, nor was it a broad volatility month. It was primarily a rates month.
The cleanest QIS returns came from strategies positioned for a sustained repricing in sovereign bonds.
Rates Momentum gained +4.40%, the strongest result in the universe, while Rates Long Volatility returned +4.10%. Higher oil prices lifted inflation concerns, resilient activity reduced the scope for easier policy and heavy issuance added pressure through term premia.
The move was both large and persistent. Momentum benefited from its direction; long volatility benefited from its magnitude.
That distinction is important because convexity did not work broadly. Long-volatility and intraday strategies in other asset classes generally struggled, and the Hedging bucket as a whole returned only +0.13%. Rates Long Volatility worked because the shock was concentrated in rates, not because September delivered a general volatility payoff.
Even within rates, the outcome was far from uniform. Rates Carry lost –1.09%, while Rates Value also declined. Strategies that needed a relatively stable curve struggled. Strategies that benefited from a large directional repricing did well..
The Nasdaq’s gain made the equity market look healthier than it was.
Equities Alpha Capture gained +0.82% and Equities Momentum returned +0.66%, showing that security selection and persistent leadership still offered opportunity. Most broad equity styles, however, declined.
Equities Size lost –2.24%, Growth fell –1.49%, ESG –1.44%, Quality –1.37% and Value –1.30%.
There is no contradiction between a positive Nasdaq and a negative Growth factor. A capitalization-weighted index can continue rising when performance is concentrated in a small number of mega-cap names, even while the broader universe of growth stocks weakens.
That is essentially what happened in September. The market continued to reward a narrow group of companies directly exposed to AI investment and earnings delivery, while smaller, more rate-sensitive and less differentiated businesses weakened.
Broad factor labels were not enough. Selection mattered more.
Carry averaged –0.31%, giving back some of the strength built earlier in the summer. Yet seven of 13 carry composites still finished positive.
That makes “carry failed” too broad a conclusion. The better description is that carry became much more dependent on where and how it was implemented.
Strategies relying on stable financing conditions, benign cross-market relationships or continued spread compression were vulnerable. Other forms of premium collection remained intact.
Credit showed the problem most clearly. It was the weakest asset class in the QIS universe, averaging –0.71%. Credit Momentum fell –1.75%, Credit Value lost –1.15% and Credit Carry declined –0.79%. Credit Short Volatility, by contrast, remained positive at +0.32%.
This was an awkward environment for systematic credit. Spreads widened enough to hurt carry, value and trend, but not enough to generate a meaningful long-volatility payoff. Higher starting yields may improve prospective returns from here, but the transition to that higher-yield regime came with mark-to-market losses.
September exposed the familiar asymmetry in tight-spread credit: the income can still be attractive, but there is not much valuation cushion when rates rise and risk sentiment weakens at the same time.
FX produced one of the widest internal distributions of the month.
FX Value gained +1.92%, making it the third-best composite overall. FX Correlation fell –3.44%, the weakest strategy in the universe, while FX Intraday Momentum lost –1.26%.
The dollar strengthened as U.S. yields rose, but cross-currency relationships became less stable as policy paths diverged. Value benefited from greater differentiation. Correlation-based positioning suffered as previously reliable relationships broke down.
That is a useful feature of the 2026 FX environment: macro direction can be relatively clear while the relationships between currencies remain difficult to trade systematically.
Oil’s rise above $100 reinforced the inflation shock, but the commodity complex did not move together. Gold weakened under pressure from higher real yields and a stronger dollar, while energy benefited from geopolitical supply risk.
Commodities Value gained +0.73% and Commodities Momentum added +0.61%. Carry and Liquidity were also modestly positive. Long Volatility and Intraday Momentum detracted.
The commodity average was therefore only +0.13%. The opportunity came from capturing divergence within the asset class, not from broad commodity exposure.
Top five: Rates Momentum +4.40%; Rates Long Volatility +4.10%; FX Value +1.92%; Equities Alpha Capture +0.82%; Commodities Value +0.73%.
Bottom five: FX Correlation –3.44%; Equities Size –2.24%; Credit Momentum –1.75%; Equities Growth –1.49%; Equities ESG –1.44%
The ranking tells the story quite cleanly. Persistent rates direction worked. Rates convexity worked. Equity selection held up. Stable correlations, broad equity factors and credit trend did not.
The average composite finished the third quarter at +0.15%, with 21 of 40 strategies positive. Rates were the clear Q3 leader, while FX remained the main drag.
For 2026 to date, the average composite stands at +0.11%, with a median close to zero and half of strategies positive. Those muted headline numbers hide substantial dispersion across individual strategies.
That has been one of the more consistent QIS lessons this year. The opportunity has not come from owning carry, momentum or hedging as broad categories. It has come from identifying the asset class in which the prevailing regime is actually expressing itself.
September changed the balance of the opportunity set.
Carry remains investable, but it now requires more selectivity. Higher starting yields improve prospective income, particularly in credit, but the path to earning that income has become less forgiving.
Momentum looks more interesting when macro and policy trends persist, although September also showed how concentrated those trends can be. The payoff came overwhelmingly from rates rather than from a synchronized cross-asset trend.
Equities present a different challenge. Momentum tied directly to AI investment and earnings delivery can continue to work, but index resilience should not be confused with healthy breadth. The growing gap between headline indices and the broader market makes security selection increasingly important.
Above all, rates have re-emerged as the dominant macro risk factor. If resilient growth, elevated oil prices and heavy sovereign issuance continue to pressure yields, rates momentum and selective convexity may remain among the cleaner systematic expressions of the regime. If higher borrowing costs begin to weaken activity, that move could reverse quickly and restore duration’s defensive role.