Citadel: US stocks may face a "tactical downtrend window" in September; it is recommended to reduce positions on rallies and buy protective equipment on dips.

Since its March lows, the S&P 500 has risen by about 22%, adding about $12 trillion in market capitalization, but the multiple tailwinds that supported this rally are fading away.

In his latest report, Scott Rubner, chief equity and derivatives strategist at Citadel Securities, stated that his long-term constructive outlook for U.S. stocks remains unchanged, but the risk-reward ratio has shifted recently. Rubner pointed out that as we enter September, the positive effects of earnings season have largely been priced in, retail and corporate buyback demand has historically weakened significantly in this month, the space for systemic fund rebuilding has narrowed considerably, and the macroeconomic calendar is converging, with seasonal patterns indicating that September is the weakest month of the year.

Rubner explicitly suggested that this was the first time since the July correction that he favored "selling on rallies and buying on dips for protection" rather than chasing the rally. He characterized September as a "tactical downtrend window" rather than the start of a broader bear market reversal, and predicted that after this window, the market might present more constructive positioning opportunities around mid-October.

With all the positive earnings reports priced in, the biggest catalyst has moved out of sight.

This earnings season has been exceptionally strong, but its boost to the market has largely been realized. According to Rubner data, 93% of the companies weighted in the S&P 500 have already released their earnings reports, with 88% of them exceeding earnings per share expectations, a median difference of 7%; while for those that missed expectations, the median difference was only 3%.

In the second quarter, the S&P 500's earnings per share grew by approximately 33% year-over-year, the strongest growth rate outside of the post-recession recovery period, and the valuation correction path was also the steepest since 2000.

With Nvidia's earnings now finalized, the most important individual stock catalysts have all passed. Historically, US stocks tend to rise in the first month of earnings season, but this momentum dissipates as the earnings calendar enters a lull. Rubner points out that the corporate calendar will become significantly lighter thereafter, removing the clearest source of positive surprises for the summer.

Both retail sales and buybacks have weakened, putting the demand structure to the test.

The two main sources of demand that supported the August rally—retail investors and corporate buybacks—are both facing seasonal contraction pressures.

In the retail sector, according to data from Citadel Securities, the average daily net nominal purchases in August were at the 65th percentile over the past year, about 10% higher than the average, indicating continued net buying. However, the average daily total nominal purchases were only at the 35th percentile, 4% lower than the average, indicating low overall participation. More importantly, September has historically been the weakest month for retail demand, with both the net nominal purchase ratio and directional skew at their lowest points of the year. Since 2019, the average net retail purchases on days when the S&P 500 index falls were only half the monthly average in September, the lowest level of any month throughout the year.

Regarding share buybacks, the total amount of buyback authorizations granted by Russell 3000 constituent stocks since the beginning of the year has exceeded $1.1 trillion, with 67% concentrated in sectors outside of technology, providing significant support for August. However, as more companies enter a quiet period ahead of their Q3 earnings reports, this source will continue to shrink. Rubner points out that the quiet window will accelerate significantly around September 12, and the largest and most stable source of structural demand will gradually diminish as the months progress.

With protection costs at a low level, hedging offers excellent value for money.

The current market is pricing downside protection at extremely low levels, a characteristic that Rubner sees as both a risk signal and a tactical opportunity to establish a hedge.

The 1-month 25-delta bearish/bullish skew of the S&P 500 has fallen to its flattest level in the past year, at the 1st percentile. At the end of August, the 1-month 25-delta downside protection cost fell to its lowest level since December 2024, while the VIX closed at 14.4, the second lowest closing level since December 2025. Implied volatility is particularly low in interest rate-sensitive sectors such as small-cap stocks, financials, regional banks, and consumer retail.

The compression of individual stock volatility is also significant. From the end of March to the expiry date in July, the average 1-month at-the-money implied volatility of the top 15 components of the Philadelphia Semiconductor Index rose from 53.0 to 77.2 in 74 trading days; then, it took only 20 trading days to give back all the gains, and after 30 trading days, it had fallen to 46.0, a decrease of 31.2 volatility points from the peak, a drop of 40%, and even lower than the starting point before this round of rise. At the same time, the VVIX (volatility of volatility) is currently at the 1st percentile since the beginning of 2025.

Rubner concluded that the market is about to enter a period of intensive macro events, but investors are buying protection at extremely low premiums, and this divergence itself constitutes an asymmetrical positioning window.

Systemic funding has been restructured, and competition now comes from bonds rather than stocks.

The systemic funding rebuilding space that emerged after the July correction has now been largely exhausted. CTA, volatility control, and risk parity strategies have all rebuilt their exposure from the July lows, with the increase mainly concentrated in the S&P 500 and Russell 2000, while Nasdaq exposure remains largely unchanged. Rubner stated that current positions are not crowded, but the market no longer possesses the systemic buying reserves that were not deployed after the July correction.

Meanwhile, a significant technical event occurred at the end of the quarter. The coverage ratio of the top 100 U.S. pension plans was approximately 112%, the highest level since 2001. This high coverage ratio continues to incentivize plans to de-glide and implement portfolio immunization, creating mechanical selling pressure on stocks and buying pressure on fixed income at the end of the quarter. Rubner points out that given the significant rebuilding of systemic equity exposure from its July lows, while duration positions remain relatively insufficient, clearer systemic positioning opportunities may have shifted towards bonds rather than stocks.

Seasonal demand for equity mutual funds is also relatively weak – since 1984, September has been the month with the lowest subscription rate for equity mutual funds throughout the year, with a median subscription volume of approximately 1.79% of assets under management.

Options expiration and the macroeconomic calendar bring double pressures.

The massive expiration of quarterly options in September could trigger another round of technical market rebalancing. Currently, approximately $9.6 trillion in US option exposure will expire by September 18th, representing about 35% of all US option exposure; of this, $6.2 trillion will expire on September 18th alone, accounting for about 23%. At the current pace, September is expected to surpass the record of $7.7 trillion with triple expiration dates set in June. As option positions expire or roll over, the long-term gamma-level hedging support previously provided by market makers may dissipate, removing another layer of cushioning below the stock market.

The macroeconomic calendar has also shifted significantly. Following the Jackson Hole meeting, a series of data releases are scheduled: non-farm payrolls on September 4, PPI on September 10, CPI on September 11, and the Fed's interest rate decision on September 16. Unlike the earnings season, which is full of positive surprises, macroeconomic catalysts offer two-way risks, and the upside potential is far less clear than before.

Historically, since 1928, September has been the only month in which the S&P 500 has a higher probability of decline than rise—a 55% decline rate, an average monthly return of -1.1%, and an average maximum decline of -4.7%, with the weakness typically concentrated in the second half of the month. Historically, September has been even weaker in midterm election years, with an average return of -1.5% and an average maximum decline of -6.2%. Rubner states that this trajectory aligns closely with his predictions: an initial tactical weakness followed by a more constructive positioning around mid-October, but this does not necessarily indicate a reversal of the broader stock market trend.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.