September’s Weak Record Returns As Bond Yields Pressure Stocks

Historically, the month of September represents the least favorable period for U.S. stocks. The S&P 500 undergoes an average correction of 0.7% and closes higher in only 44% of cases since 1950. This seasonality is associated with high bond yields, U.S. public debt exceeding 40 trillion dollars, and questions about the profitability of investments in artificial intelligence this year. However, history does not directly predict a decline.

In Brief

• September is historically the least favorable month for the S&P 500, with an average return of -0.7% since 1950.
• The technical context provides a counterpoint, with the index starting September well above its 200-day moving average.
• The increase in bond yields heightens pressure on valuations, with particularly high long-term U.S. rates.
• U.S. federal debt exceeds 40 trillion dollars, fueling questions about financing costs.
• Massive spending on AI will be monitored, particularly their ability to generate anticipated revenues.

September Shows an Average Negative Return Since 1950

The significant underperformance of the main stock indices during the back-to-school month is called the “September Effect”. This scenario mainly concerns the S&P 500, but also the Nasdaq, Dow Jones, and Russell 2000.

Various interpretations have been proposed. After the summer decline in volumes, major managers return to the market and reassess their positions. Several funds rebalance their portfolios, lock in gains, or reduce their exposure before the end of their fiscal year.

Some data allow measuring this seasonal weakness :

• The S&P 500 loses an average of 0.7% in September since 1950 ;
• The index finishes the month higher in only 44% of cases ;
• September is the only month with an average negative return over this period ;
• Pullbacks also regularly affect the Nasdaq and Dow Jones.

These averages do not constitute a market norm. Data created and constructed over many decades combine diverse economic, monetary, and geopolitical periods. Therefore, they do not confirm that September 2026 will end in the red.

A counterpoint is even provided by the technical context. The S&P 500 rose 2.6% in August and starts September nearly 8.3% above its 200-day moving average. Since 1950, if the index starts the month above this threshold, its average return is 0.2% according to the reported data.

Bond Yields Increase Pressure on Valuations

The main difference this year is related to the bond market. Indeed, the yield on the U.S. 10-year Treasury note is around 4.7%, while that of 30-year bonds exceeds 5%.

The U.S. Treasury statistics positioned these yields near 4.73% and 5.22%. These are securities that reveal low credit risk, although they remain exposed to rate fluctuations and inflation.

If bonds provide higher compensation, investors usually demand additional returns to hold equities. This increase mechanically reduces the present value of expected gains over several years. Growth companies are thus more sensitive to a further rate hike.

John Plassard, partner at Cité Gestion, emphasizes this difference in an on-chain analysis:

September will be less a calendar test than a confidence test.

Now exceeding 40,000 billion dollars, U.S. federal debt also heightens questions. Government borrowing needs keep long-term yields at a high level. This scenario increases financing costs for the state, corporations, and households.

Spending on AI Will Be Particularly Monitored

Technology companies are injecting substantial amounts into chips, data centers, and power grids useful for artificial intelligence. These expenses still need to generate the revenues necessary to justify some valuations.

A high cost of capital further disadvantages companies whose profits are expected in the long term. A probable correction would thus take the form of a sector rotation rather than a global collapse. Investors would favor profitable, low-debt companies capable of automatically generating cash flow.

Market direction will depend on various deadlines. For instance, the European Central Bank will announce a monetary decision on September 10. On September 15 and 16, the Federal Reserve will meet. As for the Bank of England, it will announce its decision on September 17. U.S. employment and inflation figures will also weigh on rate expectations.

September 2026 thus carries several risks, but its unfavorable history is not enough to announce a decline. The rise in bond yields and central bank decisions will provide stronger indications than seasonality alone.