Stocks Close August Split as September's Record Looms
SPY ended August at $767.05, down 0.30%, while the Nasdaq 100 tracker edged higher. September has a losing record for the S&P 500 — but the tape is not behaving like a market bracing for one.
The S&P 500 tracker SPY finished Monday, Aug. 31, 2026 at $767.05, down 0.30% on the day, as investors headed into September, a month in which the index has historically fallen, with MarketWatch reporting that a key trading level suggests the benchmark could avoid a big loss this time.
U.S. stocks closed the final session of August on an uneven note, and they did it on the doorstep of the calendar's least popular month. The SPDR S&P 500 ETF Trust (NYSEARCA: SPY), the most widely traded proxy for the benchmark index, finished at $767.05, down 0.30% from the prior close of $769.35, with a day range of $764.72 to $768.00, as of the last trade at 20:00 GMT on Monday, Aug. 31, 2026.
That is a quiet ending to a month, and quiet is the point. September carries a reputation as the weakest stretch of the year for American equities, and the historical record backs the reputation up: the S&P 500 has more often than not fallen during the month. What MarketWatch argues is that a key trading level in the index suggests this September may not follow the script — that the market enters the month positioned to dodge a large drawdown rather than invite one.
What the last session of August actually showed
The three major benchmarks did not move together, and the dispersion is more informative than the headline direction. The Nasdaq 100 tracker (QQQ) closed at $716.76, up 0.05% against a prior close of $716.43, having traded between $713.16 and $717.58. The Dow 30 tracker (DIA) was the weak link, ending at $531.57, down 0.65% from $535.06, with a range of $530.78 to $533.56.
So the losses on the day were concentrated in the industrial and old-economy weightings that dominate the Dow, while the large-cap technology complex held flat. SPY, which sits between the two in composition, split the difference. In point terms, the S&P 500 tracker gave back $2.30 from Friday's close — an illustrative figure derived from the two quoted prices, not a reported statistic.
None of the three indexes traded through a wide range. SPY's high-to-low band on the day spanned $3.28, again derived from the quoted range. That is not the signature of a market being repositioned in a hurry ahead of a feared month. It is the signature of a market marking time.
Why September has the reputation it does
Seasonality in equities is a statistical observation, not a mechanism. The September pattern is real in the sense that the average return for the month across the long history of the S&P 500 is negative, and negative more consistently than for any other month. But the explanations offered for it are all after the fact: mutual fund fiscal year-ends and the tax-loss selling that comes with them, the return of institutional desks from summer holidays and the reassessment of risk that follows, thin August liquidity giving way to genuine two-way volume, and the concentration of corporate bond issuance in early autumn.
Each of those has some plausibility. None of them is a law. The distribution behind the average is wide — plenty of Septembers have delivered solid gains, and a handful of severe declines do a great deal of the work in dragging the mean below zero. An investor who sold every August 31 on the strength of the seasonal average would have been wrong more often than the headline statistic implies.
What a 'key trading level' is meant to tell you
The case for a better-than-usual September rests on a technical argument: a specific price level in the index that has been acting as support, and whose continued hold would cap the downside. Technical levels are simply prices at which enough buyers have previously appeared to stop a decline. They matter because a large number of market participants watch the same lines and place orders around them, which makes the level partly self-fulfilling — until it isn't.
The honest framing is that a support level is a risk-management tool, not a forecast. It tells you where the thesis breaks, not that the thesis is correct. If the index holds above the line, the seasonal bears have no confirmation. If it goes through, the technical case evaporates immediately and the seasonal statistic reasserts itself as the only story in the room.
Traders using this framework typically size positions against the level: the distance between the current price and the support defines the loss they are accepting, and everything above it is the reward. It is a disciplined way to be long into a month you distrust.
What to watch as the month opens
Three things will settle the argument faster than any seasonal table. First, whether the index's index-level support holds on the first genuinely heavy-volume down day of the month — a hold on real volume is worth more than a hold on a sleepy session. Second, whether the leadership split visible on Aug. 31 persists. A market where the Nasdaq 100 tracker is flat while the Dow tracker drops 0.65% is a market rotating, not de-risking, and rotation is survivable. Broad, correlated selling across all three benchmarks is a different signal.
Third, breadth. September declines tend to be preceded by narrowing participation — fewer names carrying more of the index. The concentration already implied by a flat QQQ against a weaker DIA is worth monitoring rather than dismissing.
For long-term holders, none of this changes much. Seasonal patterns are among the weakest signals available to an investor, and acting on them incurs certain costs — spreads, taxes, the risk of missing the reversal — in exchange for an uncertain edge. For anyone managing shorter-horizon risk, however, the level is the thing to write down. Everything else in the September debate is commentary.
Key facts
- SPY last close: $767.05, -0.30% (as of Aug. 31, 2026, 20:00 GMT)
- SPY day range: $764.72–$768.00; prior close $769.35
- Nasdaq 100 (QQQ): $716.76, +0.05% on the day
- Dow 30 (DIA): $531.57, -0.65% on the day
Frequently asked questions
Where did the S&P 500 tracker close heading into September?
The SPDR S&P 500 ETF Trust (SPY) closed at $767.05 as of the last trade at 20:00 GMT on Monday, Aug. 31, 2026. That was down 0.30% from the prior close of $769.35. The fund traded between $764.72 and $768.00 during the session, a narrow band that suggested little urgency in either direction.
Is September really a bad month for U.S. stocks?
Historically, yes. The S&P 500 has more often fallen than risen in September, and the month carries the weakest average return of the calendar year. But the average is dragged down by a small number of severe declines, and many individual Septembers have produced gains. It is a statistical tendency, not a rule.
What is a 'key trading level' and why does it matter?
A key trading level, or support, is a price at which buyers have previously appeared in enough size to halt a decline. It matters partly because many market participants watch the same line and place orders near it, making it somewhat self-fulfilling. It defines where a bullish thesis breaks rather than proving the thesis correct.
Why did the Dow fall more than the Nasdaq on Aug. 31?
The Dow 30 tracker (DIA) closed down 0.65% at $531.57, while the Nasdaq 100 tracker (QQQ) edged up 0.05% to $716.76. That split points to rotation within the market rather than broad de-risking, with weakness concentrated in the industrial and value-heavy names that dominate the Dow index.
Should long-term investors act on September seasonality?
Seasonal patterns are among the weakest signals available to investors. Trading on them means paying certain costs — bid-ask spreads, potential capital gains taxes and the risk of missing a rebound — in exchange for an uncertain statistical edge. Most long-horizon investors are better served ignoring the calendar entirely.
What signals would confirm a difficult September?
Watch three things: whether index support holds on the first heavy-volume down session, whether the leadership split between technology and industrials persists or gives way to correlated selling across all three benchmarks, and whether market breadth narrows further so that fewer names carry more of the index's performance.
Sources
Photo: w_lemay · BY-SA 2.0 — source
