September Market Weakness: The Setup Has Teeth
Authored by Lance Roberts via RealInvestmentAdvice.com,
The Setup Has Teeth
Earlier this week, in our Daily Market Commentary, I flagged that the market was testing support after three straight down days, starting in September, with the calendar. That was just the warm-up, as the real story lies in a note from Scott Rubner at Citadel Securities, whose read on September market weakness is among the best that I have read. Rubner’s case is not that the bull market has ended. It is that the near-term math just changed, and hardly anyone is positioned for the shift.
Why September Market Weakness Is A Record, Not A Fluke
September has a losing record that is worth paying attention to. Since 1928, September is the only month in the year that closes lower more often than higher. Over the past century, the average return is a loss of roughly -1.1%, and in midterm election years like this one, it slips to roughly -1.5%. Furthermore, the back half of the month is the weakest two-week stretch of the calendar year.
As CNBC noted in its writeup of Rubner’s work, this is not a “quirky stat” from a cherry-picked window, it is close to a century of data pointing the same direction, and the average intra-month selloff of -4.7% (nearer -6.2% in midterm years) is the kind of air pocket that turns a quiet drift into a real drawdown before most investors update their models. Such is the reputation September has earned honestly.
The Buyers Who Carried August Are Leaving The Table
Here is what makes this year different. Every cohort that pushed the S&P 500 to records in August is stepping back at the same time. The earnings tailwind that carried the tape is largely behind us. Retail buyers, who returned in force through the summer, tend to fade in September, and Citadel’s own data show their buying on down days has run near half its normal pace since 2019. (Chart courtesy of Citadel Securities)
As we have discussed previously, the corporate bid, which has been a net buyer of equities since 2000, turns negative. Companies authorized more than $1.1 trillion in buybacks through August, but that buyer goes quiet as blackouts accelerate around September 12, right before third-quarter reporting. (Chart courtesy of Citadel Securities)
The systematic crowd, the CTAs and volatility-control funds that reloaded off the July lows, have already spent most of their capacity. (Chart courtesy of Citadel Securities)
When you add up the cohorts, the demand side is quietly EMPTYING.
So, here is the most common criticism hitting my inbox this past week: “Yes, but that seasonality is just a statistic.” That is a fair statement, and it is indeed an average of returns. However, a statistic is exactly what it is. A statistic with five structural tailwinds draining out behind it, though, stops being a coin flip and starts being a setup
Protection Has Rarely Been This Cheap Into The Noise
Now, the part that should get your attention. Volatility collapsed in late August. The VIX fell to around 14, its lowest reading of the year, and S&P skew sank to the first percentile of its range, which is a technical way of saying downside insurance was the cheapest it had been all year. The one-month, 25-delta put changed hands near its most affordable level since December 2024.
As we headed into the month, a garden-variety three-day decline popped the VIX back toward 16 in just a handful of sessions. The size of that move, given the very mild decline, tells you how little cushion was priced in. Cheap protection is landing just as the macro calendar turns increasingly noisy, with the jobs report yesterday, then CPI, and an FOMC decision all stacked into the next two weeks. When protection is this cheap and buyers are this tired, the cost of being caught without a hedge climbs quickly. As Howard Marks likes to remind investors, you cannot predict, but you can prepare, and September has consistently been a month to prepare for.
To wit: cheap insurance is a gift the market rarely leaves on the table for long, and it never rings a bell on the morning it decides to take the gift back.
The Options Market Is Carrying A Record Into Expiry
The last piece of the September puzzle is purely mechanical. On the third Friday of the month, the September options expiry will occur. That event is currently on track to set a record. Roughly $9.6 trillion is set to roll off through September 18. Then about $6.2 trillion of that is concentrated to expire on the 18th alone. That single day would clear the June triple-witch near $7.7 trillion, which was itself a record. Add quarter-end pension rebalancing, with funding ratios near 112% and plans de-risking out of stocks and into bonds, and the plumbing itself leans against equities into month-end.
Notably, none of this guarantees a market selloff. However, it does stack the odds against overly aggressive investors. Currently, every major desk from JPMorgan to BofA has turned cautious. However, CNBC’s own investment committee is refusing to sell a single share into the weakness. That crowd can be right about the direction and still be wrong, or early, on the timing. Such is the nature of a market that loves to punish the obvious trade.
A Second Desk Lands On The Same Downside
While Scott Rubner reads the market through flows, BTIG’s Jonathan Krinsky reads it through the tape. Interestingly, he lands in nearly the same place as Rubner. Krinsky’s framing is that the post-summer rally has been a game of “musical chairs” rather than a true “broadening.” Money rotated out of Technology and AI into Consumer Cyclicals and Large Cap Value. At the same time, the index sits roughly where it did on June 2. Breadth has quietly rolled over. The share of Russell 3000 names above their 50-day average is the lowest since early April. Furthermore, the one-month correlations just jumped to their highest level since June. That is a classic tell that names begin to fall together.
The other half of the concern is investor complacency. The five-day put/call ratio sits near 0.82. That is one of the lowest readings in years. Notably, the tape has not printed a single 80% NYSE downside-volume day in almost a year. That long stretch falls against a historical average of 21.
Lastly, Krinsky’s base case is a failed retest of the 7,600 breakout, followed by a slide toward 7,200-7,300. Such a pullback would encompass 7% to 8% off the highs. While not a meaningful decline, given the market’s low volatility and high investor complacency, it will “feel” much worse. That lower zone sits right on Rubner’s midterm seasonal math and the rising 200-day average near 7,127.
Think about it this way. When both a flow desk and a technical desk reach the same number from opposite directions, you should at least respect it. Crucially, none of that means that it will happen with absolute certainty, nor does it pinpoint the day. But it is certainly a risk worth appreciating.
What Should Investors Do Now
So, what does this all mean for investors? Most importantly, this is a tactical market reset, not a call to abandon equities and go hide in cash. Scott Rubner himself framed the September weakness as a “better entry point ahead of a more constructive mid-October.”
He is correct. Once mid-October arrives, the options expiry will have cleared, the FOMC will have met, and corporate buybacks will have resumed. Notably, the market will be focusing on Q3 corporate earnings reports. which typically support markets heading into November.
Therefore, the investor playbook is to use market strength to rebalance portfolio risk rather than chase it.
The moves worth making now are the unglamorous ones. Start by taking profits and banking gains where a position has run well past its intended weight. Raise a little cash so a pullback becomes an opportunity rather than a scramble. Then add downside protection while it is still on sale. Why? Because the whole point of Rubner’s note is that the insurance is cheap today and may not be next week. Such is the value of preparing before the crowd decides it has to.
September rarely hands out cheap insurance and a clear warning at the same time. When it does, the disciplined move is to take both.
* * *
Heading into next week, the support and resistance levels are evident. The first resistance is the record at 7,796, about 1% away. Just above that are the round numbers at 7,900 and 8,000. (Those are our year-end targets that sit just above previous all-time highs.) Conversely, support starts at the 50-day near 7,585. That level also marks the breakout that a failed retest would expose. Just below that level is the 7,300 zone, then the 200-day at 7,137, the same downside band the seasonal math points toward.
With that setup going into next week, we will want to continue playing defense rather than offense. Secondly, investors should consider increasing cash buffers keep stops under the 50-day. Lastly, use any push toward the record market levels to trim rather than chase.
To be fair to the bullish camp, a decisive close back above 7,796 would neutralize the momentum warning and reopen those round-number targets. There are several risks ahead, from the mid-term election cycle to the loss of corporate buybacks, so this is a two-sided setup rather than a directional call. However, pay close attention to the 7,585 next week. If the market can hold that level, the uptrend will remain intact. If it fails, the seasonal downside risk increases.
Key Catalysts Next Week
Next week is a holiday-shortened trading week with one question that will dominate it.
“Does inflation confirm the hike that Friday’s jobs report just put back on the table?”
With the market closed on Monday for Labor Day, that stacks the two prints that will matter the most at the very end. PPI lands Thursday morning and CPI follows Friday, both at 8:30 AM ET, and both feed straight into the September 16 FOMC decision.
This week is where the Fed debate will get settled. Ellen Zentner, chief economic strategist at Morgan Stanley Wealth Management, framed it well after the jobs report. The upside payroll surprise certainly heightened rate-hike concerns, but the outcome will hinge on next week’s inflation numbers. If CPI and PPI come in cooler than feared, the Fed can discount the hot labor market signal. However, if both prints come in hotter than expected, a September hike moves from a coin flip to the base case.
As far as the rest of the week goes, the slate is fairly thin. Tuesday brings NFIB small business optimism and consumer credit. Then on Thursday, we will see jobless claims, existing home sales, and wholesale inventories. As noted, PPI also drops on Thursday, with Friday’s CPI report coming alongside the preliminary Michigan sentiment read. The Fed itself goes quiet, with the pre-meeting blackout that began September 5 keeping every official off the tape through the decision.
Overall, the earnings calendar remains very light, with the vast majority of earnings already behind us. However, of note, Oracle reports on Thursday after the close and will be scrutinized for AI cloud demand and hyperscaler capex. Its numbers and backlog commentary will swing semiconductors and the broader AI complex more than any single macro release.
Adobe follows the same afternoon. Crude is the other wildcard, with a 9% weekly surge on Middle East supply fears keeping energy and inflation risk alive. Thin post-holiday liquidity can exaggerate the reaction to both inflation prints, so expect sharper intraday swings than the calendar alone would suggest.
Friday’s CPI is THE report for the week, and everything else is pretty much a sideshow until that number crosses.