Historically, September sucks for stocks.

The good news is, it’s not September. Yet. It will be, though. Until then, buy stocks?

That’s a question. I can’t answer it for you. Not investment advice. Do your own research. Consult one of those strip-mall financial planners who decorates his office with framed copies of Barron‘s.

No, but seriously, the August seasonal’s favorable for equities even as vol — both realized and implied — tends to inflect higher starting… well, now. The figures below, from Citadel’s Scott Rubner, show you the seasonals both for the S&P and the Nasdaq 100.

“Looking at the past 100 years, the S&P 500 tends to clos[e] the month on the highs, consistent with the number of vacations, pool parties and the general unwillingness to put on a new short during August,” Rubner wrote, in a Tuesday update, adding that the last week of this month is “historically one of the most popular long weekend vacation periods of the year.”

If you’re wondering who might be buying in August, one answer’s retail, although they might be tired after being a net buyer of stocks in 27 of the last 28 sessions, 14 of the last 16 weeks and 19 of the last, uh, 19 months. (Those statistics are from Citadel’s flows data.) Besides, Rubner remarked, “retail activity typically decelerates in August as we head into September.”

But that’s ok, because as discussed here on Monday, vol control as a cohort can continue to add exposure as long as stocks stay relatively well-behaved. After all, three-month rVol still has some catching down to do in the absence of new shocks.

And then there’s the Mag7. What about ’em? Well, they’re still highly profitable. Have a look:

The figure above, from Goldman, shows you EPS growth for the vaunted septet versus the rest of the index. It says a lot about you as a collection of companies when 26% YoY EPS counts as the worst quarter in years.

“Results so far have reflected the continued earning exceptionalism of the mega-cap tech stocks, which represent one of the key sources of upside risk to S&P 500 estimates,” Goldman’s David Kostin said, noting that the Mag7 outperformance illustrated above — i.e., the 22ppt gap between EPS growth for the mega-caps and the so-called “S&P 493” — is nearly double the disparity tipped by consensus ahead of earnings season.

Although this varies by company, there’s more than a little evidence that tens (hundreds) of billions in AI capex spend is starting to pay off for the big guns. As Rubner quipped, “AI capex is the new buybacks.”

But just in case, there’s still “old” buybacks. Corporates are coming out of the earnings blackout, which means they’ll be clear to reengage. “I believe there is capacity for increased share repurchase activity — particularly during August,” Rubner said, in the same note.

The figure above, from Nomura, reminds you that authorizations were running at a record-setting pace headed into Q2 results.

Don’t forget the irony: That’s in part a function of trade uncertainty. When you’re unsure what to do with your spare cash, you can’t go wrong by plowing it back into your own stock. Reduce the float, increase EPS, increase the value of equity-linked compensation. It’s magic!

What about the data? Does anyone care that the US economy’s maybe (probably) headed into a mild recession? In a word, “no.”

“The US stock market does not always reflect the broader economy,” Rubner wrote, adding that recent data might’ve “increased investor bearishness,” but Citadel’s client franchise “was quick to buy the dip.”