Sell in May, and go away? Vol #345
Paid subscriber analysis
Well, there it is folks, the first >1% intra-day sell-off since the end of March. And immediately, the bears are looking jumpy and excited. So naturally, time is ripe for the classic “Sell in May and Go Away” adage. Yesterday’s monthly OpEx was the ideal trigger. In hindsight, obviously.
Before we jump into the seasonality narrative, a few factors that we should take into account.
First, the 10Y yield ended the week at almost 4.6%. This is concerning. We could be at the forefront of yet another decoupling between the SPX and the 10Y (see chart below). When bond yields are this high, and especially if they keep pushing higher, this creates an enormous headwind for equities. The structural factors are still there pushing markets up, so don’t expect any huge corrections, but having yields rising at this point becomes an issue.
Second, the volatility in bonds (measured by the MOVE Index) shot up this week. And bond vol is typically a good predictor of equity volatility (see below). Yes, we got a slight VIX spike yesterday, but the move in the MOVE was more significant.
Third, and maybe the most concerning this week - markets are now pricing in one rate hike for the last meeting of 2026, and even one more rate hike (just barely) in 2027. Just a few weeks ago we were expecting rates to go to 3% in 2027. Now we are discussing 4% again. Not good. Viva higher for longer!
All these are related. Higher (and persistent) inflation creates pressure on central banks to raise interest rates again, which means that both the short end and the long end of the yield curve adapt (bonds are getting sold, which drives their prices down and yields up). The selling of bonds implicates bond volatility, and the MOVE index spikes, creating upward pressure on equity vol.
The next few weeks should be very interesting, especially with NVDA earnings coming up on Wednesday.
Should we, then, sell in May and go away?
It’s been a couple of years since I last ran this table, and I figured it was worth a refresh. Especially after this week. The original sample covered 1967 to 2022, so 56 full years of SPX monthly returns, sliced four ways: overall, post-2001, during recessions, and during recoveries. With 2023, 2024 and 2025 now in the books, we’ve got three more full years of ordinary expansion to add to the Overall and Since-2001 columns. Neither year was a recession or a recovery - 2024 was a steady grind higher, 2025 had a brief tariff-induced panic in March-April before melting up into year-end - so the right-hand columns are unchanged from last time.
A few things stand out.
