I've been noticing a lot of handwringing again on twitter and elsewhere about how the market is overvalued, about how there is an AI bubble. A few days ago we had CEOs from MS, Goldman and JPM echo such warnings. The reflexive contrarian response to this would be that this is evidence of a wall of worry for the market to climb higher still. Well, it's not always that simple. It could be that these worries are justified but early. Despite the bullish contrarian implication of the cautious narratives out there, there are pockets of excessive bullish positioning as well as poor market breadth which suggests the market may indeed be vulnerable to a correction and perhaps a sizeable one, but as mentioned, with all the hand wringing out there, perhaps the correction happens a litter later on as a watched pot doesn't boil so to speak. Let's go over the evidence.
I've mentioned how AAII sentiment has not been showing excessive bullishness in the weekly surveys but this contradicts how they are actually positioning themselves. According to the monthly allocation survey, AAII investors now have 70.5% allocated to equities. Anything above 70% is considered "excessive" territory. Now, a single month reading above 70% is not an automatic bull killer by any means; it may only suggest a correction is forthcoming as was the case in the summer of 2024 which was the last time equity exposure hit 70%+ which first occurred end of May, remaining at 70+ until end of July. Then, in early August the market had a sharp correction due to a growth scare and the unwind of the Japanese carry trade. Prior to 2024, AAII equity exposure hit 70%+ at the end of March of 2021 and stayed above 70% for the remainder of the year....we all know what happened starting in 2022. Prior to 2022, equity allocation hit 70%+ for 3 straight months starting at the end of December in 2017 leading up to the 10% correction in late Feb 2018. So, based on history, hitting 70%+ for the first time is indicative of late stage rally behavior precluding at least a significant correction but it also suggests that the market may still have a bit more more time and room to go higher first before the major downside happens! If we go by history on this indicator alone, it suggests the likelihood of a significant, 10%+ correction would be 2-3 months from now. Of course, you should not hang your hat on any single indicator, so let's look at some others.
Sentimentrader posted a chart on twitter titled "Titanic Syndrome and Hinderburg Omen" which captures the weakness we've been seeing in the internals. As you can see, this had preceded some nasty corrections.
Investor's Intelligence bull/ratio is a lobsided 4:1 which is quite excessive. This is in stark contrast to AAII bull/bear ratio which is neutral. It's quite odd how these 2 surveys are diverging so much like this...which one do you believe more?
Despite the above mentioned warning signs, there are still plenty of good reasons to believe that despite what appears to be a forthcoming correction, it wouldn't signal the end of the bull market because longer-term sentiment isn't excessively bullish enough. I've mentioned the BofA bull/bear indicator many times already, so here's another 2 charts which gives the same message
.
You can see that positioning is bullish not stretched and if we did see a significant correction, it would probably reset positioning to bearish quickly.
Let's talk about this "AI is in a bubble narrative" again. The fact that you are hearing this everywhere to me indicates that it's either not a bubble or if it is, this call is early. As I said earlier, this doesn't mean that there's no froth out there or strong warning signs of at least a correction....there is....but to me it says that if we get one, no matter how bad it looks, somehow, someway it will not be the end of the bull market because we've haven't seen the narratives which embrace an optimistic outlook, which embraces AI fully - not half ass like now where it's pretty much a 50-50 split of optimists and pessimists. We've also have not seen a flurry of AI IPOs including the big kahuna Open AI which apparently won't happen until second half of 2026 at the earliest. Again, I'm not saying there's no signs of overheating in AI or the markets in general, but we are far from the euphoria of the dot com bubble. Back then, the giddiness was palpable and you could hardly find anyone who was bearish.
