NDX relative strength vs. SPX was slowly building for several days before the explosion higher on Monday, 9/21. Semiconductors ignoring the AI safety news spread by the CEOs of the big AI firms was an early sign. META Muse news was the spark that got speculators excited about AI again. I don't expect it to last for long, as Muse is just the latest overhyped AI agent, just like Openclaw was earlier in the year. But I also don't want to risk getting run over being short if this speculative wave lasts for more than a few days.
Anthropic delaying their IPO into November has been a blessing in disguise for the Nasdaq. The delay has made tech/growth funds less urgent to raise cash to buy the IPO. That eliminates a near term bear catalyst. Anthropic is obviously having trouble building up enough of an investor pool to buy into a $2T valuation. Its an early peek into investor demand for a $2T market cap money bleeding company that has lots of competition from not only Open AI, but more importantly, from China. Its not going to be easy for them to raise as much capital as they want at their preferred valuation. They will either need to raise less at a higher valuation, or raise more at a lower valuation. Either way, I can sense that there just isn't a lot of enthusiasm to buy into the IPO when the trend is toward using lower cost Chinese open weight models.
But everyone is saying this market can't keep brushing off higher bond yields. Why ignore rising bond yields and just focus on the NDX, when CNBC is talking about bond yields every day? Its because investors are buying stocks because of the AI boom, and that's directly related to the NDX. 50 bps higher bond yields doesn't change an investor's decision to buy stocks or not. The performance of AI related stocks does. NDX strength is the most important barometer for the US stock market. It is the fuel that feeds this bull market. NDX has been outperforming SPX and IWM massively for the past 2 weeks.
NDX outperforming SPX/IWM is the simplest signal of risk appetite. Contrary to what all the technicians tell you, it is not the number of stocks going up that is important. It is the number of the most heavily owned stocks going up that is important. The most heavily owned stocks are AI related tech, in particular semiconductors. If the most heavily owned stocks by speculators are outperforming the SPX, then you are not likely to get a big down move.
Downtrends happen when speculators are losing money and forced to sell. The type of investors that are losing money is important. If value investors are losing money, that doesn't mean much. Value investors don't sell when stocks go down, they usually want to buy more. Same goes for most bond investors. But stock investors, in particular fast money stock investors, are mostly trend followers. When stocks are going up, they buy more. When stocks are going down, they either stop buying or they sell. The marginal supply and demand determining short term market direction is coming from these momentum based investors. Pod shops are now big participants in the fast money trading game. Most pod shops are momentum based. Why? Because they have tight stops, and by the very definition of having tight stops, you have to be with the trend or you get stopped out.
That being said, NDX outperformance doesn't last forever. When it turns, it can lead to vicious down moves not just for NDX, but for the broader market. The more crowded the NDX long position, the sharper the down move when it happens. My gut feeling tells me this NDX over SPX trade doesn't have a lot of room to run. Tech is just too heavily owned by retail, by hedge funds, and the biggest IPOs are all in tech. Perhaps we get a sharp reversal in the NDX over SPX trade right ahead of the Anthropic IPO in November. Similar to the NDX weakness vs SPX in June ahead of the SpaceX IPO.
Net equity demand from Federal Reserve data as of June 30, 2026. Investor inflows into stocks are doing the heavy lifting, as corporate net repurchases goes negative.
A quick word on bonds. Based on the speed and continuous downtrend in bonds, it looks like we are into the capitulation phase for bonds. The top in 10 year yields in 2006 and 2007 around 5.25% is a psychological resistance level where you could see some buyers stepping up. Bonds are not a screaming buy, like some kneejerk contrarians are stating, but we are getting close to levels where you can buy for a high probability short term trade. Long term fundamentals are still poor as supply is huge, inflation is sticky, and the AI boom still ongoing.
The original plan was to hold for a big move lower as we were in a seasonally weak window, along with a complacent market that was grinding lower, with no signs of investor capitulation, and with NDX being a relative underperformer vs SPX. That all changed, and the market is shrugging off the relentless rise in bond yields.
So I scrapped the plan and got out of all shorts this week, as NDX outperformance was just too much to bear. You have to trade the market that is, not the market that you want. Its just unlikely that you'll get that big move lower that I was expecting for September/October. It doesn't mean I'm completely giving up on shorts in the near term. Without a big move lower, one has to be more tactical and short only at certain levels, and cover when you get pullbacks.
We never got the capitulation during the 1 month slow grind lower from mid August to mid September, so I don't expect an extended rally. This week, you've already seen a total retracement of the big Monday up move in SPX. That's not common. Usually a total retracement of such a big move takes several days to develop, not 2 days. So you aren't getting any V rallies this time. It means that rallies are shortable, but that profit targets need to be reduced and you need to get out of shorts on dips, with expectations for another bounce coming again. A bounce towards SPX 7780 to 7800 would be enough to get me interested in re-shorting.
Posts
No comments:
Post a Comment