Some stocks are killing it. A bunch aren’t.
Jon Sindreu at the Wall Street Journal called it an “Investing Riddle” (subscription required) that stocks in the market are in turmoil but the stock market isn’t. Yesterday, the S&P 500 closed at a new record, getting there in the last 10 minutes when so-called “reference prices” are determined.
Those are the prices that indexes and ETFs need. I think it answers the riddle. Stay with me.
Mr. Sindreu means that the benchmark goes up while in many of the components there’s undisciplined chaos. He notes how GOOG is down, META way up.
That so-called “dispersion” – absence of correlation – is pervasive. There are ten stocks in the S&P 500 up 100% or more the past year. PLTR leads, up 370%. We measure Demand – buying and selling – on a ten-point scale. PLTR spent 87 of the 255 days comprising the trailing year at 10.0. Momentum is Demand at 10.0.
But get this: 180 of the 504 current components of the S&P 500 returned at least the benchmark’s performance of about 23% the past year. That means 320 components, or 64%, underperformed the benchmark that they comprise.
How is that possible?
I’m reminded of a funny story that I shared years ago. The annual cost of the Wall Street Journal basically doubled. We called and were routed to some off-shore center. Karen said, “It’s up more than 100%. You do the math.”
She heard silence, some clicking. Karen said, “Did you hear?”
The person on the other end of the line said with an accent, “I am doing the math.”
We still laugh about that literal response (double entendre intended).
Anyway, I’m sure there’s some math that adds up. Mr. Sindreu suggests dispersion is maybe irrational exuberance – people overreacting to news both good and bad.
That’s mathematically impossible. Every constituency is a net seller of US equities save indexes and Exchange Traded Funds. As I wrote last week, the stock market is about beta, not alpha. The benchmark.
Mr. Sindreu also muses whether a hedge-fund strategy to sell volatility in indexes and buy it in individual stocks is creating dispersion. Maybe. But it supposes hedge funds have pervasive capacity to season the entire stock market with their own blend of spices.
I don’t see it in the data.
But I’ll tell you what we do see: No rebalancing. I don’t believe indexes and ETFs are selling. At all. Even when they should.
Prior to the Pandemic, there were clear patterns and periods of selling. We measure it. After the Pandemic, those patterns and periods are greatly elongated. In fact, there’s only one since 2020. In 2022 (fitfully between Jan-Nov). We had a fleeting bit in September 2024 but no pattern.
And the standard deviation of volatility from Jul 2017-Feb 2020 averaged 17. Since then, it’s 8. In fact, on Feb 5 this year, it was 1.0. Startling.
Yet weirdly, how much SPY rises and falls daily is way up. From Jul 2017-Feb 2020, SPY rose $1.20 per day and declined $1.40 each day, on average. That’s intraday volatility – how much price moves between highest and lowest prices.
Spreads up but standard deviation down? Evidence of machines setting prices. Not humans. And realize that stocks will always show a propensity to fall more than rise daily because of the bid/ask spread. The bid is always lower than the offer.
So how do markets rise? Price CLOSES over the midpoint, despite all that price-movement making Jane Street et al rich.
Since Mar 2020, SPY has averaged $2.90 per day down, $2.52 up. More than DOUBLE pre-Pandemic levels.
Unsurprisingly, volatility is also sharply higher (but its deviation is down by more than half — and that’s how machines calculate returns on investment horizons of a day or less. And they know what’s in the ETF basket to boot). Pre-pandemic, SPY moved 0.9% daily – less than 1%. Since Mar 2020 it’s averaged 1.3%, up nearly 45%.
What the hell is going on?
I’ll tell you what I think. You have to begin with the mechanics – market structure.
Indexes and ETFs are attracting literally 100% of inflows because Actives are net sellers. How do ETFs work? Shares are issued to brokers in trade for a basket, a statistical sample of the benchmark. ETFs added $2 trillion of assets the past year, the Investment Company Institute shows. There are now 3,620 ETFs traded in the US stock market, half those built on domestic equities.
Yet the number of public companies continues to decline, falling now below 3,350 (versus over 7,400 25 years ago). Companies keep buying back their stock. Five investors own close to 40% of all shares.
What if we’re running out of stocks? It would explain everything we’re seeing.
Let’s say ETFs are being created via “cash-in-lieu” rather than shares. That means a broker gives Blackrock or whomever cash equal to the value of a basket of stocks. Because they can’t get the stocks.
But Blackrock issues ETF shares. That would explain the distended patterns without correction, the rising volatility from derivatives and machine-trading uninterrupted by selling. The dispersion in the stock market. The relentless gains.
Maybe I’m wrong. But you do the math.
If the benchmark keeps pegging reference prices while many of its stocks carom like molecules, the culprit is the statistical sample — or cash in lieu. And the cause is ETFs.
