Dry Powder · measuring weekly
Stablecoins are digital dollars — roughly $306 billion parked in crypto, each token meant to always be worth $1. When that pile grows, risky stocks tend to rise and safe ones don't. That pattern holds up. Finding an individual stock that follows it did not, and this page shows both.
Loading the latest numbers…
Every week we measure how much the crypto cash pile grew or shrank, then look at which stocks moved along with it.
There's a catch. When the whole market has a good week, nearly everything goes up — so a stock rising alongside crypto might just be rising because everything did. Before comparing anything, we subtract out the part of each stock's move that the S&P 500 and the Nasdaq already account for. What's left is the part that looks specifically tied to crypto cash.
Testing one company at a time doesn't work. Two and a half years is simply not enough history to say anything solid about a single stock. So we group stocks by industry and test the groups: more data behind each test, less noise.
The main test compares the riskiest groups against the safest — crypto companies and speculative tech on one side, household goods, healthcare and utilities on the other. We settled on that comparison before running it, which matters: try enough combinations and one will eventually look impressive by accident. A single planned test can't cheat that way. It came back clearly positive.
How to read those numbers. A sensitivity of 2.2 means that in a week when the crypto cash pile grows by 1%, that group of stocks tends to move about 2.2% beyond whatever the market as a whole did. A number near zero means no connection. A negative number means the group tends to drift the other way.
Here is that measurement applied to 94 individual stocks. It's tempting to read the top of the list as a set of discoveries. It isn't.
Flip a coin ninety-four times and some stretches will look uncanny. Test ninety-four stocks and roughly five will look strongly linked to crypto cash by pure chance — that's just what happens when you run a lot of tests. So we apply a standard statistical correction that raises the bar to account for it. Afterwards, none of the 94 clear it.
Switch the sort to “how solid” to see the bar. How big an effect looks and how sure we are about it are two different things — a large number measured badly is worth less than a small one measured well. Sorted by strength, the top stock is one we have little confidence in. Sorted by how solid the evidence is, it drops well down the list.
| Symbol | Industry | Sensitivity | Without Nasdaq adj. | Fluke odds (p) | Bar it had to beat | Verdict |
|---|
An earlier version of this page said three stocks held up: MicroStrategy, Palantir and Coinbase. Those results are gone. The reason is worth explaining, because it's the kind of mistake that's easy to make and easy to hide.
Crypto trades every day of the year. The stock market closes on weekends. Our first version lined the two up by calendar week, which accidentally let Saturday and Sunday crypto activity count toward a stock-market week that hadn't started yet. Compare a stock's Monday-to-Friday against a crypto measure that includes the weekend before, and you'll find a connection whether or not one exists.
The fix was to use the stock market's own calendar, and measure the crypto pile only on days the market was actually open. Every individual result then shrank by a quarter to a third. Super Micro dropped from 2.75 to 0.81 — almost all of what we'd measured had been that weekend effect. Three surviving stocks became zero.
The industry-group pattern held. That's the honest result, and it's the one on this page.
It doesn't predict anything. We compare a week's change in the crypto cash pile against that same week's stock moves. By the time you can see the number, the move has already happened. This describes which stocks are wired to crypto sentiment — not what any of them will do next.
It isn't investment advice. It points attention at something interesting. That's the whole job.
And it's early. This pattern only appeared in 2024. Before then the crypto cash pile actually moved opposite to Bitcoin. Two and a half years isn't long enough to know whether we're looking at something durable or one long stretch of unusual weather. That's what the “provisional” label at the top of the page means — real enough to report, not settled enough to lean on.