This is going to be another (relatively) short note, combined with a section from my new paper inspired by last week’s note. This will be my first “academic” paper, and I’ll share it here as soon as it is finished (currently being reviewed by friends), but this section on “selective cointegration” is absolutely critical to understand what is going on. Candidly, I had not fully grasped it until last week.
In the meantime, Privates are on full display, selling off sharply while the broader market has been in a holding pattern:
I went on Charles Payne’s show this past Thursday to discuss the recent sell-off. Note the vest… I wanted to dress in character (and it’s starting to get chilly!):
The private market has taken many slings and arrows over the past few years, from volatility laundering to concerns about mismarked assets. I pile on. The claimed performance for the sector is simply absurd:
As most know, even if they don’t explicitly monitor it, “private equity” is simply “levered equity” with extra fees. Following the RJR Nabisco debacle, “LBO” became a poor calling card for insitutional investors; so they rebranded:
But that’s really the only thing that changed. Well, funding sources changed as well as the post-2008 era led to bank regulatory arbitrage. Funding low quality loans became capital intensive and banks backed off. Today, the majority of funding for LBOs comes from “private” credit. Now private credit SOUNDS like a credit fund, but remember limited partners (LPs) don’t lend money, they invest equity; that equity gets levered. How? It’s banks all the way down:
At the center of all those green and red lines sits a levered operating company, responsible for making payments to owners (PE) and lenders (interest and principal (PC+Banks). The returns to private equity are thus no different than the returns to a levered operating company and the returns to private equity fund investors are no different than the returns to a fund of levered equity. The law of one price says it must be so:
The Law of One Price: In an efficient market, identical goods should sell for the same price when expressed in a common currency, once transportation costs and trade barriers are accounted for.
Obviously, this has not been an “efficient” market for some time:
There are two possible conclusions:
The barriers to entry in private equity are so high and the skill level is so extreme, that they are able to extract 17.66% in annual alpha after fees.
They are lying.
The pattern of distributions suggests the latter:
But lying has consequences, and not just for the PE or PC investor; it also impacts competition. And competing with private credit for loans to these same operating companies has largely been the province of regional banks even as money center banks have been lending low cost funds to private credit:
And we are discovering that they did a “nicht so gut” job of underwriting:
A “sweeping betrayal of trust”? No wonder we want “trustless” settlement in Bitcoin (sarcasm). Now the Zion fraud appears to be largely tied to commercial mortgages. But the underpinning facts are astonishing:
Zions’ subsidiary, California Bank & Trust, is suing Andrew Stupin and Gerald Marcil – the until now, relatively unknown managers of several funds utilizing the name “Cantor Group,” along with their associate Deba Shyam.
The lawsuit, filed in Los Angeles County on Wednesday, alleges, a “sweeping betrayal of trust by sophisticated financial borrowers who abused CB&T’s confidence, manipulated loan structures for their own enrichment, and systematically eliminated the collateral protections that were supposed to secure the bank’s loans.”
Zions and a lawyer representing the defendants didn’t respond to CNBC’s multiple requests seeking comment. The relationship in question stemmed from about $60 million in financing that Zions’ CB&T made in 2016 and 2017 to two related investment vehicles, Cantor Group II and Cantor Group IV.
Loans made in 2016 and 2017 are going bust with 100% loss? HOW? Subordinated collateral? Just like First Brands. Fraud? Tricolor. A pattern is emerging that supports interpretation two: faced with a massive influx of capital over the past decade, private credit and private equity “spent” like drunken sailors; and now the hangover is due. Frauds are exposed when financial conditions tighten. I realize many of you like to use the “official” definition of financial conditions, which are ostensibly loose as a goose. But financial conditions are just credit spreads, and credit spreads are wrong as weak primary issuance has led to secondary market buying activity in dedicated credit funds. Last time credit spreads were this tight? June 2007:
My model of credit spreads captures this, as it did in 2007, and suggests high yield credit spreads should be closer to 600 than the current 300. And the latest data has not been favorable with a substantive widening underway:
The good news? Nothing in this chart says the party ends tomorrow. But the party is ending. The wave of fraudulent defaults and weak recoveries means banks are starting to probe; in other words, credit conditions are tightening. You don’t find fraud when you’re sending money out — you discover it when you try to bring it in.
Selective Cointegration
There’s going to be math here. Apologies in advance. One of the charts that has haunted me over the past few years now has an explanation:
This chart has been problematic because my models suggest correlation should be rising. And it has for the median stock:
The chart above tracks “comovement” — how many stocks in the S&P500 are moving in the same direction on any given day. It normally tracks correlation, which is a more refined measure that includes amplitude of movement and weights; it has not done so recently:
This has been one of the most important of my concerns. I always maintain the right to simply be “wrong” in my analysis of passive — and if I was, this was the chart to focus on.
But it turns out that there’s a very “simple” explanation — selective cointegration tied to rising concentration. Here’s the math.
Introduction: The Weight-Flow Feedback
Consider Apple (AAPL) at 7% of the S&P 500 and Carmax (currently the smallest stock in the S&P500) at 0.01% of the S&P500 by market capitalization. When the index receives $1B in inflows, Apple stock receives a $70MM buy order and Carmax receives a $100K inflow. Using standard market impact rules from Bouchaud (2018), the expected price impact for each stock can be estimated as:
Populating the data with observed information as of October 2025, we can estimate the price impact as:
Keep reading with a 7-day free trial
Subscribe to Yes, I give a fig... thoughts on markets from Michael Green to keep reading this post and get 7 days of free access to the full post archives.