[Week 14 of 2026] Congestion and Credit

Welcome back to Price and Prejudice with a few musings from Week 14 of 2026. Now that the Capital Markets & Investments class at Columbia is in progress, the blog will also reference to some contents from class as well.

The Congestion Paradox

Hyun Shin, who was the head of a research department at the Bank for International Settlements (BIS), released a paper just before being nominated as the next Governor of Bank of Korea. I have been reading his work for years, and the nomination is very well-deserved. The paper and FT coverage is a nice parting gift from his time at the BIS.

We briefly discussed stablecoins in class, and the usual case against stablecoins is about reserve risk: are the assets backing Tether's USDT actually there, and what happens in a bank run? Shin's argument is different. Even if reserves are perfectly sound, the rails themselves guarantee fragmentation. He shows that USDT sits on 107 different blockchains, each version a fundamentally different, non-fungible token (USDC lives on 125). Bridging between them costs money and time, and bridge exploits apparently have cost users over $2.5 billion since 2021. Gas fees rise with transaction volumes because validators need to be paid, so as a chain gets popular, price-sensitive users flee to cheaper alternatives while security-focused users stay on the expensive, more decentralized chain. As a result, we end up having more and more blockchains, each fragmenting the liquidity that's supposed to make money useful.

Basically, what Shin formalizes is that this congestion is an important feature – blockchains must be congested to generate enough fees to keep validators honest and fully incentivized, but this means that the network effect that gives money its social value is structurally unattainable on decentralized rails. In short, there are economic limits to the scalability of blockchain, which are ingrained in the incentive structure of any permissionless blockchain.

Insurance-Private Nexus

I have covered private markets multiple times in class, and a recurring theme is the increasing partnership of insurance companies and private equity companies. This article lays out a particular concern associated with this symbiosis: life insurer investments in private credit reached $849 billion in 2024 according to a Chicago Fed estimate, more than double what it was in 2014 and close to half of the $1.8 trillion sector.

In particular, the article uses the term "doom loop" where concerns over private credit leads retirees to surrender their annuities, forcing liquidation of illiquid assets, which creates more distress and withdrawals. In fact, doom loops have become the default fear framework in finance, including a more recent one for NYC (office vacancies lead to lower tax revenue which degrades city services, leading to more vacancies).

Whether the mechanism is real here is a regulator's job to figure out, and historically, annuity surrender rates have been remarkably stable. But the opacity of private credit layered on top of the opacity of insurance balance sheets makes this genuinely hard to evaluate, and that's exactly what should concern people. Insurance markets are where enormous risk allocation decisions happen with relatively little scrutiny, and it's probably not extreme to think that PE exacerbates this opacity.