[Week 34 of 2026] Claims, Credentials, and Calculations

Welcome back to Price and Prejudice with a few musings from Week 33 of 2026.

Mandating Insurance

This WSJ analysis finds that auto insurers closed 45% of liability and medical claims without paying anything last year, up from 35% a decade ago. I actually did not know this was a huge deal — 35% even a decade ago seems like a pretty high number to me? My mental of insurance companies’ marginal cost is the probability of accident and the magnitude of the resulting payout, but I suppose there’s an extra margin that insurers have in managing their costs.

Auto insurance is one of the few products where Americans are legally required to buy, and the requirement attaches to the purchase rather than the performance. While it’s easy to compare premiums across insurers, it’s quite challenging to compare the probability of being paid since denial rates are not published. So competition ensues on attributes that customers can directly observe and leaves the unobservable component free to drift. And what the article covers is a natural consequences of these firm-level optimizations.

Another interesting angle to this is that mandating insurance can actually backfire and provide less protection. While the article talks about many factors that are responsible for this trend (e.g. litigation), one important consideration is the price of insurance. Consumers who are financially constrained will opt for the least amount of coverage that meets the regulatory requirement, which means that when there is an accident the insurer will not have to pay out as much.

Financial Advice with Charisma

An audacious WSJ journalist sat out to scrape TikTok and review close to 50 hours of money content across 200+ accounts that provide financial advice. As expected, most of the people dispensing these advice are without license or certification, but the effect is small — Fidelity credits these accounts for bringing in a big chunk of Roth IRA contributions Among Gen Z in the second quarter.

What happens this rise? One argument would be that these advice fill a gap that is not typically covered in more traditional sources of personal advice. But a more probable narrative is that the advice worth giving got much simpler: index funds, Roth accounts, high-yield savings, spend only what you have. When good advice becomes simple enough, the credentials don’t really matter since there isn’t much to certify. Something has to then replace credentials as a signal, and that’s where traits like charisma, personal relatability, and the entertainment content of the shorts come in. Probably the only observable measure of reliability is the displayed wealth of these finfluencers, which admittedly has become easier to fake thanks to AI.

Big World Intuitions

This interesting read is about what’s called “big-world” intuitions, which refer to heuristics that work when you are very small relative to the system you operate in. The article includes a finance example: a small trader should not strategic around market impact and should simply bid honestly. In finance lingo this behavior is referred to as “price taking,” which is useful if you’re a retail investor but not so much if you are, for example, Leopold Aschenbrenner. For large funds, a big chunk of effort goes into execution algorithms because they are not in a “big world.” This intuition also nicely ties back to the social media financial advice — indeed, the correct advice for a 24-year-old with three thousand dollars is a big-world heuristic, which can be dished out in a Youtube short.