More bank failures, and more predictions

Welcome to Nonrival, the newsletter where readers make predictions.

How it works

On Sundays, read the newsletter and make a forecast by clicking a link at the bottom.
On Wednesdays, see what other readers predicted and how your forecast compares.
Over time, you’ll get scores based on how accurate your forecasts are.

In this issue

Update on bank failure forecasts
What prediction markets think about SVB contagion
Nonrival will be off this weekend and next week. Back soon.

Thanks for forecasting. Send feedback to newsletter@nonrival.pub.

Well that escalated quickly. Just a day after Nonrival asked whether any more of the top 50 US banks would fail, the government announced that Signature—ranked 30th by assets—was no more.

Nonrival closed forecasting at 7pm Sunday night, since the question had been resolved. If you made a forecast before that, it will count. Forecasts that came in after that time won’t count.

In a moment, we’ll review what prediction markets can tell us about what happens next. But first, here’s where Nonrival forecasters were as of Sunday night. (No scores yet, since the Sunday night cutoff makes scoring slightly trickier.)

This was the rare case when readers were divided at the extremes. The average forecast was a 46% chance of another failure, but the most common responses were Very Unlikely followed by Very Likely. No question has divided Nonrival readers as sharply as this one.

How your forecast compares

You said there was a [031223_FINAL GOES HERE]% chance of another top-50 bank failing this year.

You predicted that the average of readers' forecasts would be [031223_CROWD GOES HERE]%. The actual average was 46%. You were closer than [031223_CROWD_RANK GOES HERE]% of readers.

What other readers said

Readers’ rationales

Here are some of the rationales among forecasters who thought another failure was likely:

70%: Banking regulations are too lax. Greed is too seductive. Panic is infectious. The crypto world is the perfect example. More dominos will fall. Call it a correction if you will.

80%: In an effort to separate from the pack, somebody is going to have made what will in retrospect be an obviously terrible choice. I expect at least a few more failures.

80%: Banks with concentrated customer bases are already struggling. First Republic and Signature are on that list. Both appear to be at risk based on market action last week, and the year ahead is long and full of interest rate stresses. Depositors must be fully reassured in the Silicon Valley Bank unwind for any hope of calmer seas ahead. If they aren't, if there are delays or spillovers to other startups (and we don't know the full list of which companies had accounts at SIVB yet), the nerves of other depositor at other institutions could fray if they are above the 250k FDIC threshold. If Signature goes down, the crypto industry is in even deeper trouble than it already was. In tangentially related news, another fintech bank is by far the largest customer of a onetime tech high-flyer. There are cracks in the system. How far they spread is uncertain, but the year is long and there are lots of ways for things to break. I may be overestimating risk here, but so be it.

90%: The base rate shows that either no bank fail or several banks fail. Therefore, it seems unlikely that this bank will be the only one.

What happens next?

Nonrival published a roundup of prediction markets related to SVB, which you can check out here. As of Tuesday afternoon, here are some of their predictions:

63% chance another bank buys SVB by March 18, per Manifold
60% chance of at least one more big bank failing other than Signature, per Good Judgment Open
More layoffs in tech: The chance that March sees elevated tech layoffs (30K+) jumped significantly on Good Judgment Open

Nonrival