Hacker News new | ask | show | jobs
by JohnFen 1191 days ago
In other words, SVB was willing to shoulder a greater amount of risk than the more staid banks. There's nothing wrong with that at all -- but taking on more risk is taking on more risk, and sometimes you end up getting burned by it.
2 comments

Not to single you out JohnFen, but I'm seeing this take in many places, and it is very wrong. The risk here is being attributed to the wrong place. A startup checking account should be low risk, it is just plain cash as it comes in from sales and/or funding round investments. No derivatives, no funky illiquid stuff, just cash. The activities of the startup _may_ be risky, and it can totally fail because of that...by running out of cash, that should be safely and easily available in a deposit account. A startups cash can hit zero, but they expect to be able to access it so that it can hit zero!

So the argument that by the mere fact SVB banked startups as customers was more risky than big banks is not quite precisely right, as big banks have business accounts and plenty of startup customers too. But as other comments have mentioned things were often made quite hard for them there, and SVB was the better product to startups...as just a literal checking account for day to day operations.

The giant risk management failure was in how SVB managed the deposits and indeed the mismatch of the assets it held against them. We all know the long dated UST issue now. Their deposits ballooned so quickly over the last 18mo, the mistake was to put them into such long locked up bonds immediately. A better approach could have been to assume that money could "come out as fast as it came in", and to have held shorter duration securities, shorter maturity bonds, money market funds, repo market commercial paper etc. Then as the steady state pattern emerges post this influx of deposits, then make a better risk management based decision of what proportion to now put in longer dated assets, medium dated assets and shorter dated ones.

And above this specific risk, clearly in hindsight, there was an overall systematic risk in having a non-diversified customer base who all consume the same information sources, highly networked and correlated with each other in their behaviour.

I was not commenting about the SVB collapse and what risks were involved with that. I was commenting generally about exhibitapp's experience.
But that is not what happened here AFAIK. They got burned by being too conservative and investing in government backed securities.

I think from their perspective, they thought they understood tech better than other banks, so thought that they could better analyze the risk (thereby not misclassifying many companies as risky) and use that as a competitive advantage. Right now, it does not look like that part of their business model failed at all.

>They got burned by being too conservative and investing in government backed securities.

Just because one is conservative, does not automatically mean one is not taking on risk.

In banking, interest rate risk is the exposure of a bank’s current or future earnings and capital to adverse changes in market rates. If they put all their money into govt backed securities, they also took on portfolio risk.

I'm pretty sympathetic to SV bank here, but wonder if it was an option to Simply hold cash as a less risky option still when their Business tripled in 2021. They were in a very unique position. When it's all gets Unwound, it will be interesting to see what was going on internally with their decision making and if their Chief risk officer was on board with this or not.
A bank would be better to not take the deposits than hold it all as cash though.
interesting, why would that ever be the case?

Cash on hand would clearly help with a bank run and could be slowly invested in treasuries over a longer duration.

It was not conservative at all! They invested short term deposits in long term bonds (MBS)
that's how banking works. and also why you have depositor insurance
Yes, my comment was about exhibitapp's experience, not the SVB collapse.