I have the same question. (It could be simply that they went longer to try and squeeze out a bit more yield, but perhaps there’s more to it than that.)
I have the same question. (It could be simply that they went longer to try and squeeze out a bit more yield, but perhaps there’s more to it than that.)
SVB had a lot of deposit growth (both from an increase in national monetary base and from a big inflow of transfers as venture capital raised a lot of money in 2021/2022) and had to match that deposit growth with something on the asset side. Normal banks are usually constantly making loans (indeed, in aggregate the loans are where the deposits come from) and have a lot more flexibility to change their asset profile—lend to businesses and households with lower credit, extend loan maturities, etc.
But if you're a tech bank and so many of your clients are fresh-faced startups who just raised millions in equity, who are you going to lend to? The startups don't want to borrow money, they just raised a ton of it. You can do boutique things like lend against startup equity or do venture debt projects but broadly speaking there's a cap on how much your clients want to borrow and its far less than what they now have in deposits.
So you go out and invest in a bunch of high quality assets with longer maturities. Treasuries and MBS have very low default risk and you squeeze out a bit more yield by holding them. It really only goes against you if rates go up rapidly, and even then you can theoretically hold to maturity without a problem. You take on duration risk instead of credit risk, and a lot of it, to make money.
Part of the core problem was the deposit base—these were not people who wanted traditional banking services, which is why they didn't bank with a traditional bank. They moved in quickly and moved out quickly, and SVB wasn't prepared for that. I am not defending the decision to invest in so many long-term assets at generationally low yields (it was a losing idea, but more importantly it was a risky idea that went unhedged and they paid dearly for that) but that was the logic.
I have the same question. (It could be simply that they went longer to try and squeeze out a bit more yield, but perhaps there’s more to it than that.)
SVB had a lot of deposit growth (both from an increase in national monetary base and from a big inflow of transfers as venture capital raised a lot of money in 2021/2022) and had to match that deposit growth with something on the asset side. Normal banks are usually constantly making loans (indeed, in aggregate the loans are where the deposits come from) and have a lot more flexibility to change their asset profile—lend to businesses and households with lower credit, extend loan maturities, etc.
But if you're a tech bank and so many of your clients are fresh-faced startups who just raised millions in equity, who are you going to lend to? The startups don't want to borrow money, they just raised a ton of it. You can do boutique things like lend against startup equity or do venture debt projects but broadly speaking there's a cap on how much your clients want to borrow and its far less than what they now have in deposits.
So you go out and invest in a bunch of high quality assets with longer maturities. Treasuries and MBS have very low default risk and you squeeze out a bit more yield by holding them. It really only goes against you if rates go up rapidly, and even then you can theoretically hold to maturity without a problem. You take on duration risk instead of credit risk, and a lot of it, to make money.
Part of the core problem was the deposit base—these were not people who wanted traditional banking services, which is why they didn't bank with a traditional bank. They moved in quickly and moved out quickly, and SVB wasn't prepared for that. I am not defending the decision to invest in so many long-term assets at generationally low yields (it was a losing idea, but more importantly it was a risky idea that went unhedged and they paid dearly for that) but that was the logic.