Hacker News Reader

Comment by bell-cot | Hacker News Reader

No mention of the years-long (and finally "successful") lobbying effort by big banks, to get Congress to let them do far riskier stuff.

No mention of SVB's high-risk bets that the shape of the (interest) rate curve wouldn't change much.

No mention of the American financial sector's long, long history of suffering big losses, after high-risk bets on that curve's shape went sour.

And (sadly, once again) there seem to be no chance of those responsible facing any serious consequences.

View on HN

> No mention of the years-long (and finally "successful") lobbying effort by big banks, to get Congress to let them do far riskier stuff.

First, the 2018 Dodd-Frank modifications did not affect the Big Four banks at all.

Second, the 2018 modifications had nothing to do with SVB's collapse. SVB collapsed because it made a bad bet on where interest rates were going (and did not hedge for risk), making it vulnerable to short-term liquidity issues. That has nothing to do with the 2018 raising of the minimum AUM for application of Tier 1 capital requirements from $50 billion to $250 billion.

To put it simplistically, Dodd-Frank has to do with solvency. SVB was solvent; it was not liquid.

It would not have passed a stress test.

The SF Fed's warnings last year were that, warnings; SVB management did not listen (quite possibly because the bank lacked a CRO for most of the year). That's not the same thing as the bank failing an actual, formal stress test.

While SVB and Credit Suisse's collapses' causes are directly unrelated, depositor outflows were involved in both, as opposed to fundamental issues with their balance sheets. Credit Suisse's Tier 1 ratio was as good as that of UBS or any other big European bank. <https://www.wsj.com/articles/credit-suisses-death-gives-birt...> At the end of the day, a loss of clients' faith in a bank is something that no stress test can directly measure ahead of time.

EDIT: SVB's Tier 1 ratio was 12%, among the highest for US regional banks. <https://www.morningstar.com/articles/1144363/which-bank-stoc...>

Larry Summers and Dan Tarullo explicitly confirm that SVB would have passed a Fed stress test in 2022. <https://www.youtube.com/watch?v=cYgz4KSRPww#t=5m10s>

I think the Banking industry should bear the responsibility for the shortfall in depositors funds. I think they should be free to figure out among themselves and when things go so wrong that the Banking industry can't figure out we need Congress to regulate. I have no problem looser regulation but I believe it comes with greater responsibility.

> I think they should be free to figure out among themselves

Why should it be normal or even ok for bankers at competing banks to come together to "save" another bank (more specifically their depositors)? That seems like a recipe for a cartel in any other industry.

> I have no problem looser regulation but I believe it comes with greater responsibility.

This seems non-sensical to me. You want some set X of measures to be respected, but you don't want that to be explicitly enforced by government regulations - why exactly? If that set X is important, then why not make it explicit and directly enforceable? If it's not important, why should bankers be held responsible if they fail without following it?

I think it's pretty clear that the proper way of doing this would be to have official government regulation + official government guarantees for depositors at any bank that follows those regulations + actual enforcement and verification of those regulations. And any regulated bank failure should mean additional regulation to prevent those failures, similar to how the airline industry (used to) operates.

I don't think we disagree. I will try to restate my point more clearly. Banking isn't life or death. It's depositors and corporations. I place the highest value on depositors. If depositors expectations aren't met the Banking industry is responsible. If the entire banking industry can't figure out amongst themselves how to meet the expectations of depositors without going bankrupt the government regulations would require the banking industry assets liquidated. The new banking industry would start with depositors made whole and what if anything if left over could be used to fund a new banking industry under new government regulations.

>No mention of the American financial sector's long, long history of suffering big losses, after high-risk bets on that curve's shape went sour.

Can you provide some examples? For instance, the GFC was caused by bad loans, not interest rate risk.

The Savings and Loan crisis of the early 1980s was kicked off by rising interest rates that simultaneously made the assets (mainly mortgages) they already held not as profitable and also made their depositors want to pull their money to get a better rate on their savings accounts.

This also led to some of the S&L’s betting more on riskier loans and assets, hoping to make it all back later.

https://www.federalreservehistory.org/essays/savings-and-loa...

> No mention of SVB's high-risk bets that the shape of the (interest) rate curve wouldn't change much.

They for sure weren't good bets, but what part of that bet is high-risk? They're government bonds, you still get the complete investment back with interest and there's (supposedly) no chance it will fail. The only thing that made it risky is their lack of liquidity combined with the subsequent bank run.

Bank risk has 2 major components credit risk and interest rate risk (duration risk is a subcomponent).

SVB was over indexed on credit risk and under indexed on interest rate risk (so they could make more profit).

This is banking 101 stuff and risk managers (including those at SVB) will point out when this imbalance happens.

2008 was about credit risk, large swaths of bank assets went down because the underlying credit defaulted. 2023 is about interest rates, large swaths of bank assets went down because they were attached to too low interest rates for too long.

If you are dealing with people who don't understand bond investing, and can disable their sense of caution with "US Govt. Guaranteed!" - then it is very easy to lose a boatload of their money. (As I learned the hard way, 4+ decades ago.)

A very basic example - take a peek at the 1-year return on investing in US Treasury 20- to 30-year STRIP bonds:

https://www.spglobal.com/spdji/en/indices/fixed-income/sp-us...

(Summary - the 1-year return is negative 23%.)

Honest, competent bond investment managers will be very cautious about investing in longer-term bonds, and will hedge interest rate risk, and etc. etc. etc. Bond investment managers who lack those virtues can - most of the time - offer better yields and returns. Their blow-ups more than ruin that, long-term...but how many American financial sector executives care about the long term, these days? Certainly not - as other news stories have made clear in considerable detail - SVB's leaders.

"You'll get your money back 10 years later" is not exactly "low risk" for any significant amount of money, even for a bank.

What's the alternative? That banks make no long-term investments at all? If government bonds are too risky to make a 10-year investment on, then what is? Should we just stop investing in projects that take 10 years to get a return entirely?

Regulation to prevent them from getting into dangerous situations.. financial regulations on capital requirements and stress tests, the sort of the things that in 2018 the SVB CEO worked successfully to be removed.

"SVB CEO and President Greg Becker "personally led" the bank's $500,000 lobbying efforts to reduce financial regulations on capital requirements and stress tests, according to the Accountable.US report." [0]

https://www.salon.com/2023/03/15/rolled-back-regulations-got...

Fully agreed there, we should indeed be focusing on the lack of regulation of the bank rather than whether 10-year investment are good bets or not. Because even a bank with lots of short-term bets can be driven into the ground without those kinds of regulations.

Bank assets are typically laddered so that they aren’t dominated by 10 year assets that were all originated at the same time.

A well run bank keeps some amount of assets in lots of different durations including longer terms, just not all at once.

Right, but that's more of a problem with diversification, it doesn't make a 10-year investment inherently high-risk.

It’s not exposed to credit risk but is exposed to interest rate risk.

What SVB did was not best practice, as shown by the fact that they are now bankrupt and the FDIC had to step in to make depositors whole. Why are you painting it as if it was?

I didn't mean to defend SVB, I'm just taking issue with the conclusion that banks should not make 10-year investments and instead only focus on investments that have short-term returns. As if banks weren't already overly focused on that.

Their CRO should have told them this was a risky strategy, or how to hedge interest rate risk. Ooops, they've been without a CRO since Q2 2022.

10 years. That's the risk.

Locking up your money for 10 years is a huge risk. SVBs risk department told the business that slapping the money in 10 year bonds was wrong and that they should put it in short dated bills but the exec team didn't like the lack of profit that would result in so decided to take a massive risk instead.

Reply on news.ycombinator.com