[dsomers]

> 4. The old treasuries decline 30-40% in present value. Oops, they're not so safe after all if you need your money back before maturity, which is often decades away.

This is because they fucked up their duration risk handling, no one held a gun to SVBs head and forced them to invest so heavily in long duration bonds. If they bought more short duration bonds none of this would be a problem. Other banks didn’t make this mistake. So why did they do it? Greed, or who knows, maybe it was plain old ignorance/hubris. Still, it was 100% avoidable. SVB fucked up bad and pointing blame at the fed is just not making them accountable.

Like fuck, who are these dumb chimps running this place. When I help my parents plan their retirement I looked at 100 years of history of what normal fluctuations are in the bond or stock market. Looking up data like that takes an hour and should make anyone with basic math skills realize rate and the bond fluctuations that follow at 100% bound to happen given enough time. It was just absolute amateur hour at SVB.

[newaccount2023]

exactly! even more amateur single investors know better than to bet the farm on ten year notes when rates are exceptionally low

SVB wasn't even operating at the level of the typical individual investor!

[unyttigfjelltol]

Maybe, but the latest Fed action violates a 40-year downtrend in interest rates, so it was exceptionally improbable from a historical perspective. From 2020 trough to 2022 peak, government interest rates increased almost 1,000%, which means the magnitude also is hard to anticipate or plan for, and the effects extreme from failing to do so. You can do a regression of interest rates from whenever to now, draw a line that's never violated until 2022 (even the 2020 lows were in-trend) and the 2022 highs were almost 50% higher than the trend.

I don't make excuses for dumb decisions, but... someone said recently when the Fed taps the brakes someone always goes through the windshield. This was a far more extreme scenario than tapping the brakes on a car. More like the moon stopped. We all knew it could theoretically, and what do you know, it just stopped.

[dsomers]

> Maybe, but the latest Fed action violates a 40-year downtrend in interest rates, so it was exceptionally improbable from a historical perspective.

imo this is very flawed thinking, a once in a 40 year event is almost 100% likely to happen in an average persons life — maybe twice. I think when it comes to either your life savings or gigantic amounts of money like banks manage it’s irresponsible to not consider economic cycles that even only happen once per 100 years because of how likely it is to happen once is your life and be absolutely devastating.

I stand by my statement, this is 100% on SVBs amateur hour monkey level thinking.

[unyttigfjelltol]

I'm not disagreeing that SVB was wrong, but it's easy for us arm-chair folks to second-guess the wisdom of creating a business model that was doomed in a once-in-40-year-catastrophe. But, "Silicon Valley" was in the name and 40 years was a long time ago, so why not.

[postexitus]

Risk management is not that difficult. Interest Rate exposure is a first order risk that even juniors should be able hedge out properly - it's not some exotic event where correlations went out of whack or something. These guys were either clueless / had no visibility into their balance sheet or outright criminal.

[Mezzie]

I'm a child of a bookkeeper with no econ training under my belt and I know about interest rate exposure and risk hedging. And there have been discussions since 2008 about how long QE and low interest rates could last, it's not like that was a new question.

[dsomers]

Exactly

[itsoktocry]

>World order is still fluctuating all as a result of these betrayals.

It seems a pretty basic assumption that someone at SVB did a WHAT IF analysis to game out different scenarios if interest rates started rising. This is not rocket science.

>You can do a regression of interest rates from whenever to now, draw a line that's never violated until 2022 (even the 2020 lows were in-trend) and the 2022 highs were almost 50% higher than the trend

Well, if your bank is relying on analysis like this, it deserved to fail.

Here's how easy this situation was:

"Historically, would 5% interest rates be seen as crazy? What happens to our balance sheet if we see those rates?"

[qwytw]

> almost 1,000%

Well 0.1 -> 1.0 is a 900% increase, want to 1.0 -> 4.0 is "just" 300% so I'm not sure if this is a very good way to measures increases in interest rates...

> regression of interest rates from whenever to now, draw a line that's never violated until

Given the data (https://www.macrotrends.net/2015/fed-funds-rate-historical-c...) I would expected such line to be nearly useless due to a very low R2 (too lazy to actually calculate it...)

But yeah, even back in last March the bond markets were not expecting that the rates will be above 2.75% in 12 months.

[huevosabio]

I don't think anyone in such position would do a simple regression to guess interests.

In 2020 and 2021, we already had tons of talk of upcoming inflation due to the stimulus pacjages. I myself took some precautions, and I am really not at all savvy.

The management team of a major bank would certainly be aware of the risk of inflation given the massive uptick in money supply and that the fed would be forced to raise interests as soon as the inflation manifests.

I think incompetence is an easier explanation. Years of ZIRP bubbled incompetent or lazy people to the top.

[newaccount2023]

> but the latest Fed action violates a 40-year downtrend in interest rates

its no big deal to make a little bet based on that, but SVB seems to have bet the farm

had they even diversified their duration they would have been fine

[fastaguy88]

It really didn't take a finance expert to appreciate that when interest rates ranged from less than 1% short term to barely 2% at 10 years, interest rates could only go up, 40 year trend or not. And being invested long, for very little gain, was extremely dangerous should rates go up. It has been clear since the 2008 crash that longer bonds had far more downside than upside when rates were so close to zero, so why take the risk?