[Goose]

This is obviously a complex situation and there were many factors involved.

Totally agree.

However, it is ludicrous to claim that SVB was hoodwinked by the Fed into buying too many long-term bonds.  This is one of the most sophisticated bank/investment houses in the country.  (Banks aren't simple banks having to cover deposits anymore of course, but that is a different discussion.)  They would have been fine if they were buying long-term bonds to hold until their maturity. I am part of a very conservative association that has been rolling 10-year bonds for many, many years with steady, modest gains because the funds used were not part of our liquidity.  SVB was instead, putting too high of a percentage of their assets in non-liquid investments to cover unexpected withdrawals.

They absolutely were not hoodwinked. However, I think they were buying bonds that they had every intention of holding to maturity (unless by some chance they could make a larger profit selling them sooner). They did not intend to use them as part of their liquidity. The fact they had to do so was not planned. That was the problem.

SVB could claim that they expected to be able to take such risks because they were never prevented from doing so before, but they knew they were taking risks.

IMO they would not be claiming they could take such a risk because "they were not prevented from doing so before". They would claim that for their bank it was not a "significant" risk because they would continually have substantial deposits coming in from IPOs and that these new deposits would mostly remain in the bank for quite a while (years) because the companies wouldn't spend the money that fast. The obvious question would be "What if the IPOs stop?". The very smart people at SVB didn't think they would. That is the second piece of the puzzle that can't be ignored. For a very long time money kept flowing in for reasons that nobody really "understood". There were crazy piles of cash being raised that had to land somewhere, and SVB had it rigged that much of it would land in their bank. Therefore, they just assumed it would continue to do so.

Whether they would or would not have passed a stress test is a separate question from their culpability for their investment decisions.  Getting away with it is not the same as being tricked into doing it.

Agree, and I don't think anybody was "tricked". However, the "stress test" question is IMO very important. It seems their is a significant body of opinion that if SVB had been subject to the "stress test", this problem would have been detected. It seems based on what the OP presented this would not have been the case because the parameters of the "stress test" used in the relevant time intervals did not adequately mirror what actually occurred. It certainly would not have detected it early enough to "undo" lots of one year maturities that were already locked in.

This is an inherent limitation of such tests, in that they test outcomes based on what the "experts" think are "reasonable" predictions about future trends. The experts are essentially always "wrong" to some degree and the exact parameters tested can be critically important. It is also impossible to simulate the effects of human perception. What actually starts a "run" on a bank? There are lots of human factors. If there was a reliable way to "test" what will actually happen, banking regulation would be easy. That doesn't mean "stress tests" are useless, but it does mean they have significant limitations.

[Goose]

This is obviously a complex situation and there were many factors involved.

Totally agree.

However, it is ludicrous to claim that SVB was hoodwinked by the Fed into buying too many long-term bonds.  This is one of the most sophisticated bank/investment houses in the country.  (Banks aren't simple banks having to cover deposits anymore of course, but that is a different discussion.)  They would have been fine if they were buying long-term bonds to hold until their maturity. I am part of a very conservative association that has been rolling 10-year bonds for many, many years with steady, modest gains because the funds used were not part of our liquidity.  SVB was instead, putting too high of a percentage of their assets in non-liquid investments to cover unexpected withdrawals.

They absolutely were not hoodwinked. However, I think they were buying bonds that they had every intention of holding to maturity (unless by some chance they could make a larger profit selling them sooner). They did not intend to use them as part of their liquidity. The fact they had to do so was not planned. That was the problem.

SVB could claim that they expected to be able to take such risks because they were never prevented from doing so before, but they knew they were taking risks.

IMO they would not be claiming they could take such a risk because "they were not prevented from doing so before". They would claim that for their bank it was not a "significant" risk because they would continually have substantial deposits coming in from IPOs and that these new deposits would mostly remain in the bank for quite a while (years) because the companies wouldn't spend the money that fast. The obvious question would be "What if the IPOs stop?". The very smart people at SVB didn't think they would. That is the second piece of the puzzle that can't be ignored. For a very long time money kept flowing in for reasons that nobody really "understood". There were crazy piles of cash being raised that had to land somewhere, and SVB had it rigged that much of it would land in their bank. Therefore, they just assumed it would continue to do so.

Whether they would or would not have passed a stress test is a separate question from their culpability for their investment decisions.  Getting away with it is not the same as being tricked into doing it.

Agree, and I don't think anybody was "tricked". However, the "stress test" question is IMO very important. It seems their is a significant body of opinion that if SVB had been subject to the "stress test", this problem would have been detected. It seems based on what the OP presented this would not have been the case because the parameters of the "stress test" used in the relevant time intervals did not adequately mirror what actually occurred. It certainly would not have detected it early enough to "undo" lots of one year maturities that were already locked in.

This is an inherent limitation of such tests, in that they test outcomes based on what the "experts" think are "reasonable" predictions about future trends. The experts are essentially always "wrong" to some degree and the exact parameters tested can be critically important. It is also impossible to simulate the effects of human perception. What actually starts a "run" on a bank? There are lots of human factors. If there was a reliable way to "test" what will actually happen, banking regulation would be easy. That doesn't mean "stress tests" are useless, but it does mean they have significant limitations.