[rogpodge]

Yes, that was my point, that the 10-year was still being touted by the Treasury in 2021 as a safe investment. Why? Because it was in their own best interests.

https://twitter.com/WallStreetSilv/statu...8886995969

https://twitter.com/balajis/status/1635228632275877888

Maybe a 2022 stress test detects the problem, but SVB's risk management department was going through a transition, AND the bonds hadn't been devalued completely at that time.  It would have required timing of the stress test in Q4 (when the Q3 declarations would be most evident see above chart). At least one quant doesn't think the stress test would have made a difference. 

https://twitter.com/StanFeldman2/status/...7961300994

The Fed's 2022 stress test scenario assumed interest rates would stay below 3%.

https://www.federalreserve.gov/publicati...narios.htm

Based on the 2022 balance sheet and the Fed's 2022 stress test scenario, would the issue have been detected? Probably not. Most of the devaluation occurred in the back half of 2022. By then, I doubt SVB could have found counter-parties willing to take on that interest rate risk (as shown by how few parties were willing to buy the devalued assets and the failed auction over the weekend). 

https://twitter.com/balajis/status/1635308195282583559

Let's pretend that Barney Frank of Dodd-Frank was not on the Board of Directors for Signature Bank. Yes, SVB was mismanaged and didn't properly hedge interest rate risk. But they were significantly suckered in by the Fed and the Treasury (and the general mismanagement of inflation and the economy). We'll see tomorrow whether the pause bond traders are pricing in will be viable, or if CPI will force the Fed's hand.

---

a large amount of 1.25% yielding Treasuries is now worthless and can't be sold. 

Not so. The Treasuries market is enormous and among the most liquid there is, if not the most. SVB's Treasuries (and other bonds) can be sold for a value low enough to generate a yield-to-maturity equal to the current yield on a similar bond of equivalent time to maturity. The problem is, selling those Treasuries won't generate the needed amount of cash because their value has decreased relative to when they were bought. By how much? Plugging numbers into an online bond value calculator, if SVB bought a 10-year Treasury when 10-year Treasury interest rates were at their low in 2020 (~0.55%), today that bond would have lost very roughly about 25% of its value. That's more than enough to fuel a liquidity crisis, but it ain't 100%. But SVB might have bought bonds with longer terms than 10 years (I don't see any published description of their actual portfolio's characteristics) in which case they could have lost a lot more than that (but still not 100%).

Maybe in general terms, in a normally functioning market, but SVB reported that they sold all available securities for sale.  This is what they yielded on a $122B bond buy.  Once these losses were realized, they needed to raise more capital, because they didn't have anything more to sell (that people were willing to buy). 

On March 8, Silicon Valley Bank parent SVB Financial Group said that it was taking "strategic actions," including selling almost all of its available-for-sale securities — $21 billion in bonds. It also said it planned to issue stock as part of the plan to raise capital and strengthen its finances.

[rogpodge]

Yes, that was my point, that the 10-year was still being touted by the Treasury in 2021 as a safe investment. Why? Because it was in their own best interests.

https://twitter.com/WallStreetSilv/statu...8886995969

https://twitter.com/balajis/status/1635228632275877888

Maybe a 2022 stress test detects the problem, but SVB's risk management department was going through a transition, AND the bonds hadn't been devalued completely at that time.  It would have required timing of the stress test in Q4 (when the Q3 declarations would be most evident see above chart). At least one quant doesn't think the stress test would have made a difference. 

https://twitter.com/StanFeldman2/status/...7961300994

The Fed's 2022 stress test scenario assumed interest rates would stay below 3%.

https://www.federalreserve.gov/publicati...narios.htm

Based on the 2022 balance sheet and the Fed's 2022 stress test scenario, would the issue have been detected? Probably not. Most of the devaluation occurred in the back half of 2022. By then, I doubt SVB could have found counter-parties willing to take on that interest rate risk (as shown by how few parties were willing to buy the devalued assets and the failed auction over the weekend). 

https://twitter.com/balajis/status/1635308195282583559

Let's pretend that Barney Frank of Dodd-Frank was not on the Board of Directors for Signature Bank. Yes, SVB was mismanaged and didn't properly hedge interest rate risk. But they were significantly suckered in by the Fed and the Treasury (and the general mismanagement of inflation and the economy). We'll see tomorrow whether the pause bond traders are pricing in will be viable, or if CPI will force the Fed's hand.

---

a large amount of 1.25% yielding Treasuries is now worthless and can't be sold. 

Not so. The Treasuries market is enormous and among the most liquid there is, if not the most. SVB's Treasuries (and other bonds) can be sold for a value low enough to generate a yield-to-maturity equal to the current yield on a similar bond of equivalent time to maturity. The problem is, selling those Treasuries won't generate the needed amount of cash because their value has decreased relative to when they were bought. By how much? Plugging numbers into an online bond value calculator, if SVB bought a 10-year Treasury when 10-year Treasury interest rates were at their low in 2020 (~0.55%), today that bond would have lost very roughly about 25% of its value. That's more than enough to fuel a liquidity crisis, but it ain't 100%. But SVB might have bought bonds with longer terms than 10 years (I don't see any published description of their actual portfolio's characteristics) in which case they could have lost a lot more than that (but still not 100%).

Maybe in general terms, in a normally functioning market, but SVB reported that they sold all available securities for sale.  This is what they yielded on a $122B bond buy.  Once these losses were realized, they needed to raise more capital, because they didn't have anything more to sell (that people were willing to buy). 

On March 8, Silicon Valley Bank parent SVB Financial Group said that it was taking "strategic actions," including selling almost all of its available-for-sale securities — $21 billion in bonds. It also said it planned to issue stock as part of the plan to raise capital and strengthen its finances.