This is what happens when you run a carry trade and they invert the yield curve

‍

SVB was a poster child of the old New Normal and its demise signals the rise of the new New Normal. A bank that, often in return for helping fund capital rounds, took large corporate deposits from Silicon Valley startups who had raised money via VCs, it mainly 'invested' its capital in carry trades, exploiting an ultra low cost of funding and an upward sloping yield curve. Because these were usually 'risk free' in terms of being government bonds, it overlooked that fact that it was taking risk through duration risk - both in terms of a liquidity mis-match and a mark to market valuation risk. Thus, by essentially focusing on 'credit', it ignored the actual business risk implicit in its structure; that it was heavily negatively exposed to rising short rates and especially to an inverted yield curve. The new New Normal is about normal companies making normal profits from normal interest rates. SVB was the opposite of this, it only made money in a world where interest rates were the wrong level and the real surprise is that anyone is surprised what just happened.

The crisis at Silicon Valley Bank is causing shockwaves throughout markets, not so much because people didn't see it coming (some certainly did, including Hedge funds, who having been selling US regional banks for nine straight weeks and the management themselves also having been unloading large amounts of stock) but more because, as with the 2008 Financial Crisis, there was an expectation that 'the grown ups were already in the room'. Once again, by layering thousands of pages of regulation and employing thousands of people to ensure compliance, government has diluted incentives for individual market participants to make their own risk assessment. Just as some of the most sophisticated Private Equity investors seemingly failed to do proper due diligence on Sam Bankman Fried, so many of Silicon Valley's smartest minds failed to do due diligence on SVB.

Having most of your corporate cash with a single institution is not exactly best practice risk management, but this is not to let the regulators off the hook, since it was well known that SVB's deposit base was almost entirely corporate rather than retail and thus, being uninsured, was actually less 'stable'. Indeed, this was actually a central part of its business model; it would come in with the debt portion of a VC raising on the basis it got the corporate account. Certainly it was not offering a notably higher return on deposits. Regulators knew this, but for some reason did not recognise the greater 'flight risk' from large corporate treasury accounts rather than individual, deposit insured, retail accounts. JP Morgan estimate that $152bn out of $173bn of SVBs deposits were not insured. The regulators also failed to acknowledge the fact that the asset base was loans but mainly fixed income securities and that, as with the LTI products in the UK last September, the increase in interest rates by the Fed was going to cause significant mark to market losses in those security portfolios. If they hadn't noticed it themselves, they could easily have read the research by JP Morgan Asset Management in q4 2022, referenced here on the Grumpy Economist Blog. Looking at the chart of the tier 1 capital ratio after adjusting for unrealised losses and it is pretty easy to spot which bank has the problem.

However, it is important to recognise that, thanks to aggressive lobbying, SVB, along with a number of other, smaller banks, was not subject to the same stringency as the bigger banks, even though it was one of the fastest growing. This of course is all part of the fall out from the more than decade long policy of QE as it collided with the equally foolish policy of Zero Covid. While deposits at US Banks rose by $5.4trn between q4 2019 and q2 2022, less than 20% was actually lent out, the rest went into fixed income securities. In effect banks were borrowing at approximately zero and lending to the US government for a nice 'carry'. This of course is banking 101, borrow short to lend long and while there is nothing intrinsically 'wrong' with this, in many senses it is what banks are supposed to do, 'duration mismatch' is a key risk that needs monitoring and it doesn't appear that SVB were doing this. The fact that they didn't have a chief risk officer for much of last year is being cited as a yellow, if not red, flag, but here again we have to ask about the regulator's role and the fact that it appears that hedging interest rate risk was regarded as too expensive (ie less profitable).

Carry trade collapses and so does the balance sheet

Then of course rates started rising and not only did the cost of deposits start to rise, but the return spread on assets started to fall and as the yield curve inverted, so did the profitability; the 'cash flow' side of the business collapsed. Meanwhile, the assets held on the balance sheet at 'Hold to Maturity' (HTM) - here JPMAM estimate that SVBs book of HTM assets grew from $14bn in 2019 to $99bn by 2022 - started to become a problem. One would have hoped that someone somewhere had some proper models on Duration and convexity of that bond portfolio, but even a crude model of duration would have signaled large losses as the short end headed towards 4% and the accounts suggest it was of the order of $15bn. Just as with the UK LDI issue, if you basically get a margin call and have to sell the HTM asset before maturity, you have to realise a loss and that has implications for the rest of the book, as well as your Equity.

Bottom line.

This was not a credit problem, or even technically a liquidity problem, but an asset liability mismatch resulting from a business model no longer fit for purpose and a lack of proper oversight. SVB had a business model that was funded by on demand corporate deposits from startups that had already received VC funding and like most of the US banking system had put those deposits to work, not in lending to companies, but in lending to government and agencies, via purchasing longer dated fixed income securities. In what was likely a 'reach for yield' ( a classic negative bi-product of zero interest rate policies) they started 'investing' further out the curve, delivering a classic duration mismatch, which they clearly did not hedge. A sharp rise in short rates delivered a hit to asset valuations via simple 'bond math', which has only now been recognised and essentially should/would have wiped out all their equity, which is why they tried and failed to raise more capital, thus triggering the 'bank run'. Like a lot of accounting 'tricks' such as non GAAP accounting, investors should and need to look through this failure to mark to market.

Systematically, the FDIC estimated the US Banks had mark to market losses of $650bn by q4 2022, split between HDM securities and what are known as available for sale or AFS securities and even though  successful lobbying has largely removed the 'smaller banks' like SVB from the necessary regulatory oversight, it is certain that a lot more oversight is coming. It also means that central banks will, thankfully, start to think about the speed and extent of tightening interest rates. As we have noted, the inflation came from the expansion in the money supply during Covid boosting demand at the same time as Covid was causing a contraction in the supply chain, it was not due to ultra low rates. They had caused the inflation in the financial markets and it is the deflation in those markets (remember the $18trn in negative yielding bonds?) that is now washing through the financial sector. Raising rates aggressively will not 'stop inflation', but it will make these sorts of problems worse. After a decade of strict regulation and with hard lessons learned, the European banks look far less exposed to these stresses. The Bank of Japan in particular must be looking closely at the events in California.