With the Summer Academy Risk Management Track under your belt, you could have seen this coming. When Silicon Valley Bank (SVB) went into receivership in March 2023, with $209 billion in total assets (December 2022), this marked one of the largest bank failures in US history, second only to Washington Mutual, which went under in the global financial crisis in 2008.
Perhaps you never heard of SVB because it was not for ordinary mortals. This bank was for private equity royalty and their techy start-ups. In other words, it is a by-invitation-only bank, with 9,500 client accounts serviced by 8,500 staff. The case of SVB is not one of extreme complexity, with off-balance sheet vehicles, re-securitisations and over-the-top credit default swaps. Nope, you could have managed this bank better after joining our half-day session on current issues in Asset Liability Management on Thursday morning.
So, what brought the bank down? Two absolute beginner’s mistakes:
By year-end 2021, inflation and rising rates were so obviously coming it hurt. Indeed, by December 2022, yields on 10-year Treasury Notes were scratching at the 4% mark, up from 1.5% a year earlier. Management believed they could hide the losses from rising interest rates in the held-to-maturity portfolio at amortised cost. Amortised cost means you do not have to mark down the bond assets to their current realisable market values and pass these losses through the income statement. But the losses are real, even if one held the bonds until maturity. This is because SVB would continue to earn only 1.5% interest for the next 10 years while everyone else is earning 3.5% – 4%. The resulting market value losses are massive: a bond with 10 years remaining maturity and a 1.5% coupon trading at par in December 2021, a year later, would have been worth =PV(0.04,10,-1.5)+100/(1.04^10)= $79.73 for $100 face value. Apply this 20% haircut to the $90 billion long-term bond portfolio, and there goes more than the entire $16 billion equity of SVB (December 2022).
What can go wrong? These deposits are cheap and „sticky”.
We’ll just hold the bonds until maturity and continue to pay the depositors a pittance in interest, and there will always be a modest interest margin left for us. Nice try, but core deposit theory only works with a granular deposit base consisting of hundreds of thousands of depositors holding small demand checking and savings accounts and a few larger time deposits. In reality, the bank had not even 10,000 clients, and the deposits were volatile corporate transaction accounts in the millions or rate-sensitive term deposits. More than 92% of the deposit base was in accounts exceeding the $250,000 FDIC insurance limit. Large uninsured deposits are always risk-sensitive and totally not sticky in a crisis of confidence. You don’t have to take my word for these behavioural assumptions about core deposits; they are actually codified in the Liquidity Coverage Ratio (LCR) under the Basel Framework. But as we all found out in March, SVB was exempted from these burdensome regulations by the Trump administration as a non-complex Category IV bank holding company. That worked out great!
All good, then.
SVB was a luxury problem for wealthy tech investors poorly managed by overpaid bank executives. What else is new? Not so fast; this could happen to you. Take the quick self-test: (1) Have interest rates in your main operating currency been going up recently? (2) Does the competitive/behavioural duration of assets exceed that of your liabilities? You may be suffering from SVB syndrome.
Join us in Frankfurt this 3 – 7 July for the Summer Academy to learn more! We’ll show you to manage credit risk, hedge interest rate risk and stress test liquidity like a pro.