Mark to Mayhem? Silicon Valley Bank and the consequences of aggressive rate hikes.
It is ironic that in the same week that 2-year US Treasury bond yields hit 5% for the first time since the global financial crisis (GFC), the US banking system saw a bank run which resulted in Silicon Valley Bank (ticker SIVB) being taken into receivership by the Federal Deposit Insurance Company (FDIC) on Friday1.
Events like this should be analysed in three parts: what happened, why it happened, and what happens next.
Following a failed equity raise, from Wednesday last week SVB suffered a classic bank run. Around $42b of deposits (a quarter of the total) were pulled in a few days, leaving the bank with fewer liquid assets than liabilities to the tune of $958mm. With the bank unable to liquidate its assets without taking further mark-to-market losses, it would have been unable to meet ongoing depositor demands. The FDIC seems to have been on the case already, closing the bank before market open on Friday.
What happens next for SVB’s investors and creditors? This is apparently the age of the bail-in not the bail-out. In receivership, equity holders will likely receive nothing, or possibly a token amount should the FDIC find another bank to take over SVB’s deposits and assets (think JP Morgan buying Bear Stearns for $2 back in 2008). Depending on what happens next, bond holders may receive some residual recovery when the bank’s assets are sold.
In the US, the federal government guarantees bank deposits up to the value of $250k. SVB, the bank of choice for venture capital firms and cash-rich start ups, clearly had some 95% of deposits excess of this figure. Unless there is a bailout of some sort, then the exact amount these larger depositors will get back depends on the liquidation value of the assets which SVB held against its deposits. In the short term, this could be highly disruptive not only to the depositors involved, but also to market sentiment in general - suddenly corporate bank balances look vulnerable. More generally though, the FDIC is highly experienced and effective at resolving these issues in a timely fashion.
SVB’s assets seem in the most part to have been in ‘safe’ instruments such as US Treasuries and mortgage-backed securities (MBS) - fine, but for the fact these securities have all fallen dramatically in price during the last 12 months as central banks around the world led by the US Federal Reserve embarked on an aggressive rate-hiking cycle in response to high inflation. This makes the ‘why’ of the SVB bank run clearer. What happened to SVB is exactly the same as what happened to the UK pension industry in October last year, and for the same reason.
Back in December (see link below), I suggested that if there were to be a financial crisis, this time around it would have a very different character to the 2007-9 banking crisis. A bank run sounds pretty similar to what happened in the GFC right? Well not quite.
The problems of 2008 in the US banking system were ultimately catalysed by losses from bad debt on institutions with too little in the way of equity and regulatory capital. It was called the credit crunch because there was a problem with credit quality as lending collapsed amid heavy sub-prime loan losses. That is not the case in 2023 - credit has held up remarkably well given how high and fast central banks have hiked rates. Corporations seem for now to be able to roll their debt and pay their ongoing coupon obligations.
Back in October, the UK pension sector suffered a liquidity crisis as UK gilt-edged bond yields rose sharply, in part due to the market taking fright at unfunded spending pledges revealed by the new Liz Truss’ government. The sharp move higher in gilt yields forced pension funds to sell other assets to provide collateral for margin calls on huge derivative positions that they had. These derivatives lost money when rates rose, and the losses needed to be funded, hence the fire sales, the market crisis and the immediate Bank of England Intervention.
While SVB is definitely a bank (ie it takes deposits and lends), its balance sheet makes it look a lot like an asset manager, hence the link between what happened last week and what happened back in October in the UK. SVB took in a huge amount of deposits during the recent tech boom and easy-money era, and much of this was then invested in high-quality but long-duration assets. Low interest rates means low bond coupons and long ‘duration’, where duration is the effective maturity of the average of coupons and principle for a particular bond.
UK pension funds got into trouble by buying receiver swaps - interest rate derivatives that make money with low or falling interest rates but which lose money when interest rates rise. Likewise, SVB’s portfolio of low-coupon treasuries and MBS has exactly the same profile, and it seems that they didn’t hedge their interest rate exposure (ie buying protection against rising interest rates), which seems either short-sighted or naïve, depending on which way you look at it. In bond-market parlance, both UK pension fund and the SVB portfolios display high levels of negative convexity due to their long average durations. In English - big rate hikes mean big losses for both.
Back in December, I suggested that UK pension funds were perfectly rational in their use of derivatives to juice up returns in a low rate environment, and SVB probably falls into the same bracket - they haven’t really done much wrong in terms of risky lending or anything like that, even if their management of interest-rate risk seems a bit village. The thing that has really changed is central-bank rate policy, and it has changed remarkably quickly in the past twelve months, with the Fed in particular hiking at break-neck speed after saying it wouldn’t need to.
The list of Fed mistakes is growing. They moved their inflation target from 2% to a more flexible approach back in 2020. They identified inflation as transitory when it was sticky. They kept at quantitative easing (QE) until March 2021 even with equities and house prices roofing. They have now embarked on the most aggressive hiking cycle in four decades leaving the the US yield curve as heavily inverted as it was in the early 1980s despite rates being nearly 20% back then but still sub-5% now.
The shape of the yield curve matters for banks. Banking is really quite simple - lend more than you borrow and at a higher rate. That usually means borrowing short (this includes customer deposits) and lending long with an upward-sloping yield curve (higher rates at longer maturities). Inverted yield curves, where short rates are higher than long rates, clearly don’t work for banks. The inverted yield curve ended Northern Rock in the UK in 2007, and its done for SVB in the US in 2023. Both suffered a liquidity crisis due to a duration mis-match to the backdrop of an inverted yield curve.
While the Rock went insolvent because of short-term wholesale funding problems, SVB went bust because short-term rates had crippled the price of its assets. For much of the last decade, banks haven’t had to fight for deposits because interest rates were at zero. With the Fed Funds’ rate heading to 5%, suddenly there is an alternative for depositors in the form of T-bills, but the problem banks have is that their assets were acquired when rates were low. Paying 5% when your asset base yields 2% say is clearly going to squeeze profits or become outright loss-making.
The banking sector can help itself by raising deposit rates to attract and retain more deposits, although for smaller banks this is going to be tough, not just in terms of profitability but more with respect to short-term issues such as confidence. On Friday, the dreaded contagion appeared in the US banking sector, with First Republic (FRC) down 53%, PacWest (PACW) down 37% and Western Alliance (WAL) down 39% on the lows, although all three stocks bounced into the close.
It is worth reiterating that the really big banks are in much better shape than in 2007-9. The global-systematically important banks (G-Sib’s) are far better capitalised now, and if anything have been turning deposits away. Despite frenetic trading in the sector on Friday, JP Morgan (JPM) actually closed up 2.5% on the day, suggesting the pain will not be felt evenly. Smaller banks with less sticky deposits and possibly less-sophisticated hedging strategies may well be in line for further market examination, and this is where the regulators need to be very wary.
While the graph above shows US bank reserves having fallen with Fed rate hikes and balance-sheet reduction (quantitative tightening or QT), the fall has steadied since the middle of last year despite ongoing Fed QT. Given the weakness in SVB and the market’s focus on the share prices of other smaller and more regionally-focused banks, it may well be that the pressure is not being felt equally across the banking sector.
If deposits are becoming scarce and banks are struggling to lift deposit rates because the asset-side of their balance sheet yields much less than current interest rates, then clearly this issue of banks potentially having to sell assets at depressed prices to match deposit withdrawals and taking losses as a result is a problem not unique to SVB. If that is the case, then the Fed and other bank regulatory authorities need to be right on top of this and aggressively so, otherwise things might start to get very sporty indeed.
What happens next? Having topped 5% as Fed Chair Jay Powell kept up the aggressive inflation-fighting rhetoric during his testimony in front of Congress, 2-year US Treasuries finished the week some 45 bps lower at 4.6%. February non-farm payrolls came in at spikey 311k (vs 205k expectations), so employment is not rolling over in a way which might give the Fed room to pause or even cut rates.
What is interesting is that the fall in Treasury yields was accompanied by a fall in the dollar, with the DXY dollar index down 0.64% on Friday. The usual flight-to-safety trade sees Treasury yields down and the dollar up. Given the current scare is in the banking sector, it was in fact gold, the one asset class with no counterparty risk, that spiked up 2% on the day. This difference alone suggests something is afoot.
The speed of rate hikes and their magnitude relative both to the starting point (zero) and the length of time they were at those levels (15 years or so) means that there must be a lot of institutions which are now deeply under water on a mark-to-market basis as the value of interest-bearing securities has fallen in the past 12 months. This is true not just of pension funds (like those in the UK) or banks (SVB being an egregious example), but of central banks too - the Swiss National Bank is for example sitting on 2022 losses of some $143b2. Central banks can theoretically recognise these losses at their leisure, but as the SVB story shows, banks have no such luxury.
The Fed’s policy is to hike rates on a data-dependent path but then to hold rates higher for longer. The market has woken up to this in the past month, and it perhaps not surprising that some of the most duration-exposed lenders like Silicon Valley Bank have started to suffer.
The September-October 2022 period marked the first hiatus in the central-bank tightening cycle as the Japanese government bond market and the UK gilt market became unstuck. It feels like a second hiatus is about to be reached, with the Fed potentially having to react to burgeoning contagion. The next week or two will be critical. If deposit flight starts to force more asset liquidation and loss recognition in the banking sector, then the logic of the Fed’s higher-for-longer policy will be called into question, with all the implications for inflation expectations that that entails.
In addition, the Fed may be forced to add liquidity, either from the standing repo facility or other emergency programmes to bolster confidence in the system. The Fed’s balance-sheet run off may have to end prematurely. With bank runs, there is no reason for them to happen other than panic, and it is thus confidence which is now key. The US banking system is in good shape overall and credit risk appears low for this stage in the cycle (it may well rise in a recession though).
There is no reason for contagion to spread, but then again, in a bank run, the most rational thing to do is to get your money out first. We may be about to learn that in trying to correct one mistake (believing inflation was transitory), the Fed has committed a worse one (hiking rates too far and too fast). The inverted yield curve has been screaming problems for some time now, even if the exact meaning of the inversion is different this time.
SVB has probably not done much wrong, even if some of the recent C-suite share sales make it look like insiders knew trouble was brewing, particularly given an attempted capital raise followed share sales by the CEO and CFO3. That there are other lending or shadow-lending institutions in a similar predicament due to the fall in asset prices resulting from aggressive central-bank rate policies may mean that this story isn't over and the war on inflation is about to take another, unexpected turn.
FT Journalists, Silicon Valley Bank shut down by US banking regulators, Financial Times, 10/03/2008,
John Revill, Swiss National Bank posts record $143 billion loss in 2022, Reuters, 09/01/2023.
Austin Weinstein / Bloomberg, Silicon Valley Bank’s CEO sold $3.6 million of stock in potentially ‘problematic’ transaction days before historic bank failure, fortune.com, 11/03/2023.