Liz Truss, Silicon Valley, and Holding to Maturity
An explainer on holding to maturity and financial stability
If you’ve been following the headlines lately, you’ve heard that Silicon Valley Bank’s (SVB) collapse was in part due to its hold-to-maturity (HTM) strategy. If you’re really a news addict, you may remember last autumn when pension funds in the UK also got in deep water for holding to maturity long-term British sovereign bonds. That fiasco undermined a government, and SVB’s collapse registered the second-largest bank failure in US history. For financial firms, particularly banks, holding assets to maturity in a world where those asset values are updated in real time can be a dangerous game, because you never know when you might have to sell.
So, what is the HTM strategy and how does it affect financial stability? Here’s an explainer for those of us who aren’t finance experts.
Debt (bond) markets and holding to maturity
Here’s the simplest type of bond example ever: You buy a zero-coupon bond for $95 and in ten years it’ll pay you $100 (zero-coupon means it won’t make payments over time, just once at the end when your principle is returned along with interest). You don’t do anything with that bond; you just sit on it. In ten years, you happily hand over your bond certificate to the issuing company/country/whatever and you get paid $100. Congratulations, you just made $5, a total yield of 5% over ten years (this would be a raw deal and you would aim for 5% per year but for the sake of example, we’ll assume this is what you get). In reality, the bond market is a lot messier. Ignoring the fact, for ease of explanation, that most bonds are not zero-coupon (so they pay out that $5 over time), the other differentiating feature of bond markets is that you can sell your bond. So, you have the same bond that’s going to pay $100 in ten years. You bought it for $95. But, two years in, you decide you don’t want it anymore, so you sell it. Assuming interest rates don’t move much, your bond is likely worth a bit more since you’re now closer to that future repayment and you’ll probably sell it at something like $96. That makes sense, because now someone is getting $4 to hold it for eight years (the clock doesn’t start over—the bond has a definite maturity date). Fine.
But markets aren’t always so rational and things change a lot in ten years. Imagine you bought a sovereign bond from a country, call it the Republic of Narnia. You buy a bond for $95 and it’s going to pay you $100 in ten years. When you buy the bond, things in Narnia are great, and the risk that Narnia doesn’t pay you back in full (credit risk) is low (they have stellar fur exports). Two years in, though, something bad happens (like, I dunno, a war) and Narnia’s credit risk is much more significant—people are afraid Narnia won’t be able to repay. After two years of war, though, Narnia is victorious and has a wonderful manufacturing boom. Finally, year ten arrives and Narnia’s doing just fine, it’s hosting the Olympics. You turn your bond in and you get $100. But if you had sold your bond during the war, you would not have gotten $100; you probably wouldn’t have gotten even $80, because the markets were concerned that Narnia and its money wouldn’t exist in a few weeks’ time. If you’d sold during the war, you would’ve lost money. So, if you weathered the storm and held the thing, you made money. But if you ditched and sold at the wrong time, you lost your shirt.
But perhaps you come to me and say that you have every intention of holding the bond fully to maturity at ten years because your whole investment strategy is about not selling bonds, and you need the money specifically ten years from now, not anytime in between (like, to pay for retirement). Maybe I believe you and that seems reasonable (and less risky than day trading or something). But again, stuff happens. You’re planning to retire, but then you get defrauded in a crypto Ponzi scheme or something and you’re low on cash and need to sell your bond. Now you may be forced into taking a loss.
There’s nothing necessarily wrong with that, and maybe that’s an okay strategy so long as you’re aware of the risk. But the inherent fragility is that you may have to sell even if it’s your intention to hold to maturity, and you might sell for a loss. The need to sell assets for cash (“liquidate” them) and the quality of the market you’re liquidating in aren’t entirely unrelated, though, and that’s the other problem. War breaks out, panic spreads, and you need some cash. But that’s the worst time to be selling an asset, precisely because of the panic that’s driving your appetite for cash.
Marking to market and the value proposition of banks
It’s useful here to take a brief detour to talk about liquidity. Liquidity is basically the ease with which you can buy and sell something. Cash is the most liquid: I can walk around the block and dump cash in return for a candy bar/debit card/deposit at a checking account/whatever very easily. The least liquid would be a partial ownership share in your uncle’s RV: it’s not easy to find a buyer because the market for such assets is very small.
Now let’s say you’re a company and you have this strategy: you’re investing in long-dated (say 20-year) Treasuries (US government bonds), a safe asset (to the extent that the US isn’t going to default . . . we hope). Now, you have to do some accounting and there are really two ways to do this: you can value “at par” or “mark-to-market.” At par means when you reveal your balance sheet to the world, you value your assets at the price you bought them (e.g., you bought $1,000 worth of Treasuries five years ago, so you say they’re worth $1,000). (Technically there’s some accounting distinctions between par values and amortized cost, particularly for non-bond securities, but we won’t get into that.) Marking to market is when you say they’re worth what you could sell them for in the market (e.g., you can sell your Treasuries for $1,100 based on prevailing market conditions, so you say they’re worth $1,100). We mostly report mark-to-market; when you go look at your Robinhood/Schwab/whatever account, that’s what they’re doing: you own five shares valued at $7 per share in the market, so you have $35. Tomorrow if the share price drops to $6, you have $30.
But if your strategy is holding to maturity, then maybe you don’t mark to market because, theoretically, you’re not subject to the volatility of the market: you’re not selling. If you’re in the business of running a dairy farm, that’s probably fine and no one cares because, you know, you can weather the storm—your time horizon is a very long time and you’re not selling the farm. And that makes sense: the value you derive from your dairy farm isn’t informed by fluctuations in land prices, it’s derived from the cash your farm produces when it makes milk and cheese. That’s not to say you might not get into a pinch when bovine medicine prices unexpectedly skyrocket—you might. But you can take out a loan and be “net negative” for a period and survive. The value of your business isn’t that you have stacks of cash on hand to fend off anything at any time. In other words, your value proposition is that you make tasty dairy products, not that you have enough liquid assets to cover any liabilities on demand at any time.
But that is exactly what the value proposition of a bank is. The whole business of a bank rests on credibility. A bank has to have liquid assets enough to cover liabilities that can be withdrawn on demand at any time; the second that depositors begin to doubt a bank’s ability to do that, the clock starts ticking and the bank’s days are numbered. In an oversimplified example, let’s say a bank is on the hook for a million dollars of customer deposits (those are its liabilities, i.e. debts). It’s got a million dollars of Treasuries to back those deposits up (those are its assets, i.e., things it owns). Then something bad happens and—mark-to-market—it takes losses on those Treasuries. Now it has $800K to back up $1 million in deposits. Uh oh. Someone takes a screenshot of the balance sheet and sends it into a group chat called “Bank Run” and it’s in trouble. And investors aren’t idiots: even though maybe you’re not marking to market, investors/depositors know that the real value of your Treasury holdings have declined. In other words, you may not be marking to market, but Treasury prices are public information so your investors probably are considering the market value of your holdings, because it’s their job to think about the worst-case scenario.
SVB and the Truss administration
Bond yields and prices move in opposite directions. If I pay $95 for a $100 bond, that’s a much smaller percentage gain than if I’d paid $80 for it. Let’s say that a country has pretty low interest rates because of economic conditions and some central bank easing due to something scary like . . . a pandemic (?). Now let’s say there’s a bank investing aggressively in long-term Treasury debt and the like. I don’t want to name names, so let’s make one up: Silicon Valley Bank. They’ve got liabilities (deposits from the tech bros of Silicon Valley) and assets (Treasuries and other long-term safe debt plus some other stuff). Now the Fed starts hiking interest rates and the value of their Treasuries falls precipitously and (inexplicably) they’ve taken no significant risk-management strategy against higher interest rates. Mark-to-market, their assets are devalued. Now, they’ll tell you that they’re holding to maturity so they’re not going to take losses because they’re not going to sell. Fair enough. But remember the value proposition of a bank: they’ve gotta have assets to cover liabilities on demand at any time. Someone looks at their balance sheet, freaks out, sends messages to other depositors through the metaverse or something and next thing you know people are pulling money out. Within days they’ve gone from surfing to FDIC receivership. And, importantly, that isn’t irrational: the bank really didn’t have enough assets to cover liabilities and someone spotted it. That’s life (or being a bank at least).
In the UK, things were a little different because it wasn’t a bank in trouble, but the mechanism of the liquidity crunch was the same—long-term asset values fall and your “hold to maturity” thing covering up losses becomes less credible. Some pension funds that were engaged in a “liability-driven investment” strategy held (to maturity) a bunch of long-term UK debt (the British version of Treasuries, called gilts) or derivatives exposed thereto. They’ve got no intention to sell before maturity because they need that money at a particular time (when people retire) and they want low-to-negligible risk. Fine. Then Liz Truss shows up and thinks up a crazy tax policy scheme and shouts about it and the world starts to freak out that the UK is going to be insolvent. In response, gilt prices fall precipitously. Fine, because pension funds aren’t going to sell anyways because they’re holding to maturity. But many of the funds had used leverage (i.e., borrowed money to buy an investment) to buy up the gilts or gilt derivatives. As the (mark-to-market) value of those gilt holdings declined, they were getting collateral calls: when a broker says you need to post more cash collateral to keep your investment position. They burn through their stacks of cash and then have only one way to get more: start selling gilts. As they flood the market with gilts, the prices fall even further, and the asset side of their balance sheet—mark-to-market—looks worse and worse. They’ve got a price-liquidity death spiral. Eventually the Bank of England steps in and Truss steps out and the whole thing is stabilized, but it could have gotten much worse.
It’s easy to confuse the issues with long-term safe assets declining in the short term with an asset liquidity problem but, strictly speaking, it’s distinct from liquidity. Treasuries and gilts are very liquid. The markets for safe government debt are huge. The issue wasn’t that institutions were exposed to illiquid assets whose values fell when rates rose (like fringe commercial real estate in Honduras or something), but just that (very liquid) safe assets were needed way before they were planning to sell them and thus were subject to lots of market volatility when folks woke up to the reality that maybe they really would have to sell.
Lessons?
You can hold stuff for a long time and, if it’s safe (which is a whole different can of worms), you’re probably fine. But you never know when you might need to sell something unless you’re in a business where liquidity really isn’t needed much—like owning a dairy farm. For financial firms, particularly banks, holding to maturity in a world where assets are marked to market can be a dangerous game. Of course, other stuff happens when bad things go down: firms fail to adequately protect against known risk, regulators can be asleep at the wheel, etc. But holding-to-maturity plans unravelling isn’t a new story.
(Disclaimer: This work is independent from and not endorsed by the Yale Program on Financial Stability or Yale University; all views are my own.)