Here’s how ‘duration risk’ came back to bite Silicon Valley Bank and led to its rapid collapse
Following the fall of Silicon Valley Bank, a lot of terms are being thrown around on CNBC and elsewhere in discussions about what went wrong. One key term is “duration risk” along the yield curve in the bond market. We don’t usually get into this level of detail on fixed income at the Club — but in this case, it’s important to understanding the second-biggest bank collapse in U.S. history. The now-failed institution, which served tech start-ups and venture capital firms for over four decades, got caught in longer-duration U.S. Treasury bonds. When there was a run on the bank late last week, SVB had to sell these securities at steep losses to raise funds quickly for its customers. In the end, the rush of clients demanding to withdraw money from SVB led to U.S. regulators stepping in to protect depositors in order to prevent contagion in the banking sector. Here’s a guide to the dynamics that led to SVB’s demise. How a bank operates First, let’s get a few key basics out of the way. Deposits at a bank are held as a liability on the balance sheet. The bank takes in deposits and is therefore on the hook when the depositor requests a withdrawal. The bank also pays out interest on these deposits. A bank is also in the business of making money and needs to generate at least enough money on those deposits to pay off the interest. The deposited money can’t just sit in cash. In order to generate a profit and more than cover the interest owed on deposits, a bank will take that money and lend it at a higher rate. As these loans generate interest paid to the bank, they are considered assets. A bank makes money on the spread, or net interest margin (NIM), between what it’s paying in interest to depositors and what it’s generating in interest from loans and other investments. The money made is referred to as net interest income (NII). The lending banks engage in can take many forms — from lines of credit to mortgages to car loans. Another option for a bank that doesn’t have much demand for consumer or business loans is to buy securities such as U.S. Treasurys, bonds backed by the full faith and credit of the U.S. That’s what SVB did: It took the deposits and bought up a bunch of Treasury bonds, which requires holding the notes for the term to get your money back or sell them at the market price. Those Treasury prices, which move in the opposite direction of yields, could be worth less than the purchase price. And in SVB’s case, they were worth much less. Duration risk in bonds Those Treasury purchases in and of themselves were not the issue at SVB. The problem occurred when depositors came calling for their money and the bank didn’t have the cash on hand. So, it had to sell Treasurys of longer durations that hadn’t matured yet and were underwater in price. That’s because the Federal Reserve’s steady campaign of policy interest rate hikes to fight inflation pushed bond yields to multiyear highs. (Ironically, bond yields have come down since the banking chaos, which sent bond prices higher). A deposit withdrawal can happen at any time. There’s essentially no duration on a deposit; the money is expected to be 100% accessible at all times since it’s considered a cash balance. The value of customer accounts does not fluctuate with the market. On the other hand, the market value of the investments the bank makes with those deposits can fluctuate greatly between the time of the initial investment and the maturity date. Even Treasury notes can see their value on paper fluctuate greatly prior to maturation. This mismatch, which always exists to some extent, is where “duration risk” comes into play. Without getting too much into the weeds, the market value of a bond falls, in percentage points, by whatever its duration is for every 1% increase in rates. In other words, a bond is expected to drop in price by its duration multiplied by the percentage change in rates. For example, a bond portfolio with an average duration of five years would be expected to fall 5% for every 1 percentage point increase in rates. Should rates rise 2 percentage points, that portfolio would be expected to fall 10% given its duration. An average duration of 10 years would see the portfolio fall by 10% for every 1 percentage point increase in rates and 20% in the event of a 2 percentage point increase in rates, and so on. These moves are not exact in the real world and the timing of cash flows can impact durations. But this is a good rough guide. As rates and bond yields rise, the value of the deposits (liabilities) is unchanged. However, the value of the bank’s investments (assets) can fall dramatically. The nature of deposits means that the liability can be called in at any time. The assets, on the other hand, need time to recover. They will recover but how long that takes depends on their duration and the interest rate environment. It’s usually not a problem if the plan is to hold the asset until maturity, as any losses between now and then are only paper losses. In the end, at the maturity date, you receive 100% of your investment back and made whatever the interest rate was at the time of purchase. However, should you need to sell those bonds before maturation, in order to meet liquidity needs – like a lot of depositors banging on the door asking for their money back – then you’ve got a real issue. Those losses on paper must be realized in order to convert the bonds back to cash and fulfill withdrawals. An experienced risk management team should have hedges in place to protect against this known possibility. This did not happen at SVB. The risk is that the duration of the investments made by the bank doesn’t match up with its potential liquidity needs. Silicon Valley Bank reached too far out on the yield curve in search of higher yield. Put another way, they tied up the deposits in bonds with longer durations than appropriate. The bank also failed to adequately hedge the risk posed by a rise in rates. Doing so would have ensured that SVB had the ability to ride out any declines seen on paper between the purchase of these assets and their maturation. In fact, according to SVB’s fourth-quarter release, the portfolio duration of its fixed-income securities was 5.6 years and the hedge-adjusted duration was also 5.6 years. A proper hedge would have shortened that duration. There was effectively no hedge whatsoever. What caused the run on SVB While SVB did not manage its “duration risk” properly as rates rose, the bank also miscalculated how much the Fed’s tightening cycle would hurt the very start-up companies that were its clientele. The central bank’s battle against relentlessly high inflation led to initial public offerings (IPOs) slowing significantly. As a result, SVB’s clients — short of raising more venture money, which has been harder to come by in the current environment — were forced to use their deposits to run their businesses. Since start-ups don’t generally make profits, they burn through cash in hopes of one day going public — the ultimate exit and liquidity event. That IPO endgame has been delayed for many of these companies. All this really started last Wednesday when SVB published a mid-quarter update saying that in order to strengthen the bank’s financial position (keep in mind any wording on a release like this is going to be through rose-colored glasses to some extent) management took actions including the sale of “substantially all” available for sale securities and announced a capital raise via the sale of common equity and mandatory convertible preferred shares. As a result of these actions, the bank realized a one-time, post-tax earnings loss of approximately $1.8 billion. Now, that may have been OK. However, the one thing no depositor, especially a tech startup that needs cash to run operations and make payroll, wants to hear is that the bank holding their money is being forced to take action due to a need to “strengthen balance sheet liquidity.” Any depositor who does hear that is understandably going to want to move their funds to a place where they feel more secure. With all available-for-sale securities sold, the bank would have to turn to securities it intended to hold to maturity, which it stands to reason had longer durations and therefore even greater losses on paper. To this point, according to the mid-quarter update, the average duration of the securities sold was 3.6 years, well below the 5.6-year duration of the total portfolio. As a result, there simply wasn’t enough market value to fulfill the redemption requests. That’s how a 40-year institution can crumble in a matter of days, leaving federal regulators to try to clean up the mess and prevent it from spreading. That’s what the Fed and the Treasury Department did on Sunday evening when they said all SVB depositors (and those at another failed bank Signature ) would be made whole. On Monday morning, President Joe Biden spoke about the bank failures saying the government backing of depositors would not cost taxpayers anything. “Instead the money will come from the fees that banks pay into the Deposit Insurance Fund.” Biden also made it clear that “investors in the banks will not be protected” because they took the risk. “That’s how capitalism works,” he added. Additionally, he said the entire banking system is sound, and the work done after the 2008 financial crisis to make it so no banks are too-big-to-fail worked. (See here for a full list of the stocks in Jim Cramer’s Charitable Trust.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
A man passes a sign Silicon Valley Banks headquarters in Santa Clara, California, on March 13, 2023.
Noah Berger | Afp | Getty Images
Following the fall of Silicon Valley Bank, a lot of terms are being thrown around on CNBC and elsewhere in discussions about what went wrong. One key term is “duration risk” along the yield curve in the bond market. We don’t usually get into this level of detail on fixed income at the Club — but in this case, it’s important to understanding the second-biggest bank collapse in U.S. history.