If Silvergate Capital Corp. fails, it might be the first bank to be brought down by bad liabilities rather than bad assets.
For anyone unfamiliar, this bank has been on a wild ride. It became known as the bank for crypto traders and exchanges after plunging into a sector few other lenders were prepared to go anywhere near. It held cash on behalf of many exchanges, including the now-infamous FTX. It also attracted crypto speculators by promising to settle the dollar side of crypto trades between clients instantly, a simple service it gave a fancy name: the Silvergate Exchange Network.
Silvergate’s deposit base exploded during the pandemic’s crypto boom — to a peak of $14.3 billion at the end of 2021 from $1.8 billion at the end of 2019. As the air rushed out of the crypto bubble, that went into reverse, especially in the fourth quarter of last year after the collapse of FTX. The bank’s shares have been on an equally wild ride.
The bank made some loans against crypto assets and in January it emerged that it faced criminal investigation from the Justice Department over its links to FTX. But the central flaw in its business model was set before this, in my view. Put simply, its deposits weren’t really deposits in the normal sense. Their character was more like the cash float held by a money-transfer business, like MoneyGram International or Western Union.
The only reason Silvergate attracted cash was to settle transactions into and out of a specific set of assets. The purpose of having funds at or passing through Silvergate was to trade that product — if you stopped trading, there was no need to keep any money there.
It is a bit different from a money-transfer business: The cash in Western Union’s float is mostly on its way somewhere. Silvergate’s deposits are likely a bit slower to move than this kind of float, but how much? Let’s think about traditional bank deposits, such as the checking account where you receive your wages and pay your bills. You might move that account to another bank for better service or higher interest, but that’s a pain and we mostly don’t do it. This funding is very predictable for a bank. If you have spare money, you might hunt around for savings accounts that offer slightly higher interest rates. People do this but often quite slowly.
Company treasurers are more motivated to ensure that their business – and shareholders – are getting the best return for their spare cash and efficient services to manage all their payments to employees, suppliers and so on. They’re more likely to hunt for better deals and guard against losing money in bad banks, so their funds are less sticky than retail customers but still fairly reliable. Importantly, all these depositors do a variety of things and are exposed to different parts of the economy – only very, very rarely do all of them suddenly want or need to move most of their money at the same time.
Silvergate only needed interest in crypto to drop off for depositors to start disappearing. It knew this and said as much in its 2021 annual report. “Deposits attributable to digital currency customer investor funds are assigned the highest potential volatility,” it said. At the end of 2021, these amounted to 82% of its deposit base.
The obvious thing to do with highly volatile funding is to put it in the most easily accessible place: Deposits at the Federal Reserve or other banks to start with, plus maybe some extremely liquid, mostly short-term Treasury bills.
Large banks must keep enough high-quality liquid assets to cover 30-days’ worth of deposit withdrawals under stressed conditions. JPMorgan Chase & Co., for example, held $733 billion of such assets at the end of 2022 against $2.3 trillion of diverse, traditional deposits. Three-quarters of those liquid assets were on deposit at the Fed and other banks, the rest was in a mix of Treasuries, and government-sponsored mortgage bonds.
Silvergate took more risk. It plowed most of its new funds into longer-term securities. By the end of the third quarter last year before deposits started to shrink rapidly, only 11% of what should have been its most liquid assets was cash at the Fed and other banks with the rest in securities. Only 11% of those securities were Treasuries. Most of the rest were mortgage-backed bonds with contractual maturities of more than 10 years. Its bond holdings had suffered $1 billion of fair value losses before Silvergate began offloading them to meet withdrawals.
Silvergate was too small to be subject to liquidity rules. It might have met them anyway, but given the extremely rapid growth and common source of its deposits, its regulators should have looked closer before the money started to vanish as quickly as it had arrived.
More From Bloomberg Opinion:
• Matt Levine’s Money Stuff: Silvergate Had a Crypto Bank Run
• Crash Course: Cryptocurrencies Vs. Reality: Timothy L. O’Brien
• The FTX Crypto Victim Card Can Be Hard to Play: Lionel Laurent
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Paul J. Davies is a Bloomberg Opinion columnist covering banking and finance. Previously, he was a reporter for the Wall Street Journal and the Financial Times.
More stories like this are available on bloomberg.com/opinion
©2023 Bloomberg L.P.