Financial contagion: Why some market crashes spread, and others remain isolated
Or why the crypto crash didn’t matter for anything expect crypto.
In late 2021, crypto’s speculative fervor began to wane, and the market took on the appearance of a slow-motion car crash. Since that time, nearly $2 trillion of crypto wealth has been destroyed, with the market falling by over 60% from its peak. Despite this violent reversal in crypto’s fortunes, the crash barely registered with broader financial markets. While the Fed’s rate hikes meant that risk-off sentiment was common to all markets in 2022, the turmoil in stock and bond markets was detached from the contemporaneous crypto blowup. Crypto’s relative isolation meant that the crash just wasn’t contagious.
Traditional financial markets are linked through debt networks, where collateral mechanics mean that rapid price declines in one market can infect another. In 2008, dealers used mortgage-backed securities (MBS) as collateral to secure overnight loans to finance their inventory of stocks, bonds, and derivatives. When home prices declined and investors soured on MBS, dealers found themselves unable to secure new financing, leading them to dump assets at fire-sale prices. Eventually, the Fed had to step in and swap Treasuries for MBS, stabilizing the market by ensuring dealers had prime collateral to borrow against.
Through this obscure link, the value of a wide range of financial assets ultimately depended on home prices. Crypto markets provide a stark contrast. Since most coins are widely understood to be speculative instruments, few lenders allow them as collateral for anything expect borrowing to purchase more crypto (that is, margin lending).1 Purchasing Ethereum on margin secured by Bitcoin as collateral offers a mechanism for contagion to spread between coins, but not beyond them.
Just like crypto debt networks are separate from traditional finance, crypto capital pools are separate from traditional capital pools. When crypto crashes and speculators race for liquidity, they are pulling funds that get allocated solely to crypto, not to other markets. As a concrete counterexample to crypto capital isolation, consider the 1994 Mexican currency crisis, when linked capital pools helped propagate turmoil around Latin America. As crisis took hold in Mexico, mutual funds faced investor withdrawals and banks sought to reduce their risk exposure to the region. With Mexican markets locking up, these capital pools sold related assets to raise liquidity, perversely punishing countries with the deepest markets, such as Brazil. This capital flight sparked its own crises in countries with few fundamental links to the original peso devaluation.
Farther back in time, but no less instructive, is the way the call-money market of New York spread the 1929 stock market crash to commodities. Through the call-money market, investors would make short-term loans to finance commodity imports, inventory storage, and margin loans for stocks. With the market crash sparking a scramble for cash and fears about margin loans going bust, investors pulled funds from the call-money market.2 With credit markets frozen, commodity inventories were liquidated en masse, amplifying existing disastrous deflation trends.3
By 1987, markets had matured to separate margin financing from commodity financing, meaning that a stock market crash had no mechanical impact on commodity prices. Similarly, since the crypto capital pool is separate from nearly every aspect of traditional finance, the pullback in crypto liquidity had no knock-on effects in other markets.
Finally, there are few institutions with meaningful exposure to both traditional markets and crypto markets. There are crypto institutions and traditional institutions, but the two don’t mix well, not the least of which because of regulatory issues. Crises often propagate through the balance sheets of firms operating in multiple markets. Look to the collapse of Long-Term Capital Management in 1998 to see how a Russian bond default could end up threatening a fire sale in American capital markets.
The institution most likely at risk of serving as a link to spread a crypto crisis is Tether, a stablecoin backed by large amounts of traditional financial assets. A mass redemption of Tether could cause the firm to dump these assets and drive their price down, although this fear has been somewhat allayed by Tether’s claims to have eliminated commercial paper from its reserves. Regardless, Tether’s size and the firm’s awkward relationship with both auditors and regulators means that contagion concerns are warranted.
Financial contagion spreads when a race for liquidity causes prices to fall past a critical point where that scramble spreads to a distinct market through a transmission link. While the crypto crash featured many critical points, few links existed through which the collapse could spread to traditional finance. Viewed through this lens, financial globalization is not an unqualified benefit. International institutions, debt networks, and capital pools have increased the flow of funds to developing countries, but have also made it easier for crises to cross borders and markets. Ultimately, understanding financial contagion is about understanding markets at their extremes, and not just at their averages.
1: ‘Backed’ stablecoins like UDSC and Tether don’t really meet the definition of speculative instruments, and so are more suitable for collateral-based lending due to low volatility, but they are still not generally used as collateral for purchases made ‘off-chain’ (that is, in markets other than crypto). Useful collateral-based lending tends to be undercollateralized, but repayment therefore involves property rights enforceable through the courts, something incompatible with code-is-law philosophies. Projects which attempt to bridge this gap do exist.
2: At the time, many margin loans were offered with as little as 10% of the purchase put down in cash, making the investments highly levered and prone to liquidation. Regulation T was passed after the ’29 crash to fix minimum margin requirements for stocks, which is currently set at 50%.
3: See Manias, Panics, and Crashes and The World in Depression 1929-1939 both by Kindleberger for more on how the link between the market crash and the New York call-money market contributed to the Depression.