In the aftermath of the 2008–09 financial crisis, many people questioned why bankers kept issuing subprime loans even when it was clear they were risky and likely to fail. “When the music stops, in terms of liquidity, things will be complicated. But as long as the music is playing, you have got to get up and dance. We are still dancing,” said Citibank’s then CEO Chuck Prince in 2007.
Did bankers dance because they were overconfident? Because they expected something different to happen than ultimately did? Neither, according to research by the late Chicago Booth PhD candidate Yiran Fan, who would have graduated with his PhD in 2021. Fan’s research explains why bankers’ actions were rational, arguing that Prince and his colleagues knowingly issued low-quality loans to raise liquidity, and that they issued more when there was a possibility of a liquidity crunch.
Fan recognized the risk embedded in the model of a traditional bank, and used it as a jumping off point for his research. Households and businesses deposit money at banks for safekeeping, and banks then lend out those funds for a profit. This works fine unless many depositors suddenly want to withdraw money that has been lent out for, say, 30 years. To manage this risk, banks often utilize the secondary market, where they sell the loans they make to other parties in order to keep liquid assets on hand.
However, Fan’s study points out that trading in the secondary market is somewhat complicated by, for starters, market expectations about what banks are selling. The market presumes that a bank is keeping high-quality loans in house and so bakes in a discount, presuming what’s sold is of lower quality. Additionally, there’s information asymmetry: the bank knows more than potential buyers about the quality of the loans, which lessens its incentive to screen borrowers carefully. Thus, when a bank wants to raise cash to manage its liquidity risk, it’s rational to scrimp on screening, issue low-quality loans, and unload them on the secondary market.
But a bank’s decisions have wider implications and drive, among other things, how much people earn and then save or need to withdraw from the bank. In circular fashion, this affects a bank’s liquidity risk and the amount it trades to raise liquidity.
Fan’s model indicates how this can play out when the economy tightens. Some households are steady savers, but others are what Fan termed runners, prone to withdrawing money, perhaps if they see the economy slowing. If something bad happens, more people will become runners, a scenario that creates liquidity implications for banks, which then issue more low-quality loans to resell.
Fan considered a hypothetical situation in which the economy picks up but then reverts back to its previous state. The growth leads banks to increase their leverage, but following the reversal, higher leverage becomes a liquidity burden, prompting them to issue more low-quality loans, despite the news that prompted them to pull back. “Thus the economy falls into a bust, even though no negative shocks ever happen at all,” wrote Fan.
Ultimately, the study argues, banks issue risky loans to manage their liquidity risk, even if doing so ultimately leads to a destabilizing bust. Unfortunately, what is rational for banks is not necessarily optimal for the banking system and those who rely upon it.