Money-market mutual funds (MMMFs) have had a turbulent couple of weeks. On March 18, the Federal Reserve System created a Money Market Mutual Fund Liquidity Facility (MMLF) to “assist money market funds in meeting demands for redemptions by households and other investors.” What’s the source of the trouble?
The story begins with an investor flight, driven by fears of coronavirus-related corporate defaults, out of private debt and into safer and more liquid Treasury debt. Prices of US Treasury bonds have been pushed way up, dramatically lowering their yields. Three-month US Treasury bonds, as of March 27, are yielding only 0.04 percent, down from 1.49 percent a month earlier. MMMFs that hold only government bonds have seen record inflows of funds ($286 billion or 7.3 percent growth in the week ending March 25; nearly 12 percent growth if we add the previous week).
The flight out of private bonds, even short-term AAA-rated bonds, includes large outflows from the mutual funds and exchange-traded funds (ETFs) that hold them. “Investment-grade bond funds experienced a record $38 billion outflow in the week through March 25,” reports Bloomberg News. Some “prime” MMMFs, which mostly hold short-term commercial paper issued by banks, have been hit by outflows of more than 40 percent.
These outflows from prime funds should not be thought of as “runs” in the usual sense. Because mutual funds issue no debts, only equity claims, there is no “me-first” reason to hurry to withdraw from fear of insolvency. A mutual fund can’t become insolvent by suffering asset losses—total claims can’t exceed total assets—when share claims are promptly marked down as soon as, and as much as, portfolio value.
Yes, there was a singular MMMF run in 2008. To summarize what I wrote about it here, the failure of Lehman Brothers drove down the market value of the Reserve Primary Fund's portfolio. At the end of the day of the bankruptcy filing, the Fund neither injected capital nor promptly marked down its share price from its conventional $1 price (it did not "break the buck" until the end of the second business day after Lehman filed). Shareholders saw that redeeming immediately would mean getting $1 per share, but those at the end of the line would likely get less. So they ran. According to a Wall Street Journal news story, on Tuesday afternoon September 16 the Reserve Primary Fund announced that its Net Asset Value had fallen to 97 cents—and that it was ending redemptions to await liquidation of the fund. Eventually the fund was liquidated and the remaining customers received 99.1 cents on the dollar.
Runs of this sort on mutual funds can be prevented by doing what Reserve Primary did not do: immediately marking down the “net asset value” (NAV) of shares, the price at which the fund will buy them back, as soon as asset losses occur. Institutional prime funds do that (they have a “floating NAV”), so the heavy withdrawals in recent weeks were not out of me-first solvency concerns. (For any fund that maintains a fixed $1 share price, marking immediately to market could be replicated by a contractual agreement to reduce each customer’s number of $1 shares to match any fall in the portfolio’s total value.)
Some part of last week’s outflows may, however, be thought of as due to runs of a different sort. They were prompted not by fear of insolvency but by fear of a regulatory barrier. After the Reserve Primary Fund failure, new legal restrictions were imposed on MMMFs in 2010 and 2014 that were supposed to prevent such a failure from happening again. One of these restrictions appears to have unintentionally made the problem worse this time around. This could have been predicted—and was in fact predicted. Here is what I wrote four years ago about that restriction, a 30 percent minimum liquidity ratio (cash or Treasuries to total assets), the breaching of which gives a fund the right to restrict redemptions:
Funds are given the discretion, whether they want it or not (they cannot contractually bind themselves not to use it), "to impose liquidity fees or to suspend redemptions temporarily, also known as 'gate,' if a fund's level of weekly liquid assets falls below a certain threshold." Hanson, Sharfstein, and Sunderam, and many other analysts, rightly reject such "gating rules" as means to diminish runs. Inability to precommit not to impose a gate can be expected to weaken a MMMF. Knowing that a gate might be imposed, especially if another fund has just done so, investors will likely become more anxious to redeem now rather than wait and see.
Which is what we appear to have just experienced. The liquid asset threshold imposed by the 2014 rules is that 30 percent of a prime fund’s assets must be cash or Treasuries. Heavy withdrawals by investors who wanted out of portfolios of corporate debt and into safer Treasuries compelled prime funds to make heavy sales of these liquid assets. As the percentage of liquid assets falls toward the 30% liquid asset requirement, investors rationally fear being trapped behind a “gate,” unable to redeem, or able to redeem only by paying a fee. Again, the restriction is that a fund is blocked from contractually committing not to impose fees or a gate.
It was not to avoid breaking the buck, but to stay above the 30 percent threshold, that Goldman Sachs purchased $1.4 billion in assets from one of its own funds, and $391 million from another. For the same reason Bank of New York Mellon bought “almost $2.2 billion in securities from one of its prime money funds, the Dreyfus Cash Management fund,” reported the Wall Street Journal. Other funds faced similar trouble.
The Federal Reserve thus stepped in not to quell runs of the Reserve Primary Fund sort, but rather to help out funds that were experiencing withdrawals made heavier in part because of the existence of the 30 percent liquidity threshold. The Fed did not want to see the predictable cascade of withdrawals that would follow any fund actually imposing gates or fees. As Cornell Law professor Dan Awrey noted on Twitter, the threshold is “the precise point at which the fund boards were required to consider imposing liquidity fees and redemption gates. The boards clearly stared down the Fed, and the Fed blinked.”
Federal Reserve support is a breeding ground for moral hazard, whether support takes the form of emergency lending to MMMFs at below-market rates or the form of purchasing assets from MMMFs at above-market prices. (If the Fed were not beating the market, the MMMFs would go to the market). To avoid repeating the current problem, end the recently imposed liquidity threshold requirement. The problem of a Reserve Primary-type failure can be alleviated more elegantly by greater rather than lesser freedom of contract, namely by a floating NAV or, for an MMMF with a fixed $1 share price, allowing a contractual share-reduction mechanism (applied strictly, without “penny-rounding”) to replicate the run-proofing property of a floating NAV.