The basic deal of a money market fund is that you give it your money, it invests it in stuff, it pays you a little bit of interest, and whenever you want you can take your money back out. (This is the basic deal of a bank account, too.) The basic problem is that, if your money market fund has $1,000 of your money, and $99,000 of other people's money, and invests it in $100,000 worth of stuff, every so often that stuff ends up being worth, like, $98,000. And then the fund has a problem when you go to it and ask for your $1,000 back.
It's a weird empirical fact that this problem rarely occurs, and almost always gets solved somehow when it does. But every once in a long while it doesn't, and people lose money. And they're not supposed to lose money, because a money market fund is where you put the money you don't want to lose. (There are plenty of other places to put the money you want to lose, and they all pay more interest than money market funds.)
So the prospect of losing money is scary. So if you think your money market fund is on the way to being worth only $98,000, you might want to go and demand your $1,000 back right now. They'll probably just give it to you. I mean, there'll still be $97,000 left over. And you don't want to be the guy holding the bag when 98 other people have asked for their money back and there's no money left. But everyone thinks the same way, so everyone takes their money out, so the money market fund has to sell all its stuff, so the price of the stuff drops, so everyone loses money. This is called a "bank run," only it's a money market fund, so it's mostly just called a "run."
Today, the Securities and Exchange Commission approved some rules that basically say, if the stuff loses value, the money market fund can't give you your $1,000 back. That is: If the fund only has $99,900 worth of stuff right now, it has to tell you that, and when you ask for your money back you can only get back $999. (This is called a "floating net asset value.") And if the fund is worried that it'll end up with only, say, $98,000 worth of stuff, then it can impose fees and delays before giving back your money. This all seems sensible enough to a lot of people.
But one important thing to keep in mind is that that last paragraph was a lie. The SEC didn't pass any rules saying that your money market fund has to have a floating net asset value, or that it can impose fees and delays on redemptions. Not if you're a human. If you're a corporation or an institutional investor, then yes, that previous paragraph is true, but if you're a human you can keep on like before. The net asset value doesn't float -- you can still treat your $1,000 investment as, like, $1,000 of "cash" in an "account" -- and no one is going to prevent you from "withdrawing" it.
So now, basically, there's a two-tier in which companies and institutions get to know the cold hard truth about how much their money market investments are worth, but individuals can continue to be coddled in the warm fiction that those investments never lose value.
Why is this? Well here's Mary Jo White:
Retail and government money market funds have not to date faced significant runs even in the worst of times .... At the same time, retail investors in particular have come to rely on the liquidity and stability of money market funds, and they lack investment substitutes with similar characteristics, including those that may be available to institutional investors.
The second sentence there means: Retail investors really like being coddled. And the first sentence means: Retail money market funds don't have to worry about runs, because retail investors don't run, because retail investors are ... the polite term is, "preternaturally calm"? "Dumb" would be the impolite term? But not dumb dumb. Just, like, rational-ignorance dumb. Regular humans spend approximately zero percent of their time monitoring the quality of their money market investments. So they don't tend to run at the first, or second, or seventeenth, sign of trouble. That makes retail money market funds pretty stable.
You have to tell companies how much their money market investments are worth, because they pay attention to that sort of thing, but you don't have to tell individuals, because honestly they couldn't care less.
Here is a good column from Brian Reid of the Investment Company Institute, responding to Bank of England chief economist Andy Haldane's call for greater regulation of systemically risky asset managers. I like this paragraph:
For capital markets regulators, the answer to such problems is to clarify what the risks are and who bears them, so that willing investors can knowingly judge and accept risks. But in the banking world, where investment and systemic risk are intertwined, bank regulators instead resort to microprudential tools to intricately manage bank balance sheet activities and macroprudential tools to manage capital flows and banks' overall levels of risk-taking. Haldane and other central bankers would bring these tools to capital markets. Haldane for example, speculates about central banks taking 'an explicit role in managing the risk-taking cycle and activity in the wider economy.'
Those are two extremely different ways of looking at the world. In one -- following Reid, let's call it the "capital markets view" -- the people whose money is at risk are supposed to decide how much risk they're willing to take, and what's a good risk. This is always hard, because there are agency costs and information asymmetries, but the job of a regulator is to try to give everyone the best information possible, and then hope they make the right decisions.
In the other, call it the "banking view," investors know nothing, and nothing will ever convince them to find out. They don't want to find out. The point of a bank account is that you don't need to know what your bank is doing with your money. The job of the regulators here is rather more involved, and more important. They have to know what the banks are doing with everyone's money, and try to make sure it's all above-board and not too risky. Also they probably have to bail out the banks when things go wrong.
People say that money market funds are part of the "shadow banking" industry. Roughly speaking, "shadow banking" refers to the parts of the world that are regulated according to the capital markets view -- lots of disclosure regulation, much less capital-and-risk regulation -- but treated by their investors more according to the banking view, where the customers try not to think too hard about the risks. You can see why this would be scary: If regulators aren't worrying about the risks (it's not a bank!), and investors aren't worrying about the risks (ehhhh it's a bank!), then ... no one is worrying about the risks? That seems bad.
So the first step, broadly speaking, is to assign bits of the shadow banking world to one world view or the other: Either force investors to think about the risks, or let regulators regulate the risks. But there's no obvious reason that everything has to get the same world view. Some things should be treated like capital markets: Inform investors of the risks, and then leave the investors to take those risks. This seems to be the fate of institutional money market investors, who are understandably grumbly about it. They want their $1,000 to be worth $1,000.
But other things should be treated like banks: Don't make the investors worry about the risks, just leave that to regulators. As a first cut, if something is actually safer when its risks are concealed, then why not conceal those risks? That seems to be the SEC's reasoning on retail money market funds. Why worry retail investors unnecessarily? For them, ignorance can remain bliss.
It rarely occurs because money market funds -- both by desire and by regulation -- invest only in relatively safe, relatively short-term, relatively liquid stuff, so they tend not to lose too much money. And when it does occur, various ad hoc things tend to happen -- the fund's sponsor makes good the losses, or everyone just sits very still until the stuff goes back up to $100,000, or I don't know there's a government bailout -- so everyone can still get all their money back.
To me, anyway, and to Barry Ritholtz here at Bloomberg View, and to Republican SEC Commissioner Daniel Gallagher, who said ""Addressing a three decade old error in a nuanced and tailored manner to reinstate market based pricing should not be seen, as some have argued, as a heavy handed act of government," which is about right. (These changes passed on a 3-2 vote of the SEC, by the way, so I guess some people disagree.)
For "prime" money market funds, that is. Money market funds that just invest in U.S. government securities get a pass, because those can't lose value. (Hahahaha no of course they can, through interest rate fluctuations, but no one worries that much I guess.) Also there's "a two-year transition period to enable both funds and investors time to fully adjust their systems, operations and investing practices."
This hugely exaggerates because there's actually only a narrow band in which the NAV doesn't float. Basically if your account's value drops below $995, then the fund "breaks the buck" and has to tell you that. But if it's worth $996, they can keep telling you that it's worth $1,000. But it's some cushion, and in practice it matters because, again, money market funds don't tend to lose all that much money.
And similarly for banks, which is why retail deposits tend to be an extremely stable funding base for banks, even troubled banks. (Even banks in Cyprus, good lord.) Obviously deposit insurance on small retail deposits helps with that too, though even uninsured deposits are probably more stable than, like, repo.
I said the other day, "The point of banking is to conceal risk." This glib line of thinking draws on much more sophisticated thinking from Arnold Kling, Steve Randy Waldman, Gary Gorton and his co-authors, etc. Also of course there's Bagehot, on the bailing out part.
Reid mentions tri-party repo, where regulators have pushed to reduce banking-like features (like the daily unwind) and increase capital-markets-like features (like longer-term repo agreements with counterparties who are thus more incentivized to do credit work).
The risk of relying on such tools in securities markets is that microprudential rules could lead more asset managers to act in a similar manner -- in other words, they could increase 'herding.' Macroprudential rules would direct capital flows by tilting the investment landscape in favour of one set of assets over another.
Obviously the objection here is that retail money market funds are not subject to lots of bank-like regulation; in particular they tend to have roughly zero capital buffer. (But of course their investments are all short-term, high-grade, etc., and there's some sort of informal capital buffer from the reputational pressures on their sponsors to avoid breaking the buck.) And tentative regulatory efforts to make money market funds have some capital buffer have not gone well.
To contact the author on this story:
Matthew S Levine at firstname.lastname@example.org
To contact the editor on this story:
Toby Harshaw at email@example.com