Moral hazard, meet adverse selection

by John Q on September 19, 2008

At a time when anyone on the cutting edge is talking quadrillions, it seems a bit petty to worry about a $50 billion component of the latest bailout (only $500 per US household!). Modest as it is, the insurance scheme offered to money market funds by the US Treasury provides the opportunity to explain a little bit more about the theory of insurance.

By now, everyone has heard about moral hazard, that is the encouragement to take risky or reckless action that arises when your losses are insured by someone else. Now it’s time meet moral hazard’s evil twin, adverse selection. That’s what happens when the people you are offering to insure already have a pretty good idea whether they are going to collect or not.

For example, if you have a money market fund with assets now worth 97 cents in the dollar, and someone offers you the chance to insure against ‘breaking the buck’ you’re going to take it, since it’s a sure thing that you can collect. (It doesn’t matter much what the premium is, since your loss will increase by an amount equal to the premium paid). As far as I can tell, that’s exactly what’s on offer from the US Treasury right now. (I’d be pretty annoyed if, like Legg Mason, I’d just tipped in $630 million of company money to protect customers against this eventuality, but I guess they can probably claw it back, and take the Treasury deal instead).

By contrast, if I have kept my customers’ money in safe assets like government bonds,it would be silly to take out insurance against a risk to which I’m not exposed. That means that the premiums have to be high, which in turn means that even moderately risky firms will find the deal unattractive, unless the public subsidy keeps increasing.

The current holdings of money market funds are around $3.4 trillion. Let’s suppose half of them are in some kind of trouble, amounting to an average loss of 3 cents in the dollar (the same as the Putnam fund that just put up the shutters). If all these funds sign up for the scheme, while the other half stay out, the losses will almost exactly wipe out the Treasury’s stake.

An interesting sidelight is that this scheme is being funded by cashing in the assets of the Exchange Stabilization Fund, established in the 1930s to intervene in foreign exchange markets. I guess we’ll never need any of those Depression era relics again.



P O'Neill 09.19.08 at 11:22 pm

Is it clear that asymmetric information, an essential component of adverse selection, is present here?


John Quiggin 09.19.08 at 11:37 pm

Pretty much built in, I think. Participation is voluntary and open to all, so the Treasury has (in effect) no information it can act on. Obviously the funds aren’t perfectly informed, but presumably they have some idea whether they are going to go below a dollar.


Stuart 09.20.08 at 12:37 am

So would this be like if the US government offered comprehensive home insurance to anyone in the country with a standard premium based on the average risk in the US? They would end up screwed because only people on the Gulf Coast, Florida and on the fault would take it up.


John Quiggin 09.20.08 at 12:43 am

Exactly right, Stuart.


P O'Neill 09.20.08 at 12:47 am

But there is a lot of public information about what these money funds hold. For one thing, they tell their shareholders, and the interest on Treasuries is deductible on state income taxes, so there is a lot of documentation floating around about the portfolio composition. If the Treasury wanted to screen on portfolio composition, it could. I have the same problem with the location example. There is information that the government could use to set premiums. They choose not to. Yes they end up with a bad risk pool but I don’t see the hidden information.


Michael Drake 09.20.08 at 1:10 am

The economics of this story are generally way above my pay grade, but I thought the very same thing when I heard they were going to allow money market funds (I thought it was mutual funds generally) to opt in. I’d assumed they’d at least make it mandatory (like FDI) going forward. But…not.


Chris Bertram 09.20.08 at 6:10 am

#2,3 Yes, although I once had an economist from Dartmouth College explain to me that it was quite impossible for Britain’s Automobile Association to exist, for exactly this reason. I insisted that it did indeed exist, and had done so for many years. He didn’t believe me.


Chris Bertram 09.20.08 at 6:14 am

I think the moral there, is that if people are sufficiently risk averse and the premium is quite low, then you can, in practice, get away with a standard charge.


Monte Davis 09.20.08 at 11:22 am

Arnold Kling expands the argument from the money-market prop to the Grand Bailout:

“It sounds to me as though [the proposed toxic-debt-collecting entity] is going to be wearing a big sign around its neck saying, ‘Adversely Select Against Me.’ “


Slocum 09.20.08 at 12:06 pm

But I wonder if firms that don’t have risky assets and don’t need insurance will really be able to opt out–not for risk reasons but for marketing reasons. That is, how does an uninsured fund compete for depositors against funds than can advertise that they’re federally insured — especially in light of recent events? No, I expect money market funds will find they have to sign up even if they’re not required to do so by law.


Barry 09.20.08 at 12:12 pm

P O’Neill 09.20.08 at 12:47 am
“There is information that the government could use to set premiums. They choose not to. Yes they end up with a bad risk pool but I don’t see the hidden information.”

Another way to phrase it is that there is *some* public information, which both parties have. Then there is *private* information, which only one party has. Adding onto that is the fact that the corporations will have put in far more lawyer/accountant/financial expert person-hours on their own situation than the government could (unless the government goes out and hires all of the laid-off staff from Wall Stree, which might be a good idea).


a 09.20.08 at 12:35 pm

If the premium is constant over all funds, then the government will be arbitraged, as funds take on more and more risk to earn higher returns and win depositors.

If the premium is supposed to be a function of the risks of the fund, then someone needs to be able to calculated the risk. I doubt the government will be able to do that. If it’s the fund itself, then obviously there is every incentive to minimize one’s risk.


Slocum 09.20.08 at 1:53 pm

If the premium is constant over all funds, then the government will be arbitraged, as funds take on more and more risk to earn higher returns and win depositors.

Yep — that’s what I’d expect. Funds will all sign on and then go for the highest returns and highest risks that regulators will allow.


virgil xenophon 09.20.08 at 4:17 pm

Slocum and a are on target. Of course this problem is exactly why everyone thought “insurance” for money market funds was a pointless exercise in the first place, else they would have been originally formulated in a way other than they were.


virgil xenophon 09.20.08 at 6:11 pm

Regarding my post @14 I should add, however, that there is another dimension to this problem that, looked at on another analytical level, reveals a different kind of risk that no one is talking about, namely the prospect that the “successful” merging of commercial banks with their troubled investment bank brethern will, rather than stabilize the likes of Merrill Lynch, endanger the likes of commercials like BOA.

One of the reasons given for the dissolution of Glass-Steagall was that our own financial entities could not compete with institutions in the rest of the world who were not so encumbered, and therefore larger by many orders of magnitude than US based institutions with which they were in competition. Yet I can remember in the late eighties/early nineties when the bloom began to come off the Japanese economy, brokers advised their customers to steer away from money market funds that had a significant composition of CD’s held by Japanese Banks. The reason? Japanese banks–hybrids of commercial and investment banks–had significant exposure to the dissolving Japanese real-estate bubble.

Sooo——it seems to me that the present “fix” while not a total illusion, is not the be all and end all. It will require a change in the corporate “culture” of these institutions as well as the proposed legal/regulatory ones. Suggestions as to how to accomplish THAT would require me to type more words than anyone here wants to read.
But it is enough to say that many of the same sort of people responsible for this mess are simply walking across the street to
work for the newly merged institutions that absorbed their own.


wcw 09.20.08 at 10:01 pm

Well… yes, but no.

Yes, there will be adverse selection from those few money funds who fucked up and are absolutely, positively fated to break the buck. But most aren’t. This isn’t health insurance. Most money funds do not have terminal cancer. Most money funds are fine. Thing is, in the current market, even the more-conservatively managed ones are worried about facing a run.

A money fund’s holdings mature fast (generally you’ve got 30% of your cash back within a week or so), but not so fast that you can’t be forced to sell to meet redemptions. Normally, you just sell and all is fine, but on some things there have been no bids out there, like on Frannie paper, despite the fact that it currently is guaranteed by the US taxpayer. If you can’t sell a Treasury-alike, a run will kill you.

Simply presenting the existence of this insurance immediately props things up. If need be, the Treasury will take the paper (which, as noted, in some cases the Treasury itself already guarantees). Redemptions no longer are a problem.. and so there is no need for a run.

The current situation is less like adverse selection, and more like the FDIC. No FDIC means bank runs, panics and attendant vicious cycles. FDIC means none of that. Yes, in the long run with poor regulation it also means the S&L crisis and the RTC, but I am pretty sure that isn’t what your post was about. You posited an equivalence to health insurance, and that just isn’t there.

Full disclosure: I’ve worked in the investment business since ~1993 and my current employer has ~$200B in money-market assets. And — not that it matters — we didn’t break the buck, and unlike several competitors (google for the names) we didn’t have to inject cash to prop up NAV.


virgil xenophon 09.20.08 at 11:05 pm


I began in the financial/investment business in 76 (I’m retired now)
and don’t disagree with you. My major worry, which federal “insurance” DOES ameliorate, is the panic that comes from lawyers and accountants, with Enron on their minds, jawboning their companies into writing down good assets (94%+ of all mortgages are currently performing) to crazely insane levels by 80-90% write-offs, all to march to the tune of the FASB which idiotically imposed “mark to the market” rules on long-term instruments designed to be held only by institutions (as one old bond trader told me when I first came into the business: “Only Institutions can truly afford to hold bonds long-term”.) Granted that these CMO’s were complicated and “opaque” financial vehicles to say the least, and the risk involved poorly understood–and those flogging them less than introspectively concerned about their “fit” for any and all investment profiles–but even so, if held to maturity almost none of them (the bundled CMO’s) would have been under the water–and if not required to be marked to market could have been used to settle interim trades based on a NAV of close to par or counted as par for accounting purposes when calculating the book value of the financial firms holding them in their portfolios–all without the sturm und drag that brought the system down.

Thus, by my lights, the current Federal “assurance” should have never been needed in the first place were it not for an Enron-leary FASB. Enron was also the reason the ratings agencies (S&P, Moody’s) made the ridiculous call to lower the ratings of firms whose stock price was under pressure (thus requiring firms to come up with additional billions) based on lowered stock prices as justification–despite the basic fundamental soundness of the firm in question. Here again panic, not rational thought proved that the “professionals” are oft more easily panicked that the main-street investor.


CMatt 09.21.08 at 6:03 am

If they are good/great assets (94+% performing), and institutions can evaluate this, why does there not exist insurance for these (obviously [!] good) investments, to hedge against the risk of [posited as badly-conceived] mark-to-market short-term [accounting gimmick] deficiencies?

As posited, they’re almost golden, so why doesn’t there exist a split-the-difference insurance regimen that guarantees positive return for everyone involved, in spite of [posited] short-term, short-sighted marking requirements?


Zamfir 09.21.08 at 5:59 pm

Perhaps I am missing something, but this scheme seems to allow that the US gov actually looses 50 billion dollar on the end, as opposed to lending it out with a few percent loss. In that case, 50 billion IS a large amount, even for current day madness.

Comments on this entry are closed.