Tuesday, October 28, 2014

Fear of illiquidity

There are thousands of fears, from arachnophobia to globophobia (fear of balloons) to zoophobia (fear of animals). What might the fear of market illiquidity look like?

Say that you are petrified that a day will come when markets will be too illiquid for you to convert your wealth into the things you need. There are two ways for you to buy complete peace of mind.

The first strategy involves selling everything you own now and buying checking deposits, the sine qua non liquid asset. Get rid of the house, the bonds, the stocks, the car, your couch, and your books. Use some of the proceeds to rent a house and a car, borrow books, and lease furniture. In renting back the stream of consumption benefits that you've just sold, your level of consumption stays constant. Negotiate the rental arrangements so that the lessor—the owner—cannot cancel them, and so that you can walk out of them at a moment's notice. Structuring things this way ensures that rental obligations in no way inhibit your ability to stay liquid. With your hoard of highly liquid deposits and array of rental agreements, you've secured a state of perfect liquidity. Relax. Breathe in. Enjoy your life.

The second way to perfectly hedge yourself from illiquidity risk would be to buy liquidity insurance on everything you own. For instance, an insurer would guarantee to purchase your house whenever you want to sell at the going market price. Same with your stocks, and bonds, your car and couches and your books. With every one of your possessions convertible into clean cold cash upon a moment's notice thanks to the insurer, you can once again relax, put your legs up, and lean back on the couch.

Since both strategies lead you to the same infinitely liquid final resting place, arbitrage dictates that the cost of pursuing these two strategies should be the same. Consider what would happen if the liquidity insurance route was cheaper. All those desiring a state of infinite liquidity would clamor to buy insurance, pushing the price of insurance higher until it was no longer the better option. If the checking deposit/rental route was cheaper, then everyone would sell all their deposits and rent stuff, pushing rental prices higher until it was no longer the more cost-efficient option.

Now I have no idea what liquidity insurance should actually cost. But consider this: liquidity option #1 is a *very* expensive strategy. To begin with, you'd be forgoing all the interest and dividends that you'd otherwise be earning on your bonds and stocks. Checking deposits, after all, offer no interest. Compounded over many years, that comes out to quite a bit of forfeited wealth. Second, you'd have to rent everything. And the sort of rent you'd have to negotiate would be costlier than normal rent. Last time I checked, most landlords require several months notice before a renter can be released from their rental obligation. But the rental agreements you have negotiated require the owner to accept a return of leased property whenever *you* want—not when they want. And that feature will be a costly one.

Since option #1 is so expensive, arbitrage requires that option #2 will be equally expensive. Let's break it out. Option #2, liquidity insurance, allows you to keep the existing flows of income from stocks and bonds as well as saving you from the obligation of paying high rent (you get to keep your house and all the other stuff). Not bad, right? Which means that in order for you to be indifferent between option #1 and #2, the cost of insurance must be really really high. If it wasn't, everyone would choose to go the insurance route.

So who cares ? After all, liquidity insurance doesn't exist, right? Wrong. Central banks are significant providers of liquidity insurance. They insure private banks against illiquidity by promising to purchase bank assets at going market prices whenever the bank requires it. This isn't full and complete liquidity insurance— there are a few assets that even a central bank won't touch—but it's close enough.

The upshot is that banks are well-protected from illiquidity. They get to keep all their interest-yielding assets and at the same time can rest easy knowing that the central bank insures that those assets will always be as good as cash. Consider what things would be like for private banks if the central bank were to get out of the liquidity insurance business. Now, the only way for bankers to replicate central bank-calibre liquidity protection would be for them to pursue option #1: sell their loan books and bond portfolios for 0%-yielding cash. But then they'd be foregoing huge amounts of income. They might not even be profitable.

With logic dictating that the cost of buying liquidity insurance needs to be pretty high, are modern central banks charging sufficiently stiff rates on liquidity insurance? I'm pretty sure they aren't. Regular insurers like lifecos require periodic premium payments, even if the event that said insurance covers hasn't occurred. But the last time I read a bank annual report, there was no line item for liquidity insurance premiums. It seems to me, and I could be wrong, that central banks are providing liquidity insurance without requiring any sort of quid pro quo. Feel free to correct me in the comments section.

Say that I'm right and that central banks are providing private banks with underpriced liquidity insurance. Central banks are ultimately owned by the taxpayer, which means that taxpayers are providing private banks with artificially cheap liquidity insurance. And that's not a fair burden to put on them. Nor is the underpricing of insurance a good strategy, since it results in all sorts of institutions getting insurance when they don't necessarily deserve it.

Does anyone know if central banks have any sort of rigorous model for determining the price they charge for liquidity insurance. Or are they just winging it? ... it sure seems like it to me.


  1. My guess: liquidity insurance is bundled in with deposit insurance, which is priced. How that price is determined, I don't have a clue.

    A line of credit is a form of liquidity insurance. you don't sell your assets; you borrow against your assets.

    1. Re: deposit insurance. I was thinking that might explain some of the missing cost. I don't believe deposit insurance fees are very high, probably not enough to cover both deposit and liquidity protection.

      Re: lines of credit. Yep, very similar concept to liquidity insurance. Having a pre-arranged line of credit to be secured by each of your assets would do the trick.


      Great interview, by the way.

  2. These are exactly the kinds of thoughts that enter one's mind when looking at a "start from scratch" set up in equilibrium, where internal strength needs to be devised and hopefully remains within scale for a small time investor as well. But the costs of complete liquidity across equilibrium tend to knock out that possibility!

  3. I'm not sure how you would even draw up such a policy. How would you define the "market price" of an illiquid asset in such a way that did not reflect its illiquidity? For some things you might be able to get two way quotes and take the mid-market average. But how can I get a two way price specifically in my house? No one else can provide an offer price but me. I don't think there is a market price that is independent of the illiquidity.

    When central banks provide liquidity insurance, I think there generally providing funding liquidity rather than market liquidity aren't they. They don't buy assets outright, they lend and take the assets as collateral.

    When I sell my house I do two things. One, I get cash which I can use to make payments. Two, I remove any future exposure to the value of my house. In theory I could do those things seperately. I could repo my house (or just borrow against it), which gets me the cash, but doesn't remove my exposure. Or I could sell it for settlement after 12 months, which removes my exposure, but doesn't get me cash. Central banks are usually just facilitating the first bit, I think.

    1. From a policy perspective we don't have to think too much about houses and couches since central banks doesn't participate in these markets. Financial assets will typically have an existing bid and offer price, so the Fed would provide a bank with liquidity by always standing ready to purchase any amount of that asset within the spread, even if the order exceeds the size of what is being bid and offered in the market.

      Yes, central banks don't buy assets outright, but the idea is still the same... they will take all collateral at market prices. The bank is insured in that it knows ahead of time that its collateral always has a market.

      Finally, I think I agree with you that central banks are "usually just facilitating the first bit." Or at least, that's what they do when the crisis finally hits and private banks call in liquidity. Since crisis are rare, central banks are usually just providing liquidity insurance, but that insurance is not being triggered.

    2. OK. So maybe the better analogy in your second way would be if the insurance company agreed to lend to you when required, against the security of your assets. A sort of committed liquidity facility.

  4. "Since option #1 is so expensive, arbitrage requires that option #2 will be equally expensive."

    I think this expectation is incorrect. Instead, there is the possibility that option #2 can be perpetually LESS EXPENSIVE to maximize market share. In other words, option #2 can have some sort of inherent cost advantage that allows enduring economic relationships to exist in an economy.

    Yes, this enduring situation can result in under-pricing or a shift in ultimate risk taker. And, yes, this shift will result in economic winners and losers. Winners and losers that can be labeled or otherwise identified.

    1. I'm not sure I get you. Are you implying something about monopoly?

  5. I was picking up on the idea that "arbitrage" results in equality of price when two paths lead to the same result. Instead, it seems to me that one path will have some natural advantage that will cause it to be the price driver. That would force the second path into an inferior position, forcing it to differentiate on other than pure price criteria.

    My concept of a common price is not quite the same as your expressions of "cost of pursuing" or "equally expensive" so to a large degree, my comment is off-topic. Sorry about that.

    But, in the back of my mind, not well expressed, was the thought that a wide swath of the economy is helped by the Central Bank providing liquidity. Yes, the taxpayer is ultimately funding this liquidity but at least some taxpayers think they are the winners in this system.

    1. "...a wide swath of the economy is helped by the Central Bank providing liquidity. "

      That's a fair point, in the same way that society might be made better off by free government-provided medical insurance. However, taxpayers understand health care... its very immediate. But they don't understand finance. So I don't know how many of them are in a position to calculate if they are winners or losers in a regime with under-priced liquidity insurance for banks. Liquidity insurance could also be provided by the private sector, but does the private sector have a comparative advantage in creating this insurance? Would market prices for liquidity insurance provide better information?