Friday, March 9, 2012

Greg Ip on Safe Assets as Money

Greg Ip has a new article in The Economist where he discusses how U.S. treasuries and other safe assets can serve as medium of exchange:
[D]emand for American government debt is driven by much more than a hunger for returns. Financial-market participants use Treasury bonds and bills as collateral to secure lending, for instance. And for risk-averse investors such as foreign central banks, money-market funds and retirees, America’s debt is uniquely suited to storing savings without much due diligence. In short, its government debt is a lot like money.
I agree with the central premise of the article, but would add a few points. 

First, I would frame the discussion in this way: there are retail money assets and institutional money assets.  Retail money assets are the traditional money assets measured by the M2 money supply and are used by households and small businesses.  Institutional money assets go beyond M2 and includes treasuries, commercial paper, repos, GSEs, and other safe assets used to facilitate exchange in the shadow banking system.  Since most of the creditors to the shadow banking system are institutional investors, these assets should be called institutional money assets. 

Second, institutional money assets include both privately-produced and publicly-produced safe assets.  If the Fed is doing its job and and providing sufficient aggregate demand to keep the economy at full employment, then there should be plenty of privately-produced safe assets.  Only if the Fed allows nominal spending to crash, which would reduce privately-produced safe assets, is there a need for the government to step in and create safe assets.  To put it differently, if the Fed were to announce today that it was adopting a nominal GDP level target and planned to restore it to its pre-crisis trend, then there would most likely be a recovery and an increase in the private supply of safe assets.  As a result, the institutional money asset supply  would increase and there would be less need to produce treasuries. 

Third, the broader context for this discussion is that there is currently a shortage of safe assets for the global economy.  And, as I noted before, there is both a long-term, structural dimension to this problem as well as a short-term, cyclical one:  
The structural dimension is that global economic growth over the past few decades has outpaced the capacity of the world economy to produce truly safe assets, a point first noted by Ricardo Cabellero.  The cyclical dimension is that the shortage of safe assets was intensified by the Great Recession, a point stressed by Gary Gorton.  I previously made the case that both the Fed and the ECB were an important part of the cyclical story by failing to restore nominal incomes to their expected, pre-crisis paths.  In other words, since 2008 the Fed and the ECB passively tightened monetary policy which caused some of the safe assets to disappear while at the same time increasing the demand for them. 
This failure of the world's major central banks means U.S. treasuries will remain in hot demand.  This seemingly insatiable demand is evidenced by the low yields on treasuries.

Fourth, there is a Triffin dilemma for U.S. debt.  The global financial system in its current setup needs increasing amounts of U.S. treasuries.  This means the U.S. government must continue to run large budget deficits.  Over time, however, these large budget deficits may jeopardize the safe-asset status of U.S. treasuries, the very thing driving the insatiable demand for them. So a tension exists between providing enough treasuries to keep the global economy going and maintaining the safe-asset status of treasuries. 

Finally, the New Monetarists like David Andolfatto have been making some of these points for awhile. Greg Ip should spend some time talking to them as well. 


  1. Perhaps I misunderstand the nature of shadow banking and repurchase agreements.

    An shadow bank sells some asset to a customer but agrees to repurchase it very soon, e.g. the next day. The repurchase price is slightly higher than the original price. If the shadow bank reneges on its agreement, then the customer gets to keep the asset.

    The difference between the repurchase price and original price is analogous to an interest rate. The asset which the customer may keep if the shadow bank reneges on its agreement is analogous to collateral. While repurchase agreements are not, technically, loans, they emulate the same relations using different accounting categories.

    In practice, repurchase agreements are often rolled over for extended periods. Assets are repurchased and resold each day earning a small interest rate for the customer and liquidity for banks.

    The relative safety and short-term character of repurchase agreements means customers come to think of them like high-falutin demand deposits, i.e. 'money in the bank'. Customers can make "withdrawals" by only partially rolling over the repurchase agreement.

    For example, if a customer wants to "withdraw" $100,000 of a $1 million dollar repurchase agreement, then he only rolls over $900,000 the next day and debits his currency or FDIC-insured accounts for the difference. The investment firm arranges for the repurchase of a different combination of assets to reflect the new lower value.

    Like ordinary commercial banking, rather than withdrawing money to make other payments, it's usually easier to just write cheques or arrange for an electronic transfer of funds. Large institutional investors, the customers of shadow banks, started using their repurchase agreements to faciliate exchange.

    For example, an investment firm could write a cheque for $100,000 (like in the previous example) to buy some asset; the seller would accept the cheque and later return it to the shadow bank for redemption.

    Since "depositors" in shadow banks were unlikely to all want to "withdraw" their money all at once, the shadow bank only needed as much money as necessary to satisfy actual withdrawals. The money-multiplier effect, then, of shadow banking is analogous to that of ordinary commercial banking.

    Okay, so here is an important question, I think: when a shadow bank receives receives a "cheque" for redemption, does it settle in base money or FDIC-insured demand deposits?

    If base money, then the shadow banking system would be just a counterpart to ordinary commercial banking, i.e. they both compete for the same limited supply of base money to settle payments. However, if FDIC-insured demand deposits, then shadow banking is kind of built atop of the ordinary commercial banking system, i.e. ordinary bank deposits would be to shadow banking what base money is to ordinary bank deposits.

  2. David,
    "...if the Fed were to announce today that it was adopting a nominal GDP level target and planned to restore it to its pre-crisis trend..."

    As a corporate Treasurer considering an ABCP security, do I care that expected NGDP growth rate is expected to be 5%; or that 5% would still leave NGDP woefully below trend?

    Two factors drive my safe-asset investment decision: illiquidity tail risk and the possibility of collateral default over an extremely short holding period (weeks). To gauge both risks, why wouldn't knowledge of just the expected NGDP growth rate be sufficient? "NGDP shortfall" is a concept that describes a deviation from a condition present before I considered the investment. Its kind of like saying, "I won't buy that stock because their earnings fell last year, even though I expect them to be up significantly going forward."

  3. You wrote that “there is currently a shortage of safe assets for the global economy.” But, as I understand the economists’ conception, a shortage of X is a disequilibrium condition in which the price of X is—stickily-- rising. There is certainly no shortage of U.S. treasury securities: their prices are hardly moving much, nor are they sticky. Many people wish there were more safe assets, such as treasurys, but that doesn’t constitute a *shortage*: after all, it would be better if there were more goods of all sorts. I suspect the real point is that not only would it be better if there were more safe assets, but a greater supply could be secured *costlessly*; so it is mere perversity on someone’s (the government’s or governments’) part that the quantity is sub-optimal. “Shortage” here is a misnomer for *political failure*.

    You added: “The global financial system in its current setup needs increasing amounts of U.S. treasuries.” But *currency* is a safe asset, and that can be produced without running a budget deficit.

  4. Philo,

    There is a shortage of safe assets because the lower zero bound (or the interest rate paid on excess reserves) acts like a price ceiling. The equilibrium interest rate for assets like T-bills is lower than its actual price.

    How do we know? Because when demand for a good is frustrated by a price ceiling, the excess demand spills over into whatever people settle for instead. For example, a binding price ceiling on apples frustrates demand, leaving some people who want to purchase apples unable to find any sellers. So what do they do instead? Maybe they buy oranges, bananas, pears, or whatever--probably something that serves a similar goal as apples. So the excess demand for apples spills over into higher demand for other kinds of fruit.

    When interest rates on safe assets approach the lower zero bound, their price becomes stuck. Even if demand continues to rise, interest rates will not fall. The frustrated demand spills over into close substitutes, and one of those substitutes happens to be money itself. When the demand for money starts rising, unless offset by increasing supply, aggregate nominal income takes a fall and instigates a recession.

    Lower aggregate income exacerbates the problem, because it lowers average nominal income while leaving nominal debts unchanged. Default risks begin rising as borrowers unexpectedly find it harder to finance repayments, and investment portfolios begin looking riskier and riskier. In other words. The shortage of safe assets spills over into a shortage of money once we approach the lower zero bound, and then the shortage of money further reduces the supply of safe assets.

  5. Lee Kelly,
    If the price of apples is held to 5 cents, then I won't be able to find anyone to sell me an apple.

    If the yield on a T-bill is held to zero, I can find an institutional deposit account at a bank that pays close to zero. Or I can ask a bank to issue me ABCP, repo's or even AAA-rated ABS. The supply of safe assets by the banking system is essentially perfectly elastic at a zero marginal cost (as long as there is sufficient yield spread).

    The supply of safe bank assets is not constrained by a price ceiling, but by the perception that bank and shadow bank liabilities no longer count as "safe" (or "information insensitive", to use Gorton's term). I would be more inclined to believe this "information sensitivity" is caused by the NGDP shortfall if the prices of risk assets generally were depressed. This is not the case.

    Note that the whole shadow bank system arose to meet the demand for AAA-rated assets. This demand could not be sated by direct credit issuers (sovereigns, corporates). However, the financial system could produce them at will, until, of course, their production led to the GFC.

  6. Diego,

    Same thing. I glossed over some institutional complexities because they weren't relevant to my point.

    Sure, investors can, for example, easily get more regular commerical deposits from banks. But that just shunts the issue along. Now it's the banks who can't find any people selling "apples" because the price has hit the ceiling, so instead they hold onto the (base) money. The demand for safe assets spills over into money, depresses nominal spending, and exacerbates the demand for safe assets.

    This is basically another way of restating the cumulative rot problem.

    If an equilibrium interest rate is below zero, then, given present institutional arrangements, we have a shortage of that asset by definition. It's a straightforward supply and demand situation. That frustrated demand has to go somewhere, and when it is a demand for safe assets, the somewhere it is likely to be directed is into money itself.