Saturday, May 29, 2010

The World Supply of Liquid Assets

I found this piece by Ricardo Caballero, which is a useful starting point for a discussion of the role of safe assets in the financial crisis and in the recent European sovereign debt crisis. An important feature of post-2000 financial markets in the United States was the multi-layered role of housing in somehow "lubricating" financial markets. Housing played a direct role as collateral for mortgage debt that was used to finance consumption expenditure, and mortgage-backed securities were very important as collateral in financial market arrangements. However, when everyone realized that incentive problems had caused us to vastly overvalue the direct mortgage debt, and the securities built up from it, we then had much less of what we consider to be safe, liquid assets. With less liquid assets, there has to be less trade on financial markets, as was, and still is, the case.

What can governments do about this? Well, if we think there are less liquid assets than is socially optimal, the government can step in and supply more. One way to do this would be through fiscal policy. In order not to deal with issues related to public goods and government spending on goods and services, consider a policy where the government increases the deficit by reducing current taxes. Now, we have to ask why this would accomplish anything. As my undergraduate students know, (see Chapters 8 and 9 of my intermediate macro book) in a world where credit markets are perfect, changes in the timing of taxes are irrelevant, and so by implication is the level of government debt. Given that the government always pays off its debt eventually (another important assumption), Ricardian equivalence holds, and everyone simply saves their tax cut so as to pay off the higher taxes they will owe in the future to pay off the government debt.

Of course, we know we are not in a Ricardian equivalence world. The financial crisis was about frictions - the frictions that make studying financial and monetary arrangements interesting. However, even standard ways of thinking about financial frictions do not necessarily lead us to any obvious role for the government in "lubricating" financial markets. Suppose we think about standard private information and limited commitment problems. If private lenders face private information problems that make lending costly or might shut it down altogether, or there is limited commitment which makes collateral useful in mitigating the resulting incentive problems, then borrowers may be constrained and shut out of credit markets altogether. Further, spreads emerge - there are significant margins between borrowing and lending rates of interest, and there are risk and default premia.

But the government is faced with the same private information problems and limited commitment problems as is the private sector. In order to somehow make credit markets work more efficiently, the government has to have an informational advantage - it has to somehow be a better banker than private bankers - or it must have an advantage in collecting on its debts. For example, if we imagine a world where the limited commitment problem is so severe that lending shuts down entirely, we might imagine the government stepping in and cutting taxes, issuing debt in the present, and collecting taxes in the future to pay off the government debt, thus essentially making loans to the private sector. However, if the government is no better than the private sector at collecting on its debts, private sector economic agents will run away from their tax liabilities just as they run away from their private debts. If that is the case, this government fiscal program will not even get off the ground.

The point is that the correct type of fiscal intervention for the government in a world with significant financial frictions (which is the world we live in, and always have) is far from obvious. If anyone tries to tell you it is, you should not listen.

Now, what about monetary policy intervention? If we view a financial crisis as temporary, this gives a natural advantage to monetary policy intervention, which can in principle be reversed quickly. Milton Friedman of course recognized the unwieldy nature of the fiscal decision-making process - tax policy is hard to turn on and off. How can the Fed create more liquid assets? You might think the answer would be open market purchases and reductions in short-term market interest rates. However, note that we think of the recent reductions in long-maturity Treasury yields as reflecting an increase in the demand for Treasuries - a flight to safety. The low yields on Treasury securities at all maturities reflects a liquidity premium, or a scarcity of safe and liquid, assets. Thus, if we eliminate the scarcity, we would be increasing yields at all maturities, not lowering them. Perhaps we are thinking of this problem in exactly the wrong way.

Now, I'll address more specifically what is in Caballero's piece. He starts off this way:
While the focus of commentators and policymakers is on secondary (although still important) regulatory and corporate governance issues, the fundamental problem in the current global macroeconomic and financial equilibrium is one of asset shortages. In particular, there is a shortage of safe “AAA” assets. The world seems to need more US Treasury-like instruments than are available.
The part of this I don't like is the notion that the regulatory issues are somehow secondary. This is at the heart of what has led to what Caballero is calling a "shortage." The regulation was bad, which created the incentive problems, which ultimately created the shortage of privately supplied liquid assets. To get these privately-supplied assets back, you need proper regulation.

As Caballero seems to recognize, "producing" AAA assets is not like producing good shoes as opposed to bad ones. Whether US debt is AAA or something worse depends on self-fulfilling expectations. If the world began to think tomorrow that the United States was not creditworthy, then it would not be. We are currently in a fortunate position, as the world treats us as highly creditworthy - so much so that the yields on US Treasuries are incredibly low, in spite of our high deficit and accumulating debt. But we should not push our luck. Higher US deficits might give our creditors pause, and cause them to seek safe haven elsewhere. Further, we don't want a large quantity of debt outstanding when credit markets get back to normal and safe private liquid assets flood the market to compete with US Treasury debt.

Where Caballero loses me is here:
We should separate the production of micro- and macro-AAA assets. The private sector is much more efficient than the government in producing micro-AAA assets, but the opposite is true for macro-AAA asset production.
What's this about micro-AAA and macro-AAA? Maybe what he means is that there was aggregate risk hiding on the balance sheets of various financial institutions - as a result the liabilities of those institutions seemed a lot safer than they actually were. But this gets back to the regulatory reforms that Caballero dismisses at the beginning of his piece. It seems to me that we can make the private sector very good at producing macro-AAA assets if we regulate the financial institutions properly - that's the basic problem.


  1. "With less liquid assets, there has to be less trade"

    Well "less trade" is fine, but don't we still prefer "FEWER assets".

    Although a radical progressive in economics, I guess I'm still a reactionary on grammar.

  2. Please explain to me how one draws a distinction between "asset shortage" and "liquidity surfeit?" Might not the illusion of an "asset shortage" stem an extended period of official suppression of interest rates in the pursuit of never-correcting asset markets?

  3. Caballero claims there is a shortage of safe assets due to flight to safety. Perhaps. The alternative view is that:

    1) this is inconsistent with the performance of emerging market, non-AAA corporate, and high yield bond spreads in the past six months. All of these "risky" securities have seen spreads decline to levels that are considered "normal". This should not be the case if there were "flight to safety" demand for AAA securities.

    2) production of AAA-rated securities (in the form of ABS and CDO tranches) skyrocketed prior to the crisis. Roughly 80% of ABS and CDO issuance were tranched at AAA, which implied, at the time, that they were extremely safe. If Caballero is right, then what explains the rising demand for "safe" securities BEFORE the crisis began?

    The alternative view is that Fed policy causes demand for safe assets. There are two ways to structure a portfolio: combine assets of varying volatility; or lever a portfolio of low-volatility stocks. The risk to the latter is that the Fed will raise the cost of that leverage in an unexpected manner. Enter the Fed. With its "extended period" and "measured pace" language, the Fed increases the demand for carry trades in safe securities. This was true from 2002-2006, and it is true again today.

  4. Caballero means that government could sell insurance against adverse macroeconomic events. He thinks that government would be an efficient producer of such insurance. I have serious doubts about it, as private company AIG has severely mispriced tail risk insurance before the crisis, quasi-public Fannie and Freddie did the same.

    I think it is much better if government would ex-post intervene in credit markets, as suggested by Michael Woodford (see ) Woodford hopes that authorities could select the specific targets for intervention according to microeconomic considerations. I think he ignores information problems. I think central banks should purchase market portfolio during credit interventions. ECB is now buying Greek bonds, they should be buying Eurozone sovereign and private bond index instead.

  5. "Of course, we know we are not in a Ricardian equivalence world. The financial crisis was about frictions - the frictions that make studying financial and monetary arrangements interesting."

    Yeah, but it's a lot more than frictions. Very few members of the public have are expert in economics, finance and government. They know nothing about Ricardian equivalence and think this way little, if at all. Plus, they are not infinitely long lived. They consider something like this very little if at all. This type of evaluation done accurately takes a great deal of expertise, info, time and self discipline. Few have the expertise, and few want to spend huge amounts of all too rare free time doing calculations like this instead of spending some time with their families and on other leisure. Few people don't make decisions like this.

    It's not just a matter of rationality. It's that so much of the behavior assumed of the public in models requires a high level of specialized expertise that few have, and even if you have it, a high level of information, a high cost of very rare and valuable time, and a high level of self discipline. Often, few people have all of these things.

    So I hope new monetarists take all of this into account, and don't take models that assume these things away overliterally.

  6. I have a question about this, if you don't mind:
    "However, if the government is no better than the private sector at collecting on its debts, private sector economic agents will run away from their tax liabilities just as they run away from their private debts. If that is the case, this government fiscal program will not even get off the ground."

    Maybe I misunderstand what you're saying, but it seems to me that you're saying that if people disregard the risk of having to pay higher taxes in the future, they'll spend the money the government's been handing out. So how does that keep the fiscal stimulus from getting off the ground? Wasn't the whole point of the stimulus program to get people to spend the money the government's handing out?