Sunday, September 19, 2010

QE the Sequel: Putting Ben’s Money Where His Mouth Is

A consensus is emerging among Fed watchers that the Fed is set to embark on a fresh round of “quantitative easing” (QE), faced with a subpar employment growth and a lingering threat of deflation.

Abstracting from whether the economic outlook is such as to warrant further stimulus, I wanted to focus here on what kind of “QE” might be more effective this time round, if it were to happen.

Pre-empting my conclusion, let me say that my proposal will probably sound like the mother of unconventional measures, but it’s actually a variant of what the Chairman himself proposed back in 2002, at his famous “it” speech on deflation. But let’s start from the beginning.

First of all, “QE” means the purchase by the Fed of a certain quantity of risky assets, funded by the creation of bank reserves. The intended objective is twofold: First, to boost the price (/lower the yield) of the assets purchased (in the case of the Fed’s first round of QE, these would be US Treasury bonds, agency debt and mortgage backed securities (MBS)); and second, to affect the price of other risky assets through the so-called portfolio balance effect.

For details on how the portfolio balance channel is supposed to work you can read Brian Sack’s speech at the Money Marketeers last year (here), but here is an excerpt from that speech that sums it up:

“[T]he purchases bid up the price of the asset [being purchased] and hence lower its yield. These effects would be expected to spill over into other assets that are similar in nature, to the extent that investors are willing to substitute between the assets.” (my emphasis)

So against this theoretical backdrop, the key questions to ask when contemplating “QE, The Sequel” are two: What assets should the QE program target in order to be effective? And, critically, who should be buying those assets?

To answer the first question, we have to have in mind the endgame, which is the desire to boost aggregate demand. In other words, the true metric for success of an LSAP program is not whether it managed to lower the yield of the security targeted (e.g. mortgage rates) but whether those lower yields translated into a material increase in aggregate demand.

By that metric, the Fed’s past LSAPs have probably fallen short. Clearly, measuring the counterfactual is impossible, but there are reasons to believe that the impact on aggregate demand was small. Why? First, because the reduction in mortgage rates boosted refinancings only by people who could refinance—i.e. people with jobs and some positive equity in their home. Not exactly the most cash-strapped ones who would have spent the extra cash.

Second, the portfolio-balance effect of the LSAPs on the prices of assets like corporate bonds or equities is at best weak, if not counterproductive. The reason (which I explained in detail here) has to do with the fact that US Treasuries and MBS are not “similar in nature” to corporate debt and equities. Unlike the latter, Treasuries/MBS have more of a “safe haven” nature—so that removing them from investors’ portfolios create demand for more “safe” assets, rather than boosting the prices of equities, high yield bonds, etc.

The bottom line here is that, while the LSAPs may have helped boost the cash position of certain financially healthy households and corporates, (a) they were not targeted to achieve a big bang for the buck; and (b) they have failed to boost agents’ risk-taking behavior—in the form of stronger consumer spending or a meaningful pick-up in employment.

Put differently, the LSAPs in their first incarnation failed (as they did in Japan) to remove the right type of risk out of the market. So what should a sequel target then?

The answer is assets in the riskiest part of the portfolio spectrum—small business loans, foreclosed properties, toxic credit card debt, and so on. Now, before I have hundreds of copies of the Federal Reserve Act thrown at me, let me touch a bit on the logistics.

Logistic #1: The “QE” operation should not actually purchase the small business loans or foreclosed properties themselves from the banks. This would be a logistical nightmare for the Fed (which would have to administer those loans), as well as giving rise to unmitigated moral hazard (why would I ever pay back my credit card debt, if the Fed stood ready to buy it off for free?) Instead, the QE operation should aim at injecting capital to banks against mark-downs on existing distressed loans.

Logistic #2: Clearly, this walks and talks like a fiscal operation, right? Well, it is a fiscal operation, with winners and losers, capital allocation to “chosen” institutions and with costs to the taxpayer. As such, the Fed is not the right institution to be in charge—it should be the US Treasury instead. So what is the role of the Fed then?

Enters Ben Bernanke, 2002:

“In practice, the effectiveness of anti-deflation policy could be significantly enhanced by cooperation between the monetary and fiscal authorities. […] If the Treasury issued debt to purchase private assets and the Fed then purchased an equal amount of Treasury debt with newly created money, the whole operation would be the economic equivalent of direct open-market operations in private assets.” (my emphasis)

Call it debt monetization par excellence. You set up a “special purpose vehicle”, which funds targeted capital injections to banks with Treasury securities. The Fed then purchases an equal amount of Treasury debt, funded by bank reserves.

The advantages of this approach are that (a) it would remove the right type of risk out of the market, in effect accelerating the “deleveraging” of the economy; and (b) it would distance the Fed from the credit allocation business. A key disadvantage of course is that many people out there (myself included) loathe the idea that irresponsible bozos would end up getting bailed out of their mortgage debt with taxpayer money.

But this is precisely why “QE, The Sequel” (and any QE for that matter) should be the explicit responsibility of the fiscal authority—i.e. an elected governing body: So as to allow voters to ultimately decide the kind of path their economy should follow. This may sound like lunacy in the current political climate, but maybe another quote from that same Bernanke speech can help strike an optimistic tone:

“In short, Japan's deflation problem is real and serious; but, in my view, political constraints, rather than a lack of policy instruments, explain why its deflation has persisted for as long as it has. Thus, I do not view the Japanese experience as evidence against the general conclusion that U.S. policymakers have the tools they need to prevent, and, if necessary, to cure a deflationary recession in the United States.”

So there.. a chance for America to show that's it's not going to become Japan.


Anonymous said...

"the portfolio-balance effect of the LSAPs on the prices of assets like corporate bonds or equities is at best weak, if not counterproductive"

Wait, aren't corporate spreads at record lows? The high yield market is booming; REIT's are easily refinancing long-term debt due; emerging markets bonds are in demand, and most of the muni market is also stable. The equity markets are up more than 50% from the lows, with small caps -- a measure of risk appetite -- leading the charge.

If the argument is a "rush to safety" is constraining credit growth, please present some evidence outside of private-label MBS. You may have a point in small business lending, although this is often a function of the need for more collateral. Unless you can levitate home prices or are arguing for unsecured small business lending(!!), extending more credit to the sector would be difficult.

Anonymous said...

BTW, you also argue, "(b) they have failed to boost agents’ risk-taking behavior—in the form of stronger consumer spending or a meaningful pick-up in employment." The latter part is clearly true. But on the former, where is the evidence that households are not "taking risk"? The recent savings rate of around 5% is, again, not indicative of abnormal risk aversion.

The set of arguments sounds suspiciously like, "savings are not low enough; spreads are not low enough; and leverage, house prices and equities are not high enough." Ironically, "enough" is defined as whatever they level they last achieved when we were creating the very problems you are trying to fix.

bb said...

a couple of follies in this theoretical work:
- QE boosts the valuation of assets by bidding their prices, it has no impact whatsoever on the value of the assets. thus the effect is only temporary and generally distortionary to market activity.
- you are assuming that the holders of trash assets (the ones to be sold to the Fed) are not simply fools, but morons: they will accept a small loss and through a 'portfolio-balance effect' will bid up the prices of other dud assets. well, so far the former holders of Freddie&Fannie's paper did purchase only treasuries, not junkier assets.

isn't it fascinating: the fantasy land of accademia in so many instances simply contradicts with common sense.

Anonymous said...

Nice article. But Bernanke, as an academic, severely underestimated the political and institutional impediments to implementing unorthodox monetary policy.

Your approach may run into the same problems.

Nick R.
Kyoto, Japan

bb said...

i am finding it a bit difficult to nail it, so i will ask: are you a socialist, fascist or feudalist in your beliefs?
it looks you are preaching something of the sort of 'take my losses, pay my loans' sort of thing, but at the same time you are in support of a system that can forcefully collect all dues and does not promote fairness or personal responsibility.

it certainly is not capitalism, bankruptcy is not an option you would ever consider as a natural market event in the case of a failure.

GF said...

I believe the aim of delivering the private sector should be the goal.

But ... " as well as giving rise to unmitigated moral hazard (why would I ever pay back my credit card debt, if the Fed stood ready to buy it off for free?)"

I ask what moral hazard?

We have fraudulently inflated mortgages induced by a predatory financial community. The FBI warned of this fraud as early as 2004.

Chevelle said...

Ok, a few clarifications here in response to some comments/emails I received:

First, as I said at the beginning, I am not calling for more QE—I’m only explaining what type of QE would work best in boosting aggregate demand, *if* it happened. In my view, the case for more Fed action at this juncture rests less on compelling evidence that the economic outlook is inconsistent with the Fed’s medium-term inflation target, and more on “risk management” considerations (ie kill the risk of deflation, even if that meant erring on the side of inflation). Tomorrow’s FOMC statement will hopefully throw more light as to the Fed’s collective thinking.

Second, when it comes to the SPV: Think of it as a “private” equity vehicle whose assets are NOT the non-performing business/credit card/mortgage loans, but stakes in the beneficiary banks. (The liabilities are US Treasuries, bought by the Fed). There are at least two advantages with this approach: First, the administration of these loans remains with the banks, which (are supposed to) have the expertise in collecting them. This preserves the incentives of borrowers to repay. Second, to the extent that banks have been mismanaged and are demonstrably useless in administering and collecting loans, the SPV’s management, as a shareholder, will have the power to replace the banks’ managers or arrange for their takeover by other banks. This will help impose market discipline to those banks that failed to do their job properly.

Third, when it comes to impact of the LSAPs on corporate credit spreads, etc: As hard as it is to prove, my view is that the rally we saw in risk assets since March 2009 was predominantly driven by (a) the removal of the negative tail risk in the return distribution of risky assets, as a result of government backstop measures to the financial system (recapitalization of systemic institutions, bank debt guarantees and so on); (b) positive surprises in the economic data last year (remember those “green shoots”?), as forecasters’ expectations had been driven to the abyss; and (c) a commitment by the Fed to keep interest rates at near zero for an extended period. The impact of LSAPs against this backdrop was at best marginal (except for the MBS spreads of course).

Anonymous said...

The point about low credit spreads is not to say that QEI was successful == I agree its impact was unclear. Instead, the point is to dispel the view that "risk aversion" or a "flight to safety" is constraining AD. If, in fact, risk aversion is not a feature of the financial markets, then what, exactly, is the purpose of the Fed lending to bank SPV's? In other words, you are ostensibly trying to fix a problem of private actors' unwillingness to take risk, and yet evidence of risk appetite abounds. Perhaps I am misunderstanding the mechanism through which the SPV will actually repair a specific problem.

Chevelle said...

The point of the SPV would be to accelerate the deleveraging of the private sector--that segment of the private sector (small businesses, households in foreclosure, etc) that has not benefited from the loosening in financial conditions we've seen in the capital markets. You fast-forward the losses to the banks, recapitalize them, change management if needed and move on... And so do the households and businesses

Anonymous said...

I see that you are truly making a fiscal case -- this or that sector needs help from the government. I think this is intellectually honest. Most interventionists try to make their case in the context of the need for more risk taking across the economy. This is why the rely on such concepts as a "shortage of safe haven assets" --- a questionable construct.

So kudos for calling intervention what it is: fiscal policy. The problem is aggressive fiscal actions --anything that requires Congress to vote--are now a political non-starter.