The government cannot be responsible for systemic credit risk

Brad DeLong‘s latest has me sputtering.  (He seems to have fallen for Caballero’s view that the government has to insure private markets against tail risk by insuring private assets — which I have addressed many times.  Note that DeLong already has a second post up on this topic.)

When there is excess demand for safe, liquid, high-quality financial assets, the rule for which economic policy to pursue – if, that is, you want to avoid a deeper depression – has been well-established since 1825. If the market wants more safe, high-quality, liquid financial assets, give the market what it wants.

The policy “well-established since 1825” to which deLong refers is the Bank of England’s practice of lending generously into a financial panic.  Unfortunately, the historical episodes to which deLong refers are not really comparable to our current situation.

(i) Bankers had personal liability for their debts in England — and because the par value of shares was rarely fully paid up by investors this was effectively true for joint stock banks too.  Thus, when England’s financial system was shaken to its core by the failure of Overend and Gurney (roughly comparable to Lehman Brothers), not one creditor lost a dime.  The panic that ended the bank had been triggered by the sale of the Gurneys’ personal assets to honor debts that they had guaranteed before selling shares.  Even so, shareholders were forced to put in additional funds to honor the firm’s remaining debts.  (Ackrill and Hannah, Barclays: the business of banking, p. 46.)

(ii)  The combination of personal liability and capital calls on shareowners was an effective preventative against “moral hazard” for most of the 19th century in England:  the banking system as a whole simply didn’t issue bad debt — though individual banks could make mistakes, or be mismanaged.  Thus 19th c crises were systemic liquidity crises, not systemic solvency crises.

(iii)  The strongest evidence that the 19th c crises were liquidity, not solvency, crises is that the central bank actions were almost always completed, such that markets had returned to normal, within three months.  The idea of providing extraordinary liquidity over a period of more than a year as the Federal Reserve has done — or worse the practice of allowing the long-term instability of the financial sector to drive interest rates to zero for years — would have been unthinkable.

(iv) In short, the 19th century environment where borrowers were held to extremely high standards is not comparable to our moral hazard ridden financial system.  For this reason, DeLong’s appeal to ancient truths is misguided.

However, the assertion that really set me off is the following:

Creditworthy governments around the world can create more safe, liquid, high-quality financial assets through a number of channels. They can spend more or tax less and borrow the difference. They can guarantee the debt of private-sector entities, thus transforming now-risky leaden assets back into golden ones. Their central banks can borrow and use the money to buy up some of the flood of risky assets in the market.

Which of these steps should the world’s creditworthy governments take in response to the asset-price movements of May? All of them, because we really are not sure which would be the most effective and efficient at the task of draining excess demand for high-quality assets.

I find it ironic that DeLong compares a government guarantee of private debt to alchemy, because those of us who are concerned about the government over-reaching in trying to support the value of intrinsically flawed private sector assets are concerned precisely because we suspect that it a project that is doomed to failure.  While it is certain that governments can spend their resources directly and indirectly reflating private sector asset bubbles, it is far from clear that anything good will come of doing so.

When Bagehot argued that the Bank of England should not be overly discriminating in the bills it purchased, he did so explicitly because he knew that the British financial system was sound and produced only a tiny fraction of bad assets.  Unfortunately the modern US financial system does not meet this standard.  What good can come of putting a government guarantee behind debt that is sure to default?  What reason is there to believe that delaying a default by a year or two or three or four is in the interests of either the debtor or the creditor?

The answer presumably is in the second post:

The hope is that, by Walras’s Law which tells us that excess demands across all markets must sum to zero, that relieving excess demand for AAA assets will produce as a consequence the relief of excess supply and full-employment balance in the markets for goods, services, and labor as well.

The problem with this answer is that it is purely aspirational.  Somehow the government purchase of bad debt is supposed to return the economy to the growth path that existed when the bad debt was being issued and nobody realized how unreasonable expectation of repayment really was.  That is, instead of recognizing that the boom times were just that and that the economy has no choice, one way or another, but to shift to a more sustainable growth path, the formula promoted by  DeLong and Caballero is for government to support asset prices until either (i) growth returns and my view is proven wrong or (ii) the government is no longer capable of supporting asset prices.  My real concern is that neither DeLong nor Caballero take the possibility of (ii) seriously — they are unwilling to consider the possibility that government does not in fact have the capacity to levitate the whole economy.  They don’t appear to have a plan b after we implement their recommendations, and it fails.  In fact, Caballero explicitly assumes the success of his proposal (my emphasis):

Instead, if the government only provides an explicit insurance against systemic events to the micro-AAA assets produced by the private sector, we could have a significant expansion in the supply of safe assets without the corresponding expansion of public debt. Of course there would a significant expansion of the notional liabilities of the government, but it is nearly certain that the ex-post cost would be much less than in any of the real alternatives.

To which I can only say that I agree with the anonymous blogger at Macroeconomic Resilience that Jon Stewart has demonstrated a remarkably accurate diagnosis of our current problems:

Why is it that whenever something happens to the people that should’ve seen it coming didn’t see coming, it’s blamed on one of these rare, once in a century, perfect storms that for some reason take place every f–king two weeks. I’m beginning to think these are not perfect storms. I’m beginning to think these are regular storms and we have a sh–ty boat.

When perfect storms (i.e. systemic crises) are taking place with ever increasing regularity, maybe it’s not a good idea to sign the government up for systemic risk insurance.

In short, I’d sputter much less if DeLong and Caballero would spend a significant amount of their time addressing the problem of bad debt and how to deal with it.  Remember that the US government has already spent almost two years underwriting mortgage refinances at extremely low rates that do exactly what DeLong and Caballero recommend.  The remaining mortgage borrowers are simply not good credit risks (look at the back end DTI here).

The evidence all points to the fact that what we have experienced is a systemic solvency crisis, that we are in a balance sheet recession and that the only way out is to rebuild private sector balance sheets.  There are two basic methods to do the latter:  (i) shock therapy, where bankruptcy wipes out debt and transfers assets to creditors and (ii) decades long stagnation, where private sector debt is rolled over repeatedly — allowing debtors to avoid bankruptcy, pay mostly interest and very slowly pay down their debt.  As long the debt is not wiped out by bankruptcy (or inflation), an economy in a balance sheet recession simply cannot flourish.  Pretending — or praying — that the economy will flourish despite the debt overhang, which seems to me to be DeLong and Caballero’s plan, amounts to extreme risk-taking of the first order and is not the domain of sound governance.

While I am extremely strongly opposed to the government getting into the business of underwriting private credit risk, I do believe there is an important role for fiscal policy in the current crisis.  Supporting the economy by protecting the jobs of municipal and state employees — especially teachers — and programs to help the unemployed, etc., seem to me to be just plain common sense, given the economic straits in which we find our economy today.  Keeping a cap on the federal deficit by allowing children to go uneducated or hungry has never, and will never, be good policy.

On the other hand, I’m a bit of an apostate when it comes to the role of monetary policy at the current juncture.  I think we will get out of the “liquidity trap” sooner if we give savers who want to put their money in safe assets a small return on their funds.  Instead of keeping interest rates at the zero lower bound, and pushing money managers into unreliable and unworthy risk assets, the Fed should raise rates to 1 or 1.5%.  Also we should not ignore the benefits of the bankruptcy process in allowing firms and individuals to get a fresh start while transferring assets to creditors who are not crushed by debt.

At the same time aggressive fiscal policy needs to be used to offset the more nefarious consequences of these policies.  But the overall task must be to help the economy find its new sustainable growth path.  Because zero rates are most definitely not part of this new path, they really only succeed in creating new distortions.  The Fed should give the economy a sustainable baseline from which to work and allow market forces to sort out the details, with relatively generous support via fiscal policy.