The Moral Hazard Effects of Recent Central Banks’ Liquidity Injections

The massive injections of liquidity by the ECB, Fed and other central banks have continued today.  Whenever there is a seizure of liquidity in financial markets central banks face a tradeoff: they are the only potential providers of lender of last resort support to the financial system and thus the only ones to be able to stop a liquidity run; but, like in any form of insurance, such liquidity provision may lead to moral hazard, i.e. the risk that reckless investors will expect to be bailed out (see the WSJ article today on "Fed Treads Moral Hazard").

To reduce such moral hazard the famous British economist of the 19^th century, Walter Bagehot, suggested that such liquidity support should be provided at a penalty rate and against good collateral. History also suggests that liquidity support is more effective and justified if the economy suffers of a liquidity crisis rather than an insolvency one: if many agents in the economy are insolvent trying to bail them out will create moral hazard, will not solve the problem of insolvency and will postpone the necessary insolvencies and create a bigger set of bankruptcies down the line. Moreover, the lessons of international financial crises in emerging market economies suggests that international lender of last resort support is more effective – and less prone to moral hazard – if it is “conditional” on the troubled borrower taking economic and financial actions that reduce its fundamental financial problems: this is what is called “IMF conditionality”, i.e. lending against macro and financial reforms. IMF support is also limited in its amount – rather than being unlimited - to reduce the moral hazard distortions.

So a sound lending of last resort support that minimizes moral hazard requires:

- an illiquidity rather than insolvency problem

- penalty rates

- good collateral

- conditionality in the lending support

- limits to the amount of support

On all those dimensions the current injections of liquidity by the Fed and the ECB partly fail the test of sound lender of last resort support.

Concerns about a “Bernanke put” and “Helicopter Ben” have been – mostly unfairly – in the minds of investors. The Fed has been criticized in many quarters for feeding the tech bubble of the 1990s, then easing liquidity too much and for too long after the tech bubble burst, thus creating another housing and credit bubble that is now bursting. So, liquidity injections – however justified by the recent liquidity seizure – bear the risk of causing moral hazard and bailing out investors that speculated recklessly. The Fed and ECB – to their credit - have not – so far – cut policy rates but rather injected liquidity to prevent a spike in short rates. But the size and forms of these large liquidity injections has certainly increased moral hazard in a number of ways.

First, as I have discussed in a previous blog we are now partly facing an insolvency crisis, not just a liquidity one. With many households, mortgage lenders, home builders, hedge funds and other leveraged investors, and some non-financial corporations being financially distressed and near insolvent, not just illiquid, a generalized liquidity injection is not justified. Reckless borrowers and investors should suffer serious pain – and insolvency if needed - to prevent future reckless behavior; but the Fed coming to the rescue will not prevent the eventual default of insolvent agents and will feed expectations of further bailouts.

Second, the liquidity support did not occur at penalty rates and in limited amounts (see also De Grauwe’s oped in the FT today on this point). The Fed could have provided tons of liquidity at a higher discount rate at the discount window; it instead decided to provide unlimited liquidity at an unchanged discount rate.  Some penalty rate and more limited liquidity injections would have punished those borrowers that behaved recklessly and did not have buffers of liquidity.

Third, the lending did not occur in all cases against good collateral. What was particular bothersome about the repo operations that the Fed undertook last week on Thursday and Friday and again this Monday morning is that the Fed accepted mortgage backed securities (MBS) as collateral against the lending. The Fed did not tell which MBS it accepted as collateral and at which prices (face value when markets are now pricing discounts even of highly rated MBSs?) Did the Fed accept junk rated MBS? Maybe yes, maybe not. The reality is that we do not know. But the perception of having used risky and now very illiquid instruments such as MBSs – rather than only safe liquid Treasuries - for open market operations goes against the Bagehot principle of good collateral.

Fourth, there was no conditionality in this liquidity support: any bank that needed liquidity could get as much as it wanted at unchanged rates. Actually, of the three interventions on Friday, apparently, the most effective one was the third and last one when the Fed pushed the Fed Funds rate – for a short period – down to 5%, i.e. below its target of 5.25%. A more selective use of the discount window – as the Fed does in regular times – to limit the support given to institutions that had imprudently low levels of free reserves would have been more appropriate. The Fed – whenever it wants – can use its moral suasion tools to limit the support given via the discount window to institutions that are imprudent. None of that moral suasion limited support apparently occurred this time around; rather unlimited liquidity provision.

The Fed had little choice – given the liquidity squeeze in the market – to intervene and provide some lender of last resort liquidity support. But the way such support was provided has increased moral hazard distortions rather than kept them under check. Now financial market participants can happily expect that the Fed will provide unlimited support at unchanged interest rates whenever liquidity crunches occur. This is not a good signal for the long term stability of the financial system: it reduced a short run liquidity crunch at the cost of increasing medium term moral hazard related expectations of bail-out of reckless investors.

Update on Collateral:

I am being told that the Fed accepted three types of securities in these repos: safe Treasuries, agency (GSEs) debt and mortgage backed securities (MBS) guaranteed by Fannie and Freddie. These latter are not subject to credit risk given the guarantee; but can still technically default if the underlying assets are impaired. Still, there are still anomalies in this use of guaranteed MBS in repos: the rate at which the Fed accepted these MBS was apparently the same as the rate at which it acccepted safer Treasuries and Agency debt. This, in turn, implied that mostly MBS were offered and used in the repos; especially on Friday all submitted and accepted securities were MBS and in the last intervention the average rates on this MBS repos was below the Fed Funds rate. Even today about half of the accepted collateral was MBS. See Kevin Drum for more details.