‘Ordinary times’: Bagehot, liquidity premiums, and the window May 27, 2012 at 9:37 am

FT Alphaville had a good post last week, referencing a new paper by Brad DeLong that in turn discusses the early history of central bank liquidity provision. One paragraph in particular gave me pause for thought:

Bagehot never argued that central banks should lend only against good collateral. Bagehot was subtler than that, and less exclusive. What he argued was that central banks should lend against collateral that would be good “in ordinary times”.

What does this mean? It’s not entirely clear, but I would suggest that a modern version of this doctrine is that as lender of last resort the central bank should advance sums against collateral which are conservative given the expected eventual realised value of the collateral, but it should not consider the market value. Or, to put it another way, the Central Bank should ignore the non-default component of the credit spread in assessing collateral haircuts. That’s a pretty good formulation of the lender of last resort function: central banks should be willing, for a solvent institution, to keep it liquid long enough for it to be able to monetise the spread between assets and liabilities (or for the market to come to believe that that spread is positive, and so be willing to fund the bank privately).

2 Responses to “‘Ordinary times’: Bagehot, liquidity premiums, and the window”

  1. “the Central Bank should ignore the non-default component of the credit spread in assessing collateral haircuts”

    Would that actually be enough? In a balance-sheet driven recession the actual default component would broadly rise across the board, meaning the eligible collateral at the central bank would fall sharply. The central bank would therefore be pursuing a pro-cyclical policy, being least accommodating in supporting banks when the problem is systematic but more accommodating when supporting a bank that is in trouble for idiosyncratic reasons. The graph in your linked post suggests that central banks following your doctrine would have been least helpful in April 2009, when the crisis was reaching a peak.

    My interpretation is that Bagehot is saying that if a central bank wants to end a crisis it should proceed by making a public commitment to act as if no crisis was happening, and assess collateral on a ‘through-the-cycle’ basis. If this commitment is credible then it will help stop the rush for safe-haven assets. Inevitably this means taking on a degree of credit risk, something that seems to have become an anathema in the world of central banking, at least in their public statements.

  2. JH – I take your point, but I am not sure I would go that far, not least because of the Eurozone example. No central bank could with a straight face take sovereign bonds of a state that is close to a credit event. Therefore while I agree with you that there is a procyclical aspect, some measure of that is I think inevitable given the over-riding need for central banks to keep themselves (somewhat) safe.