The Fuss about Market Liquidity

The recently released PwC “Global Financial Markets Liquidity Study”, sounds a warning.  Financial regulation, while perhaps well-intentioned, has gone too far.  Banks may be safer but markets are more fragile.

At the moment, this fragility is masked by the massive liquidity operations of world central banks.  But it will soon be revealed as, led by the Fed, central banks attempt to exit.  Now, before it is too late, additional regulatory measures under consideration should be halted (Ch. 5).  And existing regulations should be urgently revisited with an eye to achieving better balance between two social goods, financial stability and market liquidity, rather than the current focus on stability at the expense of liquidity (Ch. 3).

The bulk of the report consists of market-by-market empirical documentation of the reduction in market liquidity in past years (Chapter 4, pp. 51-104).  Pretty much all markets have been affected, even sovereign bond markets, but especially markets that were already not so liquid.  “There is clear evidence of a reduction in financial markets liquidity, particularly for less liquid areas of the financial markets, such as small and high-yield bond issues, longer-term FX forwards and interest rate derivatives. However, even relatively more liquid markets are experiencing declining depth, for example US and European sovereign and corporate bonds” (p. 104)

“Bifurcation”, meaning widening difference between vanilla markets now supported by central clearing and everything else, is a repeated watchword, as well as “liquidity fragmentation” across different jurisdictions.  Both are taken to be obvious bads.  But are they?

The central analytical frame of the report is that market liquidity is always and everywhere a good thing, and that more of it is always and everywhere better than less.  “We consider market liquidity to be invariably beneficial” (p. 8, 17).  “We consider market liquidity to be beneficial in both normal times and times of stress. For this study we therefore work on the premise that market liquidity is invariably beneficial” (p. 23).  Accept this premise, and everything else follows.  But why accept the premise?

To be sure, economics quite regularly adopts the simplifying assumption that all markets are fully liquid, so that supply always exactly equals demand and markets always clear.  (On page 17, the report cites the venerable Varian microeconomics text as authority.)  It’s a good assumption if you are concerned about something other than market liquidity.  It is a terrible assumption, and a terrible premise, if you are concerned exactly about market liquidity.

In fact, the idealization of full liquidity in every market is logically impossible in a world where market liquidity is provided by profit-seeking market makers.  In such an ideal world, market-making profit would be zero, so no market-maker would be willing to participate! The idealization thus makes most sense as a world where liquidity is provided for free by government.  It is thus quite inappropriate as a measure of how far current reality falls short of optimum.

The report complains that regulators have addressed each market in isolation, without considering the overall impact of all the regulations added together on markets as a whole.  But the report itself commits exactly the same analytical error.  It thinks about market liquidity as something supplied by individual profit-seeking dealers, along the lines of the famous Ho and Stoll (1981) model (Ch. 2).  But it never asks about the properties of the system comprised of many such profit-seeking dealers competing with another.  They have a dealer model, but they have no money view.

What they call bifurcation, I would simply call hierarchy or tiering, which is a natural and ubiquitous feature of all credit systems.  What they call fragmentation, I would simply call essential hybridity, different local balancing of money interest and public interest.  Maybe there is too much bifurcation and fragmentation, but maybe also there is too little; that’s what we need to discuss.  What we can say for sure is that zero bifurcation and zero fragmentation is impossible.

The financial crisis made clear to everyone that the existing regulatory apparatus was inadequate for the emerging market-based credit system, as well as the larger financial globalization trend.  In the regulatory response that followed, we shifted the matched book dimension of market-making substantially to central clearing counterparties, and the speculative book dimension off of the balance sheets of banks (the Volcker Rule).  This shift  was not an inadvertent mistake, but rather a deliberate attempt to separate the liquidity risk of matched book (which arguably requires and merits public backstop) from the solvency risk of speculative book.

That said, it is clear that we are feeling our way into the future, all of us.  The crisis revealed to everyone that abstraction from liquidity is abstraction from an essential feature of reality.  Markets have already responded by pricing liquidity more explicitly, appreciating that, like any scarce good, we need to make sure it is allocated wisely.  The question is not whether market liquidity is “invariably beneficial” or not, but rather whether the social benefit is greater than the social cost.