Sunday, 20 March 2016

Liquidity and central bank policy

These things a twenty first century central banker knows:

The corporate world is gotten more complex since the advent of financial engineering.  The continuous expectation of investors to maximise risk adjusted returns leads to increased chances that the firm might succumb  to the temptation of gearing, with the result that the firm goes out of business.  Not all firms are exposed evenly to this risk - it clusters, and those industries most at risk are those with funding and asset mismatches.  This is almost the definition of the base business model of retail banking sector.  So banks have to set up shop at the foot of the volcano.  That's their job.  And those banks all have to do this, to a greater or lesser extent.

If a central banker want to avoid systemic risks, he tries to put in place measures which address this risk.  Perhaps a demand for higher capital buffers (some of which capital operates a s a liquidity buffer).  But banks respond to these capital requirements by grossing down their balance sheets rather than taking the direct hit to their share price as a result of the likely reduction in equity returns which happens when the regulator asks banks to sit on more capital.  

In the limit, the banks become not only heavily regulated by government, but the key mechanism for allocating credit (capital) to those parts of the economy that need it becomes a quasi-government function.  Bank returns become largely policy driven and the financial services industry starts to resemble a government department with ludicrously paid employees.  It is not that regulator imposition has unforeseen effects on liquidity of firms and systemic liquidity per so, it is that these banks need to place their business at the foot of the volcano in the first place.  You can't legislate geography away, so to speak.

When firms go bust, they often go through an illiquidity phase on their way to extinction.  This must continue to happen for Schumpeterian reasons, so policy setters need to distinguish when to act and when not to act.  This determines their actions vis-a-vis their role as final liquidity providers - i.e. lenders of last resort.  But if you do this too generously you end up with an economy of zombie corporations - Japan has faced this situation for decades.  And if you don't do it too actively, you enable avoidable contractions and recessions.  The ease with which various central banks pull this trigger largely drives the modern discussion of progressivism versus the Austrian or neo-classical approach in politics. 

Finally, even if central bankers and policy makers decided that they'd run the risk of forcing banks to hold dramatically more capital, as a kind of ground zero solution to the negative consequences of the inherently systemic nature of banking, the industry would migrate increasingly to the shadow banking sector.  This is already happening to some degree - witness the birth of so-called peer to peer lending.

Liquidity in context IV - the life of a de facto corporate liquidity manager

This posting is about dumbing down liquidity management in language which  most people can easily understand and relate to.  Liquidity management is mostly about the maintenance of good operational cash flow balances to cover the expected and predictably unexpected vicissitudes and seasonalities of corporate life.  There was a time not that long ago (up until the 1960s) when operational cash flow management was a private little secret of the treasury department.  They skimmed some free cash flow from operations and kept a store of it to meet more or less expected corporate cash call events.  When looked at this way, you suddenly realise that financial demands can in theory be a lot more predictable than operational events.  It is more certain knowing when you need to repay your bonds than it does when you'll need to pay for repairs to an uninsured industrial accident.  

These events in question (cash calls) each and every one of them can have an uncertainty added to the cash-amount and time schedule you'd normally think of as the definitive parameters.  If you could model all cash call events somehow, then the aggregate cash schedule and its concomitant variance would feed into a pretty decent corporate liquidity management model.  At its most fundamental, these cash calls are modelled as call options on zero coupon bonds.  Each event will have its own notional value, volatility expectation, strike price.  When you aggregate this portfolio of real options up you've got your funding liquidity mostly modelled.  One must be realistic about just what fraction of the operating corporate environment is amenable to modelling, and also with respect to just how fast the situation could change.  The more chaotic the likelihood of change is, then the more difficult it is to extract value from a liquidity management regime.  Or to put this more dramatically, there's a level of chaos in the liquidity environment above which it doesn't make much sense to model liquidity.  What you modelled today becomes largely detached from which realistically could happen tomorrow.  In general, if the future state of some system is so unpredictable based on today's models, then those models aren't much use.

Why did liquidity management stop being a private skimming operation of the treasury department in the 1960s?  Partly because of the advances in financial engineering from the 1960s onward (Treynor) which paved the way for more sophisticated financial engineering at corporate finance departments.  Similarly, macro-economic  climate became incredibly volatile following Nixon's decision to end Breton Woods agreement, leading to currency volatility and destabilising inflation.  Corporate treasurers responded by bringing some basic financial engineering to the largely in-house management of corporate cash calls.  Finally, financial engineering was also focusing the minds of corporate executives at technology companies starting in late 1950s silicon valley, via the issuing of executive stock options, which accounting bodies valued as stock price minus strike - effectively ignoring the intrinsic value element (we had to wait for the Black-Scholes equation for that).  This caused them to tilt in favour of investment returns over (liquidity) risk.  In essence, to really manage a firm's liquidity so that there is always a sufficient cash buffer in the end detracts from short term investment gains.  Corporate executives, especially in 'innovative' technology companies were now personally incentiveised to maximise precisely these short term investment returns, in ways which used less capital.

What new tricks did they come up with?

How about taking those ideas in fixed income financial engineering and look to calculate the duration of cash flows with a view to matching those cash flowsOr perhaps finding some third party to write some liquidity options for you so you can have them as a form of cheap liquidity insurance.Or renegotiate the clauses around commitment in the contracts you have with banks over your loans.Or get loans from the capital markets, dis-intermediating your firm's normal pool of lending banks.Perhaps stealing that idea of Markowitz and paying close attention to the free lunch you can achieve through diversification - in this case, the diversification of your funding sources.Lastly, don't just have a pool of liquidity buffer cash sitting at a bank earning perhaps a negative real rate in times of high and volatile inflation, which not buy assets with this pool of cash, gaining a higher return which maintaining the average liquidity profile of the pool.