Extreme Events – Specimen Answer A.9.3 –
[this page | pdf | references | back links]
Q. You work for a bank and
have become worried that different business units might be focusing too little
on the liquidity needs that their business activities might be incurring. How
might you set an appropriate ‘price’ for the liquidity that the bank as a whole
is implicitly providing to its different business units?
Liquidity risk is a risk that was arguably underappreciated
and under catered for in regulatory frameworks in the run-up to the 2007-09
credit crisis, see e.g. Market Consistency.
Some changes have been mandated or proposed since then by regulators seeking to
address some of the issues that were uncovered during this crisis.
However, relying exclusively on regulator mandated criteria
potentially risks failing to heed one of the lessons of this crisis. This is
that personnel (even up to Board level) may often be overly willing to design
business strategies around what might look good from a regulatory perspective
rather than what a more robust analysis of business needs might dictate.
The most effective way of setting a price for liquidity is
likely to be to identify what the ‘market’ would charge if the business line in
question had to source the liquidity externally. Any other value potentially
runs the risk of undercharging for the implicit support that the business unit
might be receiving from the corporate centre (if too low), or stifling activity
and potential business opportunity (if too high). Superimposed on this are
likely to be refinements set at a corporate level explicitly designed to reduce
or increase the tendency of each business unit to market products that
‘consume’ liquidity, to control the overall level of liquidity risk that the
firm is likely to be running.
In practice, this requires proxies capable of providing a
guide as to the amount of ‘liquidity’ that any particular product consumes.
Liquidity risk has the characteristic that it typically manifests itself with
low probability but high severity, i.e. painful outcomes disproportionately
correspond to outcomes in the tail of the distribution of potential future
outcomes. It may therefore be very hard to estimate precisely. However, this is
no reason not to try to do so; indeed if anything it makes the need to do so
even greater, even if the resulting proxies may need to rely more on stress
test outcomes and other sorts of scenario analyses than many other types of
One outcome of the 2007-09 credit crisis was the development
of a ‘term structure’ to liquidity risk, with interest rate swap rates
differentiating according to the ‘refresh’ frequency implicit in the swap in
question. Thus swap rates involving exchange of fixed for floating payments
became differentiated according to whether the reference rate underlying the
floating leg was overnight, 1 month, 3 month, 6 month Libor etc. (for the same
overall swap maturity date). The shorter the period between successive resets,
the easier it is for an investor to move his money away from a bank that
appears to be heading for default before it gets there. This could be used as a
way of identifying the market-implied price of the liquidity benefit that a
bank might gain from locking up (floating rate) funding sources for longer time
periods or might expect to be compensated for if it does the same in reverse.
Contents | Prev | Question