Best Practices for Liquidity Risk Management
In the previous articles we focused very much on the conceptual framework and
the causes of liquidity risk. We even suggested a term structure for liquidity
risk that will be developed further with more research.
This article focuses on best practices for liquidity risk management. After
all, integration into the existing risk management framework is essential because
only then can the connection between market, credit and liquidity risk be seen
on the full scale. Throughout the article we will see that there are analogies
in credit and market risk management, thereby making the integration of liquidity
risk management practices less complex.
Managing liquidity serves ultimately two purposes:
a) Optimize the cost / benefit for creating / placing liquidity
b) Keep the banking institution afloat
In credit risk management terms, the twin objectives are to minimize the expected
loss and increase the returns and to prevent the bankruptcy of the institution
due to credit risk issues.
Figure 1: Desired cost / benefit regimes for liquidity management
To keep the cost of liquidity low, one must consider both the cost of creating
liquidity and the opportunity cost involved in having excess liquidity. This
implies that a banking institution either wants to refinance short positions
within a certain price level or gain from long positions at an acceptable yield.
Fig. 1 refines this view further. The distribution shows the number of transactions
conducted to cover short positions or place money in the market. Three regimes
are defined that mark different cost/benefit profiles. The majority of transactions
should of course happen in the desired green area. The yellow area is to be
avoided if possible and the red area must definitely be avoided. There is no
need to restrict oneself to only three areas – any number that is felt to be
sufficient will do.
How does this picture tie into the liquidity at risk concept outlined in previous
articles? Usually, the aim of managing risk in the financial markets is to reduce
the volatility of a certain variable. In this context this can only mean reducing
the volatility of short and long positions from a treasury point of view. This
enhances the predictability of positions and hence lowers the risk of having
too much liquidity or not enough liquidity available. This should eventually
push the banking institution into the desired green areas shown in Fig.1.
In order to manage liquidity risk within a banking institution, the whole risk
management process must be institutionalized. This means that various practices
and processes within the organization must be established and executed.
Figure 2: Liquidity risk measurement is at the heart of liquidity risk management
Fig. 2 shows a high-level diagram of a liquidity risk-controlling framework.
At the heart of the framework is the liquidity risk measurement process, which
is based on the term structure discussed in a previous article.
The whole procedure starts with a data-gathering exercise for cash in- and
out-flows, similar to today’s procedure for determining end-of-day positions
at certain cut-off times. What is different, however, is that these flows are
categorized by four types of cash flows – from completely deterministic to the
various types of stochastic flows[1]. If the required data is not available,
one must determine whether it can be obtained indirectly through a mapping technique.
As with other types of risk management, data integrity is key to the whole gathering
process.
Based on this data, net funding requirements (and consequently funding gaps
or surpluses) can be calculated and mapped onto a maturity ladder. One particular
issue that has to be taken into account is contingent liabilities. These do
not constitute a part of the daily cash flow dynamics. However, they can be
quite substantial and should be part of the net funding requirement calculation
to obtain a more conservative picture.
Two other aspects are very important for calculating net funding requirements.
Business won or lost can have a substantial influence on the liquidity situation.
An estimate about business development should therefore be part of the net funding
requirements, especially for the longer dates on the maturity ladder. The second
aspect is credit lines. Despite an accurate calculation of net funding requirements,
the counterparties of these cash flows (usually nostro banks) must have sufficient
credit appetite, i.e. credit lines. If this is not the case, gridlocks may arise
where long positions in one place can’t be mobilized to cover short positions
in another.
Net funding requirements feed into liquidity risk measurement. At this point,
non-deterministic cash flows are modeled. Based on the term structure of liquidity
risk, Liquidity at Risk is calculated and funding gaps are determined. Also
important is the measurement process that will cover different time frames from
days to weeks to months and even years. The results of the measurement process
have to be inline with the risk policy and can be an important input for the
liquidity management process.
More importantly all results feed directly into the back testing procedure
that reconciles the actual net funding requirements against the predicted or
modeled requirements. Back testing is a crucial step (as in all risk management
procedures) in the whole structure because it validates the underlying assumptions
of the measurement process. Where there are differences between actual and predicted
gaps or surpluses the model has to be adjusted and the validity of the underlying
assumptions verified.
So far we can’t give more details on the reporting process, due to the
limited experience and data that is available. But from a generic point of view
it is clear that a tool set of reports will be necessary to satisfy the needs
of an individual trader, but also to give senior management a condensed information
set to make an informed decision.
The reports will very likely look at portfolios and gauge the liquidity risk,
much like market risk reports do today. Liquidity risk has to be understood
in much greater detail before we can outline specific types or reports.
We can however take a closer look at the basic principles of a liquidity risk
policy.
It is not enough to implement a technical process to measure and monitor liquidity
risk. The results are meaningless unless they are put into a context that is
specific to a particular banking institution. The question the liquidity risk
policy has to answer is: how much risk appetite does an institution have?
The risk policy should cover three basic aspects:
There are a variety of issues to be discussed and analyzed in creating the
appropriate risk policy. Here are a few highlights:
These questions highlight the fact that there is no such thing as a standard
risk policy. The framework has to fit the individual institution and support,
even enhance, the business.
The goal of managing liquidity (and therefore liquidity risk) is to optimize
the cost / benefit for short / long positions for the liquidity manager. We
suggested a framework based on pro-actively monitoring the costs and benefits
of the liquidity management process. This requires a solid risk controlling
procedure that is implemented on a firm-wide basis – just as credit and market
risk procedures are implemented today.
[1] Cash flows are basically determined by two variables: a) The time when
they occur and b) the amount. When time and amount are known then we call it
deterministic cash flows. Otherwise it is called stochastic cash flow. Altogether
four categories can be established. They are described in more detail in a previous
article.
