Liquidity At Risk - Managing Future Cash Flows With Realism

 |  Includes: XLF
by: Suresh Sankaran


Liquidity At Risk – A stochastic look at cash flows.

Liquidity at risk is a concept that evaluates cash flows on a dynamic basis.

Customer behavior analysis plays an important role in liquidity management.

This is a novel approach to modeling liquidity through a stochastic process.

Introduction & background

Liquidity risk led an unremarkable existence up until 2007-8, when suddenly, all regulatory hell broke loose. From being a risk that no one talked about, it became highly regulated over a period of 4-5 years. What changed?

Every financial institution collapse added to the mystique of liquidity risk; experts opined, 'alas, that organization did not have adequate liquidity to foot expenses', and this added to the furor surrounding the lack of regulation around liquidity risk.

A risk that was measured through prosaic ratios like loans to deposits is now being subject to haircuts, 3-notch downgrades, run on deposits, concentration of funding and so on. How did this come about?

The answer to both questions lies in the fact that there is now a belated acceptance of the fact that the end product of every other risk in the marketplace in financial terms is a lack of liquidity. It has been clearly established that when customers default, it results in liquidity risk; when there is a fraud within the organization, it impacts liquidity, when there is funding concentration, there is clear evidence of illiquidity if the funders do not renew credit lines, and when markets change, it changes the liquidity profile of an organization.

Liquidity is a second order risk, and one does not manage second order risks without adequately monitoring, measuring and controlling primary risks. If an organization controls credit risk, it is, in part, controlling liquidity risk.

The Financial Select Sector SPDR (NYSEARCA:XLF) presents an interesting investment scenario in the current market environment and therefore, it is absolutely crucial that the liquidity of this sector (in the context of the banking industry as a whole), is better understood before making any investment decisions. We refer to further analysis (like Dr. Donald R van Deventer's) for other risk analytics before any investment optimization decisions are considered.

The speed of illiquidity

As a risk, liquidity impacts organizations more quickly than any other. The impact is felt most and furthest from a cash-flow standpoint, and this makes liquidity very difficult to manage in isolation. Moreover, the change from being liquid to being illiquid is debilitating and rapid. This means that organizations are unable to cope with the speed with which liquidity deserts them, and fail to survive eventually.

The time between liquidity and illiquidity is rarely over 90 days, and this is the time it took for Northern Rock from being a poster child for creativity to being the first English institution in over 150 years to suffer the ignominy of a bank run. This therefore, makes laughable the regulatory assertion that liquidity models should be constructed over a longer term. If at all liquidity has to be managed, it is in the short term.

Liquidity perception

The problems for an organization stem from the fact that historically, liquidity has been regarded as a matter of compliance, and therefore, subjective to a great extent. 'Highly liquid assets' are sought to be held, without realizing that what matters is the markets' perception of what constitutes highly liquid assets and what does not. Just because a regulator signifies sovereign holdings as liquid assets, does not ensure that Greek government debt is any more liquid than any other risky asset holding. This was brutally reinforced during the spectacular meltdown of Iceland and Greece, and it is unfortunate that both practitioners and regulators seem to have learnt little from these mishaps.

The importance of cash flows

It is now well understood that ratios matter little; the lifeline for liquidity management is a better understanding of cash flows, and these need to be subject to all the primary risks that a financial institution has undertaken. For instance, if an American company has taken on in its portfolio Korean Development Bank bonds, it has taken on, apart from counterparty risk:

  • Sovereign risk
  • Foreign exchange risk if the bonds are in Korean Won
  • Interest rate risk, irrespective of whether the instrument is fixed or not; if fixed, changes impact value and if floating, impacts income. In a post IFRS-9 world, everything impacts the balance sheet
  • Korean equity risk, the risk of Korean markets impacting the price of the development bank bonds
  • Transfer risk, arising on account of Sovereign risk, where the counterparty is not in default, but is precluded from foreign exchange remittances on account of sovereign controls
  • Operational risks, which largely are internal to the investing entity, but any operational lacuna in the Korean Development Bank will have an impact on price and yield

All of these risks have a direct liquidity impact, and therefore, if these primary risks are managed, liquidity will be managed automatically. There is no reason to subject an organization to additional liquidity measures. Each of these risks can be hedged individually but that constitutes nothing more than risk transfer. To construct a combined hedge that takes into account all the above-mentioned factors would be ruinously expensive.

Why cash flows? Because from a contractual standpoint, there is a clear pool of future cash flows that is expected if the markets are functioning, if the counterparty is solvent and liquid, and if all operational controls are present; if not, there is going to be a change to future cash flows, and these can be modeled to better understand the changes wrought on account of each risk.

Each scenario that is constructed from a risk management perspective changes the cash flow profile of this investment, and presents a change in the cash flow profile expected by the investor bank. As markets change and as the creditworthiness of Korean Development Bank changes, the impact is felt by the cash flows that reflect market perception of the probability of default of this institution, and this default may be on account of any of the risks enumerated above.

We can take this construct further to understand all possibilities of deterministic cash flows, where the cash flows are impacted by macro factors, market factors, and idiosyncratic counterparty behavior which changes the cash flow profile, and if these changes can be handled by the counterparty, then you will get a changed cash flow profile, and if these changes are too much, then the counterparty defaults on its payments.

Risk interrelationship

A simple example illustrates the fact that risks are interlinked. An employee with IBM in the US takes a mortgage for USD 800,000 on a home in Washington DC for 30 years, fixed at 4.5%, for example. This essentially means that every year, this employee repays a portion of the principal and pays interest on the outstanding amount with the bank or the credit union. Let us suppose that interest rates go up by 50 basis points, and this means that on average, the interest payments per annum will increase by USD 30,000. At last review, I daresay that IBM's salary increase policy did not cover interest rate increases (as would be the case with most Corporates!), and what is certain is also the fact that the expense profile of this customer will at least remain the same if not increase in line with interest rates. This essentially impacts the creditworthiness of this customer, and if he or she can take the additional USD 30k cash burden then the interest profile of this customer changes, and if he or she cannot, then default occurs.

All risks are interconnected and interrelated and the smart risk manager applies a holistic structure to manage cash flows and thus, manage liquidity. By managing primary risks more effectively, you can manage liquidity risk, and any attempt to understand liquidity risk in isolation is a folly of gargantuan proportions.

From deterministic to stochastic

What are we attempting to achieve? A more structured approach to the understanding of liquidity would be to first structure the behavioral cash flows as your base case. If, for example, you have a portfolio of residential real estate mortgages, your base case will look like this:

Portfolio of residential real estate mortgages - cash flowsClick to enlarge

Source - Kamakura Corporation

We can then run a stochastic process that provides you with many scenarios, including the impact on this portfolio due to changes in market conditions, macro factors and counterparty creditworthiness, thus:

Stochastic Analysis - RREsClick to enlarge

Source - Kamakura Corporation

We can then specify a confidence interval to see how the cash flows change based on all the risk factors that impact them, and we can arrive at an "at risk" number that is very similar to an earnings or value at risk number. The holding period is replaced by the modeling horizon, and for each period you will have a specific base case, a scenario that provides the cash flows for the given confidence interval, and therefore the difference between the two which is nothing but the liquidity at risk, and this can then be cumulated to find out the risk for any specific combined period.

Liquidity at RiskClick to enlarge

Source: Kamakura Corporation

Structural Benefits of this methodology

This approach gives you a fundamental insight into the way cash flows change with changing risk factors, and provides you with the following advantages:

  • Instead of a ratio, the user can review the actual cash flows
  • It provides an insight into the various scenarios and how these impact liquidity cash flows
  • It is more structured, and takes into consideration customer behavior patterns
  • Cash flows include the impact of prepayments and early withdrawals
  • Just as we have illustrated above for a single asset class, the entire balance sheet can be analyzed, thus:

Balance Sheet Liquidity at RiskClick to enlarge

Source: Kamakura Corporation

  • This approach provides a good alternative to the standard gap analysis that is traditionally undertaken to understand cash flows, and these change on the basis of scenarios. For instance, a 99% confidence interval as opposed to the 95% that we have assumed above, will deliver the following results:

Alternative Confidence Interval - Liquidity at RiskClick to enlarge

Source: Kamakura Corporation

  • It takes into account the organization's risk appetite and risk tolerance
  • This approach can easily be linked to the amount of capital that is used by each asset class, and can form a simple basis for capital allocation
  • It models liquidity for what it is, which is a second order risk, by correctly identifying the key risks associated with each asset class
  • It manages liquidity through the management of other risks, and correctly identifies the relationships between risk categories
  • The approach is rooted in a well-accepted approach already popularized through value at risk


Approaches to model liquidity have been varied, some successful and others not so successful. An approach that seeks to understand liquidity and its changes on account of changes to primary risk drivers is an approach that will always provide you with more credible results than when you seek to model liquidity on a standalone basis. The approach also gives you clear decision parameters as limits can be set for each point in the modeling horizon based on the 'best effort' liquidity estimates that you provide, and finally, this approach seeks to integrate with the value at risk already in use within most organizations.

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.