Brief Introduction

Setting the Scene

Where does liquidity risk enter the banking activities? Are some banks incurring more liquidity risk than other banks? Is liquidity risk and isolated risk? Can it be traded?

What is a bank anyhow?

A commercial bank is an institution that engages in two activities on each side of the balance sheet: deposit-taking and lending.

Taking deposits means issuing claims, which are risk-less and demandable. Lending is a more complex process, because it requires the bank to acquire information about borrowers and estimating the credit worthiness based on this information.

The fact that banks engage in both types of activities is at first sight not clear. It can be shown that loans can be provided by a non-bank, which is funded by short term debt and not demand deposits. On the other hand an intermediary can take demand deposits to invest in mutual funds which contain liquidity securities only (government bonds, high grade commercial paper). So there is no need to be active in the lending business if one were to take deposits. Needless to say that these two options represent interesting business models that cater to a niche market.

In fact there are financial intermediaries, which don't have demand deposits.

Synergies for Deposit Takers and Lenders

Nevertheless the reason why banks are taking deposits and giving out loans is very simple. There are synergies created by doing both. Every bank has to have some sort of an "overhead", which consists of cash and securities as a buffer to bridge the imperfections on the capital markets, i.e. to absorb liquidity shocks. This is necessary for taking deposits, because there might be a sudden need from many depositors.

So there will be time periods where withdrawals from deposits won't require any back up by the liquidity buffer and as a matter of fact this rarely happens. As a result the "idle" buffer can be used for other activities like giving out loans. The arguments is true for the reverse situations. Hence there is a synergy created for the bank by doing both.

Can we quantify the synergy? Lets look at a simple example: We mentioned that giving out loans could also be accommodated by a finance company that relies on issuing short term debt or alternatively long term bonds, which have to be serviced at market rates / coupon.

The bank on the other hand is not forced to pay market rates to the depositors. In fact all banks pay considerably less than market rates, as you might have found out with your own bank account.

This spread can be quite substantial and as a result if both companies compete for the same client base a bank has arguably an advantage as the refinancing is cheaper for them. This is by the way one of the reasons why banking giants like JPMorgan or Citibank were able to muscle their way into investment banking. Investment banking especially for corporates always includes lending and even more important: committed lending.

Being able to exploit this cost advantage can therefore create tremendous revenue opportunities and consequently huge profits.

Liquidity Risk is Inherent to Banks

However, this synergy works very well only as long as the correlation between withdrawals from deposits and demand for new loans (especially committed loans) are not highly correlated. This will eat into the liquidity buffer very quickly and might bankrupt the bank. In the simple agent based model we have created a similar situation.

This correlation is also called fragility by some researchers. They point out that these two business activities are incompatible, because frictions are pre-programmed. They also point out that this fragility allows banks to provide credit and liquidity to the market. Or more precise: Financial fragility is a desired feature of banks!

We already see that this goes towards the direction of risk and risk management. If banks are fragile then there will be the chance also known as risk that the synergy breaks down.

If the correlation between withdrawals and loan commitments is positive then the bank will face a liquidity crisis and is hence subject to liquidity risk. If we further argue that this correlation is reflected by a timing issue then we will arrive at a handy definition for liquidity risk:

"Liquidity risk is the risk that a bank is not able to honour its commitments when they fall due. "

Consequently keep your liquidity in order and make sure that all expected cash flows are well timed and no harm will follow.

Liquidity Risk is a Consequential Risk

So can we regard liquidity risk like any other type of risk we know? Is it similar to credit risk, market risk, operational risk or any other risk you come up with? Can it be analysed and isolated like for example credit risk?

The answer unfortunately is NO. If we follow through our model from above we can see the following:

The withdrawals by the depositors are stochastic. Other factors can include the credit rating of the institution and its reputation.

Loans and especially committed loans / committed credit lines are a very complex financial product. If and how much is drawn depends very much on the credit rating of the client. They can be replicated with complex derivative transactions, which would be notoriously difficult to hedge.

We see that liquidity risk is a consequential risk. It can essentially be the unpleasant add on of any other type of risk event. It is highly asymmetric and one could argue that due to the non-linear behaviour of other risks, liquidity risk shows even more leverage, because it crystallises in those moments when a bank has already taken a hit.

Liquidity Risk is Inherent to Banking