Liquidity Risk Explained

Liquidity Risk

Summary

If you obtain a contractual cash flow report for a bank you will see that the liabilities have shorter maturities than the assets. There is an assumption that short term borrowing can be rolled over without difficulty. But the 2007 crisis and aftermath proved this wrong. Banks would have been better able to weather the storm had they held more high quality liquid assets. At the heart of the crisis was a lack of confidence. It is why stress testing liquidity is in vogue. Stress testing isn’t simple and it can’t predict the future but it can give you an estimate for your liquidity buffer. It’s likely to be a lot bigger than previously and it’s going to cost your firm more, that’s unless you can pass it on. 

Why borrow short and lend long? 

Banks have traditionally engaged in short term borrowing and long term lending. That’s because it makes money. It works as long as depositors have confidence and they don’t all come knocking on the door at once asking for repayment. If this happens banks simply don’t have the money to hand and that’s liquidity risk. It can lead to failure. A risk both firms and regulators became complacent about.

Before the crisis banks could receive Libor plus 25 basis points by investing for five years. Funded for three months at Libor the resulting profit of 25 basis points appeared to many as an arbitrage. The limits got larger and some firms took things further. They set up asset backed commercial paper programmes and structured investment vehicles to exploit the perceived liquidity arbitrage. Coincidentally firms reduced their holdings of liquid assets and widened the definition to include illiquid securities in an effort to squeeze more out of their money market books.

Why did this happen?

It happened because liquidity risk was perceived as minimal. That is because liquidity risk is all about trust and behaviour. If the banking system is trusted, short term depositors whether they are wholesale or retail will continue to roll over their deposits with few questions asked. Confidence must be badly damaged before investor behaviour changes. But confidence is fickle and if you think there is a chance your three month deposit is in danger of becoming a three year investment or worse, you won’t renew. This occurs when banks are in danger of losing more money than their capital will support. It happens more frequently than you expect. There have been banking crisis throughout history. Latin America in the 1970s, US Saving and loans and Scandinavia in the 1980s and Japan in the 1990s.

In many cases liquidity risk is a symptom and not the cause of failure. The cause of failure is normally a balance sheet weakened by bad lending, insufficient capital, bad management, fraud or a combination of these factors. If a bank can’t absorb its losses it will eventually run out of money and suffer a liquidity crisis.

This is exactly what happened in the 2007 crisis.  But this time the crisis emerged very quickly and banks had insufficient liquidity to buy time in order to repair their balance sheets. If you were able to point at one factor that triggered events it was probably the realization by investors of how individual banks and the system were leveraged. It came as a big shock to most. When they found out they reacted rationally. They weren’t getting paid to take this risk so they withdrew their funds. It left governments with the stark choice. Either provide emergency finance to cover the outflows and more capital to cushion the credit losses or let the banks fail. Lehman Brothers demonstrated that the latter strategy was too risky so the bail outs began in earnest. 

It is that forced support that has made regulators focus on the amount of liquidity held by banks and the system as a whole. It’s important for the following reasons.

1. A bank that fails to meet its contractual payments endangers itself and the system.

   
   

2. Improved liquidity management reduces the need for the central bank to act as lender of last resort and this reduces the risks and cost to the taxpayer as private markets become the lender in the first instance.

   
   

3. Increasing liquidity requirements makes leverage more expensive for banks and scales back one of the prime causes of their failure.

    

In the new financial world running liquidity light is not an option. Not only will the market discipline you but regulators can and will change management when liquidity risk is inadequately handled.  So just how do you measure and manage it?

Measuring liquidity risk

Liquidity risk can only be fully hedged if the maturity of your assets and liabilities is identical and depositors cannot ask for early repayment. But that doesn’t happen. When you lend long and borrow short you assume you can roll over the borrowing. But in reality the depositor has that option not you. In this sense when you run liquidity risk you have sold an option, you are short the refinancing risk and the depositor is long. Unfortunately standard option models do not fit this risk very well. That’s because you are looking at behaviour and it doesn’t fit the standard distribution functions used. We need to adopt a different approach, an approach that makes assumptions about behaviour.

If these assumptions prove to be inaccurate and a firm’s liquidity comes to be tested it is likely that the firm will suffer failure.

This means that whilst the probability of failure may be low the cost of failure will be disproportionately high unless you can rely on the central bank as lender of last resort.

Whilst this happened in 2007 it is not clear whether similar reliance is justified in the future. Indeed boards that make that assumption are indeed short of on option themselves – a regulatory put option.

That’s why in future the final measures of liquidity risk can be expected to err on the side of caution and it is most likely that firms in the long run will carry more liquidity than is strictly necessary.

Contractual cash flows

The assets and liabilities on your balance sheet will have contractual repayment dates. These are the dates on which the respective amounts are due to be paid and received. Using this information you can build a cumulative cash flow report or cash “ladder”.

If you had a simple balance sheet where you borrowed $100m for three months and lent $100 for one year the ladder would show a net cash outflow of $100m in three months time: 

That’s useful information. In this example you have three months grace before your need to “roll-over” your borrowing. Your liquidity risk would crystallise if the funding is not renewed.

The earlier you run into a cumulative net outflow the earlier your potential liquidity risk arises

How could you avoid this problem? One solution is match funding, where the assets and liabilities have the same maturities:

This is a simple solution but a costly one (because term money is more expensive to borrow). Can you do anything else? Yes, you could spread the funding out over several time periods, preferably from different sources and hold some short term assets as well. Your report could now look like this: 

Now your liquidity risk has been pushed further out in time. The short term asset can also be used to meet the three month liability should it be repaid without replacement. The refinancing risk is now in month six and it is for $25m not $100m (as in our first example).

This may look simple, it is and it’s what banks do. The assets are generally longer in maturity. The funding gets spread out and diversified. A proportion of short term assets are held in case of need.

In this contractual model as long as you hold positive short term cash balances and you are not overextending your refinancing risk all appears to be well. But it isn’t. This contractual model is only half the picture. It fails to consider the behaviour of the parties that provide you with cash and that is a serious weakness.

Where liquidity risk bites

Taking our example a step further will highlight the problem. The bank makes a $75m one year loan and parks the $25m balance in short term assets. But this time it is financed as follows. $20m retail funding with immediate access for the customer and $80m wholesale funding spread out over the remaining time periods. Our cash flow report looks like this:

It appears as if the liquidity risk is well managed but that could be far from the truth. First it’s clear that with an 80/20 split this bank is excessively reliant on short term wholesale funding. Second have the dealers bought an illiquid three month asset?  Third how mobile are the retail savers? These are difficult questions to answer.

Managing  the risk using daily cash flows will help. You can then see exactly when you have a liquidity shortfall and you can fill the gap with a suitable liquid asset. Many dealers will be familiar with this approach it’s a classic way of managing short term liquidity. But you are still facing uncertainty.

Deposits over and above any government protection scheme, internet accounts and money brought in by offering high rates are all considered to be of higher mobility.  Short term wholesale money may not be refinanced. How big does the liquidity buffer need to be to cover these possibilities? That’s something your cash ladder will never tell you and it’s where stress testing comes in. 

Stress testing

Stress testing goes beyond contractual reporting. It’s a way of assessing behaviour in extreme conditions. It should tell you what may happen in times of extreme disquiet. That’s how much retail and wholesale funding that could be withdrawn over a given period of time and how much liquidity you would need to cover it. It’s an estimate of the liquidity buffer you need. The purpose? To buy time. It allows you to take action to resolve the outflow, it allows the market to normalise and it allows regulators and central banks to consider the options available to them. 

In the UK (and other jurisdictions) liquidity stress testing is now mandatory for financial institutions.

The UK regulator, the FSA, insists that banks undertake regular and robust stress testing of their liquidity. The stress tests that must be made are explained in the FSA’s Consultation Paper 08/22 - Strengthening Liquidity Standards. Firms must undertake three stress tests, they are:

1. A firm related stress test

2. A market related stress test

3. A combination of the firm and the market stress tests

   
   

The firm is obliged to assume the short term firm specific stress lasts two weeks and the market wide stress for three months. The FSA expects firms to identify and report on the sources of liquidity risk that would crystallise under the scenarios. In particular the FSA highlights the “drivers” of liquidity risk to be as follows:

• Wholesale secured and unsecured funding risk

• Retail funding risk

• Intra-day liquidity risk

• Intra-group liquidity stress

• Cross-currency liquidity risk

• Off –balance sheet liquidity risk

• Franchise viability risk

• Marketable assets risk

• Non marketable assets risk

• Funding concentration risk

  
  

Firms are required to make certain assumptions about the stress in order to calculate the net outflows before management action for each risk driver. The size of each outflow will then determine the liquidity buffer needed. This is the level of liquid assets that must be held. It is assumed that in the first two weeks of the stress period the firm will have to meet all its liquidity needs entirely from this liquid assets buffer. After this period mitigating action by management can be brought to bear on the situation.

What are the merits of this approach? 

The firm is provided with the stress scenarios and liquidity risk factors that it must consider. But it has a choice about how it undertakes the analysis and how it determines the relative importance of the factors that create liquidity risk. This allows each firm to assess the risks in light of its business circumstances. Therefore there is a common industry benchmark for stress but individual firms have the flexibility of self assessment in so much as the stresses affect their business.

Instrumental in obtaining an accurate assessment of the liquidity buffer are the assumptions made by management in the stress testing and the level of liquidity risk that the board considers to be acceptable.

An incentive 

Whilst firms may be tempted to manufacture an artificially low liquidity buffer there is a big incentive not to do so. Firms must report what they consider to be an appropriate buffer to the regulator. This must be fully supported with the analysis undertaken. In the event the regulator disagrees it can set the liquidity buffer higher.

Should the regulator need to resort to this, the quality of the firm’s management is put into doubt. In other words firms must align their liquidity risk appetite with that of the regulator.

The liquidity buffer 

Having determined the size of the liquidity buffer what can it contain? Pre-2007 firms often maintained liquidity in a number of instruments. These included interbank deposits, certificates of deposit, commercial paper, treasury bills, government bonds, repo and in some cases high quality structured assets. This is no longer considered acceptable.

A narrow definition of “liquid asset” is now being applied. In the UK the regulator has taken the view that only the highest quality bonds will qualify. This means for sterling based banks with sterling liabilities, UK government bonds (Gilts), UK treasury bills and deposits with the central bank. Why is this?

The regulator argues with some justification that in times of stress should a bank need to obtain short term liquidity it has two choices. It can either sell or repo assets. And only the highest quality assets can guarantee a sale without loss of value or be used as collateral in the repo market.

This will have far reaching consequences. The factors that cause liquidity stress determine the liquidity buffer. The more liquidity risk a firm has the larger that buffer needs to be. Weak management can also increase the buffer required.

Low yielding assets in the buffer make it expensive to maintain. This will encourage firms to factor the cost of liquidity into the business they conduct. For example, short term wholesale and retail funding will become less attractive and “off balance sheet” liquidity guarantees will become prohibitively expensive.

What else can be used to reduce liquidity risk? 

A further mitigant of liquidity risk is a firm’s Contingency Funding Plan (CFP). The CFP details the course of action that will be taken by the firm should a liquidity stress arise.  A well prepared CFP should mean that a firm can respond to a liquidity crisis faster and with greater certainty. Here are a few things you would expect to see in a CFP:

• It should be properly documented. A CFP that is not fully explained is probably one that has not been adequately considered. 

• The documentation should explain what triggers the plan. The triggers should be clear and quantifiable. For example if the liquidity buffer falls below a given percentage the plan is invoked. 

• The management roles, actions and responsibilities should be clear. 

• The sources of funding that are available and the time frame involved in activating these sources should be clear. 

• There should be signs that the CFP has been designed as a result of considering the stress tests that have been carried out. For example if the stress testing shows that wholesale funding is unavailable then the CFP should not assume the wholesale market as a source of funds. 

• It should be tested to ensure it works effectively.

  
  

The effect of the CFP on the liquidity buffer 

In times of stress a firm may need to rely on the CFP so it is a document that regulators take seriously. The validity of the CFP will have important implications on the liquidity buffer. A badly drafted CFP with unrealistic assumptions will mean that it cannot be relied on. From a regulatory perspective this will translate into a higher liquidity buffer and consequential cost for the firm.

Liquidity costs

The quantity of liquidity a firm must hold will depend on a number of factors, they include the following: 

• The degree of maturity transformation in the firm’s balance sheet

• The source and stability of funds that the firm can access

• The severity of the stress testing that is applied

• The quality of the CFP and the firm’s management 

  
  

High quality liquid assets held in the buffer will provide an income that is lower than the firm’s funding cost. How great is this difference? That’s the million dollar question. A “back of an envelope” calculation shows you it’s not cheap.

Here are the assumptions. A firm funds $1bn of assets, 40% through wholesale funding and 60% through retail. The difference in yield between government bonds and the firm’s cost of funding is 2.00%.

Depending on the size of the liquidity buffer that must be held the cost will be: 

The Net Interest Margin (NIM) shows the reduction in income that the firm will incur as a result of the liquidity buffer requirement.

For a firm that is 60% retail and 40% wholesale funded it is easy to see that a severe stress test could push the liquidity buffer required by the regulator towards 30%. This figure is of course dependent on a long list of factors. But if you consider that before the crisis some firms were running high quality liquidity ratios of around 5% this is a significant change. No wonder that the implementation is planned to be phased in over a number of years. The implications are clear.

The industry will start to pass on the liquidity cost to those who create it. Borrowers will face more expensive loans. Liquidity commitments on a firm’s balance sheet will also reflect the liquidity cost they incur.

There is also a general assumption that short term wholesale financing is risky. This should mean that firms seek to issue longer dated bonds and retail savers who are prepared to lock up their funds for a significant length of time will benefit from higher returns.

First Published by Barbican Consulting Limited 2009