Stress tests – the benefits
In response to the financial crisis, the Basel Committee has drawn up new standards to mitigate the liquidity risks that banks face. The first of these standards, the Liquidity Coverage Ratio (LCR), should help banks hold sufficient liquid assets to weather a severe, but relatively brief period of stress. The Net Stable Funding Ratio (NSFR) limits structural liquidity risk by promoting the funding of long-term exposures by means of stable liabilities on a bank's balance sheet. Both standards are based on fixed principles for measuring liquidity risk.
Stress tests are a useful addition, as they offer flexibility in terms of time horizon and scenarios, and allow for simulation of economic or financial shocks. They thus help to identify financial system vulnerabilities before they actually emerge, and can be used to measure how long a bank will be able to continue satisfying regulatory minimum liquidity requirements under stressful conditions. DNB has long been using its own in-house stress test framework as part of its regular supervision duties.
Liquidity stress tests in the financial sector are not new, but there is no uniform approach to, for example, determining the scenario assumptions that institutions use. As a result, approaches diverge widely between countries and institutions, and there are blind spots to methodologies used for liquidity stress testing – shortcomings which this study by the Basel Committee aims to address. For one thing, stress scenarios should factor in second-round effects, which might emerge when markets become unstable as affected parties respond or if their responses trigger reputation damage. Institutions have been known to overlook the negative effects of their own behaviour on the markets in which they operate, and to – erroneously – assume they will be able to sell off assets quickly and without incurring too great a loss.
The study also warns that banks' stress tests should not unquestioningly assume that central banks will always be at the ready with loans to help them out. Today's situation, in which many euro area banks rely on central bank funding, should not be taken as a mere given, and conditions of stress might in fact restrict access to central bank funding due to a shortage of eligible collateral.
Solvency and liquidity
It should also be noted that although liquidity and solvency risk are closely related in practice, most stress tests approach these risks separately. An example of this is a stress test of the capital position of banks that only takes into account credit losses on bonds. The interaction with liquidity risk should be considered here. A bank's capital position may for instance come under pressure due to losses on liquid assets if they must be sold at fire sale prices in order to safeguard the bank's liquidity position. Stress also influences a bank's funding costs depending on its liquidity and funding position. High funding costs in turn have a negative effect on the bank's future capital position.
The study also offers several practical suggestions. Institutions would be well advised to test their liquidity positions for each individual currency in which they do business. They are recommended to perform stress tests both at group and entity level as it cannot be taken for granted that under stressful circumstances cash will at all times continue to flow between the different units of a bank. Business with shadow banks, such as money market funds or special purpose vehicles, should also be included in the stress scenarios. And last but least, liquidity stress tests must be incorporated in institutions' risk management if they are to have genuine added value.
DNB's liquidity stress tests in some ways already apply the suggestions made by the Basel Committee. In our stress tests, banks experience higher liquidity outflows if they have lower ratings than their peers and if they are dependent on non-euro financing. In this way, capital and liquidity are linked up with the currency mismatch risk. Second-round effects play a part, as the fact that banks will be confronted with higher losses on assets if they have not succeeded in raising finance after one month is factored in. Solvency risk is also important as it is assumed that banks with more capital and fewer assets used as collateral (asset encumbrance) will be able to raise finance more quickly than other banks. Institutions with fewer capital and more encumbered assets are dependent on their liquidity buffer, making them vulnerable to liquidity stress and increasing losses upon the sale of liquid assets.
DNB intends to use the results of the Basel study to further refine its stress tests and also expects the banks to take the Basel recommendations on board. Hence, the study may serve to promote liquidity stress tests being performed in a more comparable way.