There has been a lot of focus on non-financial risks like cyber and conduct over recent years, but that doesn't mean that traditional risks like credit and market liquidity should not be on a risk manager's radar. Here, Sorina Zahan, CIO of Core Capital Management, discusses the problem with misunderstanding liquidity risk.
Today, market liquidity is one of the most acknowledged and feared risk factors. Paradoxically, it is also the most poorly understood and measured, not only by market participants, but also by regulators. The most basic and widely used market liquidity measures are based on transaction volumes. Other liquidity measures have been proposed; these may incorporate elements like funding liquidity, margin, daily returns, etc. However, measuring liquidity is only the starting point; it is by no means the solution to the problem. What is often forgotten is that liquidity measures are only indicators of the level of market liquidity, not of the liquidity risk.
Liquidity level is not liquidity risk
A costly confusion among market participants is that liquidity level is the same as liquidity risk. As such, a portfolio of liquid instruments is thought to carry little liquidity risk. This is not only false, but also unfortunate, as for each crisis since the 1990s, research shows that liquid stocks and bonds do not necessarily offer a good hedge against liquidity crises.
What is often forgotten is that liquidity measures are only indicators of the level of market liquidity, not of the liquidity risk.
Liquidity risk, like all risks, involves the presence of unexpected changes. For each security, investment or portfolio, liquidity risk manifests itself through the covariance of returns with unexpected changes in aggregate liquidity. Liquidity risk can be measured by the sensitivity of returns to changes in aggregate liquidity, or liquidity beta.
Liquid securities do not carry less liquidity risk
Various research papers published in the last 20 years, both in academia and by the Federal Reserve, show that irrespective of their historical liquidity level, securities with low liquidity beta outperform in crises. Stronger findings, referring to stocks, show that in crises liquid stocks actually underperformed less liquid stocks from the same category (e.g. large caps). Essentially, all research - be it in the weak or strong form, reveals two findings:
- Equally liquid securities do not carry equal liquidity risk;
- Liquidity risk, i.e. sensitivity of an investment to changes in aggregate liquidity, is a better predictor of performance during crisis than liquidity level.
As such, the ubiquitous confusion between the liquidity level and liquidity risk can prove costly for all. Areas of concern are not limited to systemically important institutions, pension funds or insurance companies – all of them being forced to tilt their holding towards the same, highly liquid securities, which in effect may carry more liquidity risk, but also include less obvious places like ETFs. We do not refer here to the asset-liability mismatch, or the “flow risk” of ETFs. This has become more acknowledged in the market place and certainly by regulators. We refer to pure market liquidity risk.
The ubiquitous confusion between the liquidity level and liquidity risk can prove costly for all.
The growth of ETFs, now heavily present in both retail and institutional portfolios, has steadily forced them to buy the most tradeable stocks or bonds in their particular investment space. Paradoxically, the more an ETF grows, the more its investable universe shrinks, increasingly including from the original universe only those securities which have the highest liquidity level. This fact creates a liquidity risk feedback loop, as it increases the liquidity beta of both the ETFs and that of the underlying securities. In our view, cross-sectionally, securities with high liquidity level carry now even more liquidity risk than they have historically.
Analyzing and measuring the liquidity risk
Since the Asian crisis, practitioners, regulators and academics have paid more attention to liquidity. However, taking the leap from measuring liquidity level to developing a practical framework for the analysis of liquidity risk has proved to be non-trivial. Our research bridges this gap. Its main objective was to build a framework which can be used to measure and manage liquidity risk for both transparent and opaque portfolios. Ultimately, the goal of any such framework is dual:
- Optimizing liquidity risk
- Hedging liquidity risk
In the CoreLabs framework we construct a model of a given portfolio or security in the liquidity space. The liquidity space can be defined by any aggregate liquidity measure. The model is called Liquidity Beta Profile®. It is this profile that investors should attempt to optimise, or, at minimum, be aware of.
A major conclusion of our research is that, for practical objectives like optimization and hedging, it is best to analyze the liquidity risk in a regime dependent framework. Dozens of analyses performed on complex portfolios spanning diverse asset classes - from liquid public equities and fixed income to private equity, credit and derivatives portfolios - show that the Liquidity Beta Profile®, despite being built without visibility on the underlying individual securities, strongly discriminates cross-sectionally. Its accuracy and discrimination power increases with the time granularity of the underlying data, but even for highly infrequent data, it remains a powerful basis for optimization.