Page 1 of 17
Archives of Business Research – Vol. 10, No. 6
Publication Date: June 25, 2022
DOI:10.14738/abr.106.12513. Hassanein, M., Bouaddi, M., & Aziz, H. M. A. (2022). Systematic Market Risk Adjusted for Liquidity Cost in the Banking Sector.
Archives of Business Research, 10(6). 62-78.
Services for Science and Education – United Kingdom
Systematic Market Risk Adjusted for Liquidity Cost in the
Banking Sector
Medhat Hassanein
Distinguished University Professor
Professor of Finance, Management Department
American University in Cairo
Mohammed Bouaddi
Associate Professor
Department of Economics, American University in Cairo
Heba Mohamed Abdel Aziz
Research associate, American University in Cairo
ABSTRACT
The emergence of one or more risks in the financial markets during a specific period
causes a financial crisis. Financial crises impact financial stability, which is a key
concern for all financial authorities, including central banks. One way to mitigate
the risks any economy faces is to understand the origin of risk and how it spreads
through the financial system. Liquidity risk goes hand-in-hand with market risk, as
they affect each other during a crisis. After the 2007-2008 global financial crisis,
banks showed that they need monitoring and efficient liquidity management during
both stress and normal conditions; that is, they require better integration of bank
liquidity and market risk management. In this study, we present a new
methodology to forecast the systematic market risk adjusted for liquidity cost
based on the Conditional Value at Risk (CoVaR) risk measure and asymmetric
conditional copulas. We analyze a sample of international banks based on asset size
in the US, EU, and Asia. Our hypothesis confirms that liquidity risk goes hand-in- hand with market risk, as they affect each other during a crisis. The results show
dependence in the tails of the banks’ returns and the market returns. These are very
high generally, and even higher for the Covid-19 period than for other periods. The
effect of both market risk and liquidity risk on banks during a crisis period is less
than in non-crisis periods because banks are well informed institutions and can
anticipate a financial crisis and mitigate risk, which explains our results.
ACKNOWLEDGMENT
The authors of this article acknowledge the financial grant of the American University in Cairo
to assist the authors to devote part of their own time to finalize this article.
INTRODUCTION
The increased complexity of the financial market makes the measurement of market and
liquidity risks a pressing topic. Regulators increased the level of control over banks to ensure
that sufficient capital to hedge against market and liquidity risks. Although it is important to
have a single risk measure to capture the risk a bank may face, it is also very important to have
Page 2 of 17
63
Hassanein, M., Bouaddi, M., & Aziz, H. M. A. (2022). Systematic Market Risk Adjusted for Liquidity Cost in the Banking Sector. Archives of Business
Research, 10(6). 62-78.
URL: http://dx.doi.org/10.14738/abr.106.12513
a more granular level of detail that reflects performance and how to avoid risk. Risk measures
and controls are used concurrently for internal control and by regulators to monitor banks’
performance.
Market and liquidity risk management is one of the most challenging tasks in quantitative
finance. The Basel Committee on Banking Supervision (BCBS) (2019) defines market risk as
“the losses in on- and off-balance sheet risk positions arising from movements in market prices,
the risks subject to market risk capital requirements include but are not limited to, default risk,
interest rate risk, credit spread risk, equity risk, foreign exchange (FX) risk, and commodities
risk for both trading and banking book instruments”. While the BCBS (2008) defines liquidity
risk as “the ability of a bank to fund increases in assets and meet obligations as they come due,
without incurring unacceptable losses. The fundamental role of banks in the maturity
transformation of short-term deposits into long-term loans makes banks inherently vulnerable
to liquidity risk, both of an institution-specific nature and that which affects markets as a
whole.”
Banks were free to use their own market risk models to estimate the regulatory capital for their
trading book positions. Such internal models were designed to calculate the Value at Risk (VaR)
using a confidence level of 99%. According to Wimmerstedt (2015), VaR measures the
“threshold loss over a time period that will not be exceeded with a given level of confidence.”
The VaR model is very well known because it is an easy concept to understand and apply.
Moreover, VaR is very easy to back test because it relies on the historical number of losses,
which can then be compared with the actual versus forecasted losses. Tavana et al.(2018) state
that liquidity risk poses a devastating financial threat to banks if not well managed and can lead
to irreversible consequences, such as the 2007-2008 Global Financial Crisis (GFC). Therefore,
an accurate measurement to liquidity risk is needed. However, liquidity risk measurement is a
complicated process, and must be well defined in order to provide a precise measurement. One
of the problems of defining liquidity risk is to determine the factors that create an appropriate
function that can predict and approximate its true value. Another challenge is that the definition
of liquidity risk is vague and uncertain. It is vague because the term liquidity can refer to many
dimensions simultaneously, especially when used with market risk or systemic liquidity risk. It
is uncertain because it may have different meanings in different contexts.
Our study focuses on the integration of market risk and liquidity risk. We also measure liquidity
risk by identifying the factors that capture liquidity risk accurately and determine the impact
of the overall integration of market and liquidity risks.
The rest of this paper is organized as follows. Section 2 provides a literature review. Section 3
discusses the systematic liquidity risk measurement, which depends on the state of the market.
Section 4 provides the model specification, in which we specify the marginal distributions and
the Copula of returns. In Section 5, we apply our methodology to quantify the systematic
liquidity risk conditional on the state of the market and report the results in Section 6. Finally,
we offer our conclusions in Section 7.
LITERATURE REVIEW
The literature on market risk measurement and estimation is vast. The most common model
for quantifying market risk is VaR, proposed by Jorion (1997) and Dowd (1998). Brien and
Page 3 of 17
64
Archives of Business Research (ABR) Vol. 10, Issue 6, June-2022
Services for Science and Education – United Kingdom
Szerszeń (2017) introduced the more accurate generalized autoregressive conditional
heteroskedasticity (GARCH) based VaR. Tripe and Malone (2017) measure systemic risk in
banks using the z-score, building on prior works by Roy (1952); Boyd and Graham (1986);
Hannan and Hanweck (1988); and Boyd, Graham, and Hewitt (1993). Mainik and Schaanning
(2014) use the Conditional VaR (CoVaR) to capture market risk after Tobias and Brunnermeier
(2016) introduced CoVaR to capture risk spillovers across financial institutions. However,
Löffler and Raupach (2016) find that given the large weight of banks in a market, CoVaR can
imply lower systemic risk in the market if banks increase their idiosyncratic risk. Teply and
Kvapilikova (2017) derive a robust market-based measure of systemic risk during different
financial cycles, called the wavelet CoVaR (WCoVaR). Karimalis and Nomikos (2018) propose a
new methodology to estimate the CoVaR based on copulas and VaR under financial stress
conditions.
Measuring liquidity risk has been an area of focus since the 2007-2008 GFC. Many academics
showed interest in modeling how to capture liquidity risk accurately. According to Tavana et
al. (2018), the standard way to measure liquidity risk is by comparing the available funding
sources against the expected cumulative cash shortfalls over a specific period. They conclude
that artificial neural networks (ANNs) and Bayesian networks (BNs) can find the relevant
critical risk factors that can measure liquidity risk accurately through functional and
distributional estimations. DeYoung, Distinguin, and Tarazi (2018) study liquidity risk through
the interrelationship between bank capital in the US and their liquidity prior to the Basel III
regulations, which restricted liquidity positions. Bai, Krishnamurthy, and Weymuller (2018)
develop a liquidity measurement system through the liquidity mismatch index (LMI). The LMI
measures the mismatch between the funding liquidity of liabilities and the market liquidity of
assets. Khan, Scheule, and Wu (2017) examine the effect of changing funding liquidity risk on
bank risk taking capacity and the impact of bank size and capital buffers on bank risk capacity
and funding liquidity risk through a panel regression. Their results support the idea that banks
should move away from the short-term funding to reduce their riskiness and improve their
assets quality. The findings of their study show that the bank size and capital buffer help curb
the bank’s risk-taking behavior in response to decrease the funding liquidity risk. This shows
the importance of studying the interrelation between the market and liquidity risks. If banks
size and capital buffer affect their liquidity risk profile it will impact their market risk as well.
Aniunas et al. (2017) created a model of liquidity risk management based on VaR and the best
practices convenient for Lithuanian local commercial banks. Dionne, Pacurar, and Zhou (2015)
propose the liquidity adjusted intraday VaR (LIVaR) as a measure of liquidity risk for high
frequency data. It accounts for both the ex-ante liquidity risk of liquidating a position and
market risk. Jobst (2014) developed the systemic risk-adjusted liquidity (SRL) model to
measure systemic liquidity risk. Ruozi and Ferrari (2013) examine three well-known
approaches to measuring liquidity risk: the cash flow matching approach, stock-based
approach, and hybrid approach.
Many research efforts aimed to incorporate market liquidity risk into VaR models. Jarrow
(1997) introduces a liquidity adjusted VaR that includes the volatility of the liquidity discount
and the volatility of the liquidation time horizon considering both the execution lag and effect
of trade size on the portfolio liquidation value. Almgren and Chriss (2001) estimate a simple
linear cost model in which they construct an efficient frontier for a period that depends on the