What do the value-at-risk measure and the respective legislative framework really offer to financial stability? Critical views and pro-cyclicality

Abstract

In this paper, we examine how value at risk (VaR) contributes to the financial market's stability. We apply the Guidelines on Risk Measurement and the Calculation of Global Exposure and Counterparty Risk for UCITS of the Committee of European Securities Regulators (CESR 2010) to the main indices of the 12 stock markets of the countries that have used the euro as their official currency since its initial circulation. We show that gaps in the legislative framework give incentives to investment funds to adopt conventional models for the VaR estimation in order to avoid the increased costs that the advanced models involve. For this reason, we apply the commonly used historical simulation VaR (HVaR) model, which is: (i) taught at most finance classes; (ii) widely applied in the financial industry; and (iii) accepted by CESR (2010). The empirical evidence shows the HVaR does not really contribute to financial stability, and the legislative framework does not offer the appropriate guidance. The HVaR model is not representative of the real financial risk, and does not give any signal for trends in the near future. The HVaR is absolutely backward-looking and this increases the stock market's overreaction. The fact that the suggested confidence level in CESR (2010) is set at 99 percent leads to hidden pro-cyclicality. Scholars and researchers should focus on issues such as the abovementioned, otherwise the VaR estimations will become, sooner or later, just a formality, and such conventional statistical measures rarely contribute to financial stability.

Inaccurate Value at Risk Estimations: Bad Modeling or Inappropriate Data?

“…Could the CDNVaR model become more accurate and representative if we remove these limitations? Vasileiou and Pantos ( 2020 ) and Vasileiou ( 2019 ) show that the mandatory minimum of 250 data inputs may lead to inaccurate VaR estimations, 11 and following their suggestions we remove the 250-observation threshold from our estimation process.…”

“…A reason could be that, in practice, the most popular models applied in the financial industry are the conventional and simplest VaR models: the Historical, the Variance–covariance (or Delta Normal), and the Monte Carlo Simulation. 2 Conventional models generate low VaR estimations during the days before a crisis and high VaR estimations when a crisis has already emerged (Vasileiou & Pantos, 2020 ; Vasileiou & Samitas, 2020 ). 3 However, the purpose of VaR should be to timely and accurately inform market participants that a stress period is approaching.…”

Do econometric advances lead to more accurate VaR estimations when the legislation requirements are applied? The case of Turkish Lira FX Market.

“…-studies that address the deficiencies of VaR legislation (Vasileiou (2016)) and the procyclicality caused by VaR legislation (Adrian and Shin (2014), Vasileiou and Pantos (2020)), and -studies that try to apply advanced econometric models for more accurate VaR estimations: extreme value theory (GARCH family models (Engle (2004), Assaf (2009), Diamandis et al (2011)), Markov Switching Regime (Billio and Pelizzon (2000)), data filtering models (Vasileiou (2017(Vasileiou ( , 2019, Extreme Learning Machine (Zhang et al (2017)) etc.. The importance of the TRY risk has been documented in several studies: Günay (2017), Yildirim (2015), Gün (2020) etc..…”