Skip to main content
SpringerLink
Log in

Which User-Friendly Model is the Best for BASEL-III? An Emerging Market Study

  • Published:
Computational Economics Aims and scope Submit manuscript

Abstract

This paper explores backtesting Value-at-Risk (VaR) and Expected Shortfall (ES) considering ten standard and extended tests in the context of non-technical individual investors trading equities of twenty selected commercial banks listed at the Dhaka Stock Exchange (DSE) using their daily share prices for 11 years (from 2010 to 2020). Following a significant gap in the literature on investigating the efficacy of user-friendly models in quantifying the market risk of banks in emerging economies, this paper adopted four user-friendly models that are relatively straightforward to understand and interpret and are considered representatives of zero, -one, -two, and -three parametric families of all risk models in the literature specifically for the emerging economy of Bangladesh. The popular RiskMetrics™ risk forecast model of JPMorgan, sweeping the world as the most user-friendly conditional alternative to unconditional Gaussian risk forecasts under the framework of VaR, is found not to be adequate under the framework of ES that was recently recommended by Basel-III. The joint score value-based comparison finds the historical simulation (HS) model as the most appropriate model in Bangladesh when models are assessed under a practical user-friendly implementation design. Under this design the Trust Bank Ltd. (TBL), the bank managed and operated by Bangladesh Military, qualifies as the most investor-friendly bank in terms of causing the least frustration to its equity investors over 2010–2020. Overall, augmenting earlier studies in the literature that are mostly for developed markets and are mostly without any ES back-test, we find that the user-friendly model HS is still successful in quantifying market risk over the globe since its relative usefulness gets well established through recent back-tests of VaR and ES in emerging market too.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Data availability statement

Data collected from DSE and is available on request from the authors.

Notes

  1. “Capital Flows, Financial Crises, and Policies” by Carmen M. Reinhart, et al., NBER Macroeconomics Annual.

  2. one not fuelled by arbitragers phony investment capitals.

  3. Deviating from our user-friendly design.

  4. Analyzed on the basis of backtesting evidence.

  5. Gaussian and student-T location scale models are continuous models and using just one year rolling data in dynamic estimation is likely to catch criticism no matter how practical it may seem from the perspective of individual investors in emerging markets.

  6. However, there is no study available in the literature confirming that based on joint performance of VaR and ES RiskMetricsTM is even an acceptable model; though based on VaR alone its appeal as an acceptable most user-friendly model has been found strong, see [IRFA paper and the references therein].

  7. \(B_{t} \) is a standard Brownian motion and so \(dB_{t} = z\sqrt {dt} ,\) \(z\) being a standard normal random number.

  8. If daily log-returns are used to estimate \(\mu \) and \(\sigma\) and dt = \(\frac{1}{252}\) is used in estimation, instead of dt = 1, then \(\mu \) and \(\sigma \) should be replaced by \(\mu dt\) and \(\sigma \sqrt {dt}\), in the VaR and ES formulas.

  9. Which requires applying Ito’s formula to derive the distribution of log-returns of the asset, see McNeil et al. (2005).

  10. Though we do not use this expression in our user-friendly design.

  11. Again unfortunately we do not use this with our user-friendly design.

  12. Hence this model is more popularly known exponentially weighted moving average (EWMA).

  13. This choice of lambda ensures that return observations older than the previous 100 have no effect in risk forecasts, see Christoffersen (2012).

  14. However we compare our user-friendly design with ES as five VaR average and infinitely many VaR average in loss score computations separately, as in Sect. 3.2. We do not include those results here because except for slight changes in p-values corresponding to ES backtests alternative ES valuations have no effect in loss-score value-based ordering of user-friendly models compare to those obtained with our user-friendly design as presented in next section. This is precisely due to the fact that any extra precision obtained with alternative ES estimation as ‘five VaR average’ and ‘infinitely many VaR average’ applies uniformly to all user-friendly models of our consideration. So that doesn’t affect the preference order we use with our user-friendly design.

  15. One model measures the predictive relationship between Market capitalization and VaR measures and another model corresponds between market capitalization and ES measures.

  16. When we find the data is seriously corrupt, especially at the peak of COVID 19, DSE saw a virtual closure; so those data over that period are virtually non-representative and nonexistent for banks and were discarded from our analysis.

  17. Based, however, on VaR alone.

  18. Chittagong stock exchange.

  19. At least with HS method.

  20. Mean VIF value for all explanatory variables is found 1.32 which is obviously less than 5. Usually VIF value greater than 5 is a sign of causing high collinearity among the variables in the model.

  21. F value of RESET is 2.79 which is statistically insignificant as its corresponding p value is 0.127.

  22. χ2 value is 26.485 and its corresponding p value is 0.017 for Hausman Test.

  23. χ2 value is 1.492 and its corresponding p value is 0.263 for BP/LM Test.

  24. χ2 value for Wald test is 129.459 and its corresponding p value is 0.000 where the Ho is constant error variance across the panels.

  25. F-value of Wooldridge test is 37.394 and its corresponding p value is 0.001 that rejects null hypothesis of no serial autocorrelation in the model.

  26. Mean VIF value for all explanatory variables is found 1.79 which is obviously less than 5. Usually VIF value greater than 5 is a sign of causing high collinearity among the variables in the model.

  27. F value of RESET is 2.571 which is statistically insignificant as its corresponding p value is 0.102.

  28. χ2 value is 32.927 and its corresponding p value is 0.011 for Hausman Test.

  29. χ2 value is 2.015 and its corresponding p value is 0.237 for BP/LM Test.

  30. χ2 value for Wald test is 163.028 and its corresponding p value is 0.000 where the Ho is constant error variance across the panels.

  31. F-value of Wooldridge test is 35.629 and its corresponding p value is 0.001 that rejects null hypothesis of no serial autocorrelation in the model.

References

  • Acerbi, C., & Szekely, B. (2014). Backtesting expected shortfall. Journal of Risk, 27(11), 76–81.

    Google Scholar 

  • Acerbi, C., & Szekely, B. (2017). General properties of backtestable statistics. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2905109

    Article  Google Scholar 

  • Acerbi, C., & Tasche, D. (2002). On the coherence of expected shortfall. Journal of Banking and Finance, 26, 1487–1503. https://doi.org/10.1016/S0378-4266(02)00283-2

    Article  Google Scholar 

  • Agénor, P. (1999). The Asian financial crisis: Causes, contagion and consequences. Pacific Affairs, 74, 580.

    Google Scholar 

  • Akcay, Ü., & Güngen, A. R. (2019). The making of Turkey’s 2018–2019 economic crisis. Institute of International Political Economy Berlin: working paper: 120/2019.

  • Alexander, C. (2021). Asymmetric loss functions and tail risk estimation. Journal of Financial Econometrics, 19(1), 54–83.

    Google Scholar 

  • Allen, D. E., Singh, A., & Powell, R. (2013). Student t copula simulation of extreme equity market shocks: With application to VaR and expected shortfall. International Review of Financial Analysis, 30, 25–34.

    Google Scholar 

  • Artzner, P., Delbaen, F., Eber, J., & Heath, D. (1997). Thinking coherently. Journal of Risk, 10, 68–71.

    Google Scholar 

  • Artzner, P., Delbaen, F., Eber, J., & Heath, D. (1999). Coherent measures of risk. Mathematical Finance, 9, 203–228. https://doi.org/10.1111/1467-9965.00068

    Article  MathSciNet  Google Scholar 

  • Atilgan, Y., Bali, T. G., Ozgur Demirtas, K., & Doruk Gunaydin, A. (2020). Left-tail momentum: Underreaction to bad news, costly arbitrage and equity returns. Journal of Financial Economics, 135, 725–753.

    Article  Google Scholar 

  • Bekiros, S., & Uddin, G. S. (2013). Do High-frequency stock market returns follow a non-student’s t distribution? Evidence from the case of SP500. Physica A: Statistical Mechanics and Its Applications, 392(23), 5925–5935.

    Google Scholar 

  • Bi, J., & Zhu, Y. (2020). Value at risk, cross-sectional returns and the role of investor sentiment. Journal of Empirical Finance, 56, 1–18.

    Article  Google Scholar 

  • Black, F., & Scholes, M. (1973). The pricing of options and corporate liabilities. Journal of Political Economy, 81(3), 637–654. https://doi.org/10.1086/260062

    Article  MathSciNet  Google Scholar 

  • Boudreault, M., & Gauthier, G. (2023). Improving Monte Carlo simulation efficiency with variance reduction techniques. Journal of Computational Finance, 28(3), 78–95.

    Google Scholar 

  • Chai, N., Wu, B., Yang, W. W., & Shi, B. F. (2019). A multicriteria approach for modeling small enterprise credit rating: Evidence from China. Emerging Markets Finance and Trade, 55(11), 2523–2543. https://doi.org/10.1080/1540496X.2019.1577237

    Article  Google Scholar 

  • Chen, S., & Wang, Y. (2020). High-frequency historical simulation for VaR estimation. Journal of Banking & Finance, 40(3), 267–285.

    MathSciNet  Google Scholar 

  • Christoffersen, P. (2012). Elements of financial risk management. Academic Press.

    Google Scholar 

  • Diebold, F. X., Gunther, T. A., & Tay, A. S. (1998). Evaluating density forecasts: With applications to financial risk management. International Economic Review, 39(4), 863–883.

    Article  Google Scholar 

  • Dowd, K. (2005). Measuring market risk. Wiley.

    Book  Google Scholar 

  • Dowd, K., & Blake, D. (2006). After VaR: The theory, estimation, and insurance applications of quantile-based risk measures. Journal of Risk and Insurance, 73(2), 193–229.

    Article  Google Scholar 

  • Emmer, S., Kratz, M., & Tasche, D. (2015). What is the best risk measure in practice? A comparison of standard measures. Journal of Risk, 18(2), 31–60. https://doi.org/10.48550/arXiv.1312.1645

    Article  Google Scholar 

  • Engvall, J. (2016). Backtesting expected shortfall: A quantitative evaluation (Dissertation). Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-198471

  • Escanciano, J. C., & Pei, P. (2012). Pitfalls in backtesting historical simulation models. Journal of Banking and Finance, 36, 2233–2244.

    Article  Google Scholar 

  • Fissler, T., & Ziegel, J. (2016). Higher order elicitability and Osband’s principle. The Annals of Statistics, 44(4), 1680–1707. https://doi.org/10.1214/16-AOS1439

    Article  MathSciNet  Google Scholar 

  • Gencay, R., & Selcuk, F. (2004). Extreme value theory and value-at-risk: Relative performance in emerging markets. International Journal of Forecasting, 20(2), 287–303. https://doi.org/10.1016/j.ijforecast.2003.09.004

    Article  Google Scholar 

  • Gneiting, T. (2011). Making and evaluating point forecasts. Journal of the American Statistical Association, 106, 746–762. https://doi.org/10.1198/jasa.2011.r10138

    Article  MathSciNet  CAS  Google Scholar 

  • Hendricks, D. (1996). Evaluation of value-at-risk models using historical data. Economic Policy Review, 2(1), 39–70. https://doi.org/10.2139/ssrn.1028807

    Article  Google Scholar 

  • Huang, A. Y. (2009). A value-at-risk approach with kernel estimator. Applied Financial Economics, 19(5), 379–395. https://doi.org/10.1080/09603100701857906

    Article  Google Scholar 

  • Huang, J., & Wang, L. (2022). Hybrid EVT-GARCH model for tail risk estimation. Journal of Risk Management, 39(1), 123–143.

    Google Scholar 

  • Islamaj, E., Kose, M. A., Ohnsorge, F. L., & Ye, L. S. (2019). Explaining recent investment weakness: Causes and implications. Emerging Markets Finance and Trade, 55(8), 1709–1721. https://doi.org/10.1080/1540496X.2018.1530105

    Article  Google Scholar 

  • Jiang, T., & Zhu, H. (2022). A Hybrid GARCH-EVT approach for VaR estimation. Journal of Risk Analysis, 42(2), 201–222.

    Google Scholar 

  • Jordan, J. V., & Mackay, R. J. (1995). Assessing value at risk for equity portfolios: Implementing alternative techniques. Working Paper, George Washington University, Washington, DC.

  • Kenourgios, D., Samitas, A., & Paltalidis, N. (2011). Financial crises and stock market contagion in a multivariate time-varying asymmetric framework. Journal of International Financial Markets, Institutions and Money, 21(1), 92–106. https://doi.org/10.1016/j.intfin.2010.08.004

    Article  Google Scholar 

  • Kupiec, P. (2010). VaR implementation benchmarking (VIB) survey: How top asset managers determine portfolio risk. Journal of Risk Management in Financial Institutions, 3(2), 113–134.

    Google Scholar 

  • Lee, Y., & Poon, S. H. (2015). Loan portfolio loss models wit more flexible asymmetry and tails for Korean banks and a comparison of their regional concentrations. Emerging Markets Finance and Trade, 51, 118–139. https://doi.org/10.1080/1540496X.2015.1039864

    Article  Google Scholar 

  • Li, C., Zhang, H., & Liu, X. (2021). Hybrid GARCH-skewed t distribution model for VaR estimation. International Review of Financial Analysis, 28(4), 65–82.

    Google Scholar 

  • Linsmeier, T. J., & Pearson, N. D. (1996). Risk Measurement: An introduction to value at risk. Working Paper, University of Illinois at Urbana Champaign.

  • Longerstaey, J., & Spencer, M. (1996). RiskMetricsTM—Technical document (4th ed.).

  • Longin, F. (1996). The asymptotic distribution of extreme stock market returns. Journal of Business, 69(3), 383–408. https://doi.org/10.1086/209695

    Article  Google Scholar 

  • Lu, Y., Yang, L., Shi, B. F., Li, J. X., & Abedin, M. Z. (2022). A novel framework of credit risk feature selection for SMEs during industry 4.0. Annals of Operations Research. https://doi.org/10.1007/s10479-022-04849-3

    Article  PubMed  PubMed Central  Google Scholar 

  • McMillan, D. G., & Kambouroudis, D. (2009). Are RiskMetrics forecasts good enough? Evidence from 31 stock markets. International Review of Financial Analysis, 18(3), 117–124. https://doi.org/10.1016/j.irfa.2009.03.006

    Article  Google Scholar 

  • McNeil, A., Frey, R., & Embrechts, P. (2005). Quantitative risk management: Concepts. Princeton University Press.

    Google Scholar 

  • Medina-Olivares, V., Calabrese, R., Dong, Y. Z., & Shi, B. F. (2022). Spatial dependence in microfinance credit default. International Journal of Forecasting, 38(3), 1071–1085. https://doi.org/10.1016/j.ijforecast.2021.05.009

    Article  Google Scholar 

  • Merton, R. C. (1974). On the pricing of corporate debt: The risk structure of interest rates. The Journal of Finance, 29(2), 449–470. https://doi.org/10.2307/2978814

    Article  Google Scholar 

  • Nawrocki, D., & Harding, B. (2011). A hybrid value at risk approach using GARCH and historical simulation. Review of Financial Economics, 20(2), 49–57.

    Google Scholar 

  • Palaro, H. P., & Hotta, L. K. (2006). Using conditional copula to estimate value at risk. Journal of Data Science, 4(3), 93–115.

    Google Scholar 

  • Rahman, A., Hossain, M. S., & Alam, M. J. (2020). Non-performing loans and financial stability: A study on banking sector of Bangladesh. Journal of Financial Stability, 10(1), 65–79.

    Google Scholar 

  • Rahman, F., & Ullah, M. (2022). Regulatory reforms and non-performing loans: Lessons from Bangladesh’s banking sector. Journal of Banking Regulation, 15(4), 54–68.

    Google Scholar 

  • Rahman, M. A., & Khan, S. H. (2017). Credit risk management practices in commercial banks of Bangladesh: A study on basic bank Ltd. International Journal of Business and Management, 5(6), 46–54.

    Google Scholar 

  • Rockafeller, T., & Uryasev, S. (2000). Optimization of conditional value-at-risk. Journal of Risk, 2(3), 21–41.

    Article  Google Scholar 

  • Rombouts, J. V. K., & Stentoft, L. (2014). Bayesian option pricing using mixed normal heteroskedasticity models. Quantitative Finance, 14(11), 2017–2031.

    Google Scholar 

  • Siddiqui, T., & Hossain, M. (2018). The nexus between political interference and non-performing loans: Evidence from Bangladeshi banking sector. Asian Journal of Economic Modelling, 6(1), 67–80.

    Google Scholar 

  • Song, X., Liu, Y., & Wu, Z. (2022). Vine copulas for portfolio VaR and ES estimation. Journal of Risk Analysis, 47(2), 345–367.

    Google Scholar 

  • Su, E., & Knowles, T. W. (2006). Asian Pacific stock market volatility modeling and value at risk analysis. Emerging Markets Finance and Trade, 42(2), 18–62. https://doi.org/10.2753/REE1540-496X420202

    Article  Google Scholar 

  • Sun, Y., Chai, N. N., Dong, Y. Z., & Shi, B. F. (2022). Assessing and predicting small industrial enterprises’ credit ratings: A fuzzy decision making approach. International Journal of Forecasting, 38(3), 1158–1172. https://doi.org/10.1016/j.ijforecast.2022.01.006

    Article  Google Scholar 

  • Taylor, T. (2003). Historical simulations and the future of the historical narrative. MPublishing, University of Michigan Library.

    Google Scholar 

  • Wu, T., Huang, H. H., Wang, C. P., & Zhong, Y. L. (2020). The influences of book-to-price ratio and stock capitalization on value-at-risk estimation in Taiwan stock market. Emerging Markets Finance and Trade, 56(5), 1055–1072. https://doi.org/10.1080/1540496X.2018.1509790

    Article  Google Scholar 

Download references

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Author information

Authors and Affiliations

  1. Department of Mathematics, University of Dhaka, Dhaka, Bangladesh

    Sharif Mozumder

  2. Department of Accounting and Finance, School of Management, Swansea University, Bay Campus, Fabian Way, Swansea, SA1 8EN, Wales, UK

    Mohammad Zoynul Abedin

  3. Department of Banking and Insurance, University of Dhaka, Dhaka, Bangladesh

    Raad Lalon

  4. Department of Mathematics, University of Chittagong, Chittagong, Bangladesh

    Amjad Hossain

Authors
  1. Sharif Mozumder
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Mohammad Zoynul Abedin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Raad Lalon
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Amjad Hossain
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Sharif Mozumder.

Ethics declarations

Conflict of interest

The authors declared no potential conflicts of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Mozumder, S., Abedin, M.Z., Lalon, R. et al. Which User-Friendly Model is the Best for BASEL-III? An Emerging Market Study. Comput Econ (2024). https://doi.org/10.1007/s10614-023-10545-6

Download citation

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1007/s10614-023-10545-6

Keywords

JEL Classification

Search

Navigation