⚠ Official Notice: www.ijisrt.com is the official website of the International Journal of Innovative Science and Research Technology (IJISRT) Journal for research paper submission and publication. Please beware of fake or duplicate websites using the IJISRT name.



Volatility Analysis of Cryptocurrency Markets: A Comparative Study of EGARCH, LSTM and Hybrid Forecasting Models for Bitcoin and Ethereum


Authors : Keith Reuben Muyungi; Hu Junjuan

Volume/Issue : Volume 11 - 2026, Issue 8 - August

Google Scholar : https://tinyurl.com/zf5stpew

DOI : https://doi.org/10.38124/ijisrt/26aug1506

Semantic Scholar

ResearchGate

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : This paper compares econometric and deep learning approaches to one-day-ahead volatility forecasting for Bitcoin (BTC/USD) and Ethereum (ETH/USD) over the period from January 2016 to December 2025. Seven models are evaluated: a symmetric GARCH(1,1) with Student-t innovations, an asymmetric EGARCH(1,1) with Student-t innovations, a Long Short-Term Memory neural network, two hybrid LSTM specifications, a constant unconditional-variance benchmark, and a naive random walk. Hybrid-A augments the LSTM input with the EGARCH conditional variance forecast. Hybrid-B applies the LSTM to the standardised residuals of the fitted EGARCH model. To guard against the welldocumented sensitivity of recurrent neural networks to random initialisation, the LSTM-based models are estimated under ten different random seeds and the median forecast is reported. Squared daily log returns serve as the primary volatility proxy, with the Parkinson and Garman-Klass range-based estimators used as robustness benchmarks. Forecasts are produced through an expanding-window rolling scheme with monthly refits over the test period from January 2023 to December 2025, yielding 1,096 out-of-sample forecasts per model and asset. Statistical inference is based on the DieboldMariano test with the Harvey-Leybourne-Newbold correction and the Model Confidence Set procedure of Hansen, Lunde, and Nason (2011). The principal finding is that the relative ranking of cryptocurrency volatility forecasts is highly sensitive to the choice of loss function but stable across alternative volatility proxies. Under MSE, the LSTM achieves the lowest error and is the only model retained in the 90% Model Confidence Set for Ethereum, while several models survive jointly for Bitcoin. Under QLIKE, which is robust to noise in the squared-return proxy and penalises variance underestimation, both GARCH and EGARCH survive the Model Confidence Set jointly with the LSTM and Hybrid-B for Ethereum, and GARCH retains a place for Bitcoin. The LSTM-based models beat the constant unconditional-variance benchmark by 37% to 72% across proxies, confirming that the LSTM captures genuine volatility dynamics rather than merely predicting the mean. The variance-as-feature hybrid is dominated by the pure LSTM and exhibits substantially higher seed-to-seed variability than the other deep learning specifications, while the residual-based hybrid effectively collapses to the pure LSTM. Subperiod analysis confirms that these patterns are stable across the 2023, 2024, and post-spot-ETF 2025 windows. The findings indicate that the apparent superiority of LSTM models in cryptocurrency volatility forecasting is partly an artefact of evaluation under squared error and that simple parametric models retain a useful role under loss-consistent inference.

Keywords : Cryptocurrency Volatility; EGARCH; Long Short-Term Memory; Hybrid Forecasting.

References :

  1. Alqaralleh, H., Abuhommous, A. A., & Alsaraireh, A. (2020). Forecasting cryptocurrency volatility: A comparison between GARCH-type and machine learning models. Investment Management and Financial Innovations, 17(4), 165–176.
  2. Alves, R. P., Brito, F. C. C., Mendes, P. S., & Cajueiro, D. O. (2025). Volatility transmission between cryptocurrencies and traditional financial assets: Evidence from a TVP-VAR-DY framework. Finance Research Letters, 65, 105827.
  3. Amendola, A., & Sensini, L. (2020). Conditional heteroskedastic models for cryptocurrencies volatility. International Business Research, 13(2), 71–82.
  4. Amirshahi, B., & Lahmiri, S. (2023). Hybrid deep learning and GARCH-family models for forecasting volatility of cryptocurrencies. Machine Learning with Applications, 12, 100465.
  5. Andersen, T. G., & Bollerslev, T. (1998). Answering the skeptics: Yes, standard volatility models do provide accurate forecasts. International Economic Review, 39(4), 885–905.
  6. Andersen, T. G., Bollerslev, T., Diebold, F. X., & Labys, P. (2003). Modeling and forecasting realized volatility. Econometrica, 71(2), 579–625.
  7. Aras, S. (2021). Stacking hybrid GARCH models for forecasting Bitcoin volatility. Expert Systems with Applications, 174, 114747.
  8. Aras, S., Özdemir, M. O., & Çılgın, C. (2025). Uncertainty or investor attention: Which has more impact on Bitcoin volatility? Research in International Business and Finance.
  9. Araya, H. T., Aduda, J., & Berhane, T. (2024). A hybrid GARCH and deep learning method for volatility prediction. Journal of Applied Mathematics, 2024, 6305525.
  10. Ardia, D., Bluteau, K., & Rüede, M. (2019). Regime changes in Bitcoin GARCH volatility dynamics. Finance Research Letters, 29, 266–271.
  11. Ayadi, A., Ghabri, Y., & Guesmi, K. (2025). Volatility of top market-cap cryptocurrencies during COVID-19 and the Russia–Ukraine war. Risks, 13(3), 57.
  12. Babalos, V., Bouri, E., & Gupta, R. (2025). Does the introduction of US spot Bitcoin ETFs affect spot returns and volatility of major cryptocurrencies? The Quarterly Review of Economics and Finance, 102, 102006.
  13. Bates, J. M., & Granger, C. W. J. (1969). The combination of forecasts. Operational Research Quarterly, 20(4), 451–468.
  14. Bergsli, L. Ø., Lind, A. F., Molnár, P., & Polasik, M. (2022). Forecasting volatility of Bitcoin. Research in International Business and Finance, 59, 101540.
  15. Bernardi, M., & Catania, L. (2018). The Model Confidence Set package for R. International Journal of Computational Economics and Econometrics, 8(2), 144–158.
  16. Bollerslev, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31(3), 307–327.
  17. Brauneis, A., & Sahiner, M. (2024). Crypto volatility forecasting: Mounting a HAR, sentiment, and machine learning horserace. Asia-Pacific Financial Markets.
  18. Buczynski, M., & Chlebus, M. (2024). GARCHNet: Value-at-risk forecasting with GARCH models based on neural networks. Computational Economics, 63(5), 1949–1979.
  19. Caporale, G. M., & Zekokh, T. (2019). Modelling volatility of cryptocurrencies using Markov-switching GARCH models. Research in International Business and Finance, 48, 143–155.
  20. Catania, L., & Grassi, S. (2022). Forecasting cryptocurrency volatility. International Journal of Forecasting, 38(3), 878–894.
  21. Cheikh, N. B., Zaied, Y. B., & Chevallier, J. (2020). Asymmetric volatility in cryptocurrency markets. Finance Research Letters, 35, 101293.
  22. Chronopoulos, I., Raftapostolos, A., & Kapetanios, G. (2024). Forecasting value-at-risk using deep neural network quantile regression. Journal of Financial Econometrics, 22(3), 636–669.
  23. Conrad, C., Custovic, A., & Ghysels, E. (2018). Long- and short-term cryptocurrency volatility components: A GARCH-MIDAS analysis. Journal of Risk and Financial Management, 11(2), 23.
  24. Corbet, S., Lucey, B., Urquhart, A., & Yarovaya, L. (2019). Cryptocurrencies as a financial asset: A systematic analysis. International Review of Financial Analysis, 62, 182–199.
  25. Corsi, F. (2009). A simple approximate long-memory model of realized volatility. Journal of Financial Econometrics, 7(2), 174–196.
  26. Cretarola, A., & Figà-Talamanca, G. (2020). Bubble regime identification in an attention-based model for Bitcoin and Ethereum price dynamics. Economics Letters, 191, 108831.
  27. D'Amato, V., Levantesi, S., & Piscopo, G. (2022). Deep learning in predicting cryptocurrency volatility. Physica A: Statistical Mechanics and Its Applications, 596, 127158.
  28. Diebold, F. X., & Mariano, R. S. (1995). Comparing predictive accuracy. Journal of Business & Economic Statistics, 13(3), 253–263.
  29. Dudek, G., Fiszeder, P., Kobus, P., & Orzeszko, W. (2024). Forecasting cryptocurrencies volatility using statistical and machine learning methods: A comparative study. Applied Soft Computing, 151, 111132.
  30. Fakhfekh, M., & Jeribi, A. (2020). Volatility dynamics of crypto-currencies' returns: Evidence from asymmetric and long memory GARCH models. Research in International Business and Finance, 51, 101075.
  31. Fiszeder, P., Małecka, M., & Molnár, P. (2024). Range-based DCC models for covariance and value-at-risk forecasting. Journal of Empirical Finance, 75, 101452.
  32. García-Medina, A., & Aguayo-Moreno, E. (2024). LSTM-GARCH hybrid model for the prediction of volatility in cryptocurrency portfolios. Computational Economics, 63(4), 1511–1542.
  33. Granger, C. W. J., & Ramanathan, R. (1984). Improved methods of combining forecasts. Journal of Forecasting, 3(2), 197–204.
  34. Hansen, P. R., & Lunde, A. (2006). Consistent ranking of volatility models. Journal of Econometrics, 131(1–2), 97–121.
  35. Hansen, P. R., Lunde, A., & Nason, J. M. (2011). The model confidence set. Econometrica, 79(2), 453–497.
  36. Harvey, D., Leybourne, S., & Newbold, P. (1997). Testing the equality of prediction mean squared errors. International Journal of Forecasting, 13(2), 281–291.
  37. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780.
  38. Huang, J.-Z., Huang, W., & Ni, J. (2024). Predicting Bitcoin volatility using ensemble machine learning. Journal of Forecasting, 43(5), 1357–1374.
  39. Katsiampa, P. (2017). Volatility estimation for Bitcoin: A comparison of GARCH models. Economics Letters, 158, 3–6.
  40. Katsiampa, P. (2019). Volatility co-movement between Bitcoin and Ether. Finance Research Letters, 30, 221–227.
  41. Kaya Soylu, P., Okur, M., Çatıkkaş, Ö., & Altintig, Z. A. (2020). Long memory in the volatility of selected cryptocurrencies: Bitcoin, Ethereum and Ripple. Journal of Risk and Financial Management, 13(6), 107.
  42. Khaldi, R., El Afia, A., Chiheb, R., & Faizi, R. (2019). Forecasting of BTC volatility: Comparative study between parametric and non-parametric models. Progress in Artificial Intelligence, 8, 511–523.
  43. Köchling, G., Schmidtke, P., & Posch, P. N. (2020). Volatility forecasting accuracy for Bitcoin. Economics Letters, 191, 108836.
  44. Kristjanpoller, W., & Minutolo, M. C. (2018). A hybrid volatility forecasting framework integrating GARCH, artificial neural network, technical analysis and principal components analysis. Expert Systems with Applications, 109, 1–11.
  45. Liu, L. Y., Patton, A. J., & Sheppard, K. (2015). Does anything beat 5-minute RV? A comparison of realized measures across multiple asset classes. Journal of Econometrics, 187(1), 293–311.
  46. Maciel, L. (2021). Cryptocurrencies value-at-risk and expected shortfall: Do regime-switching volatility models improve forecasting? International Journal of Finance & Economics, 26(3), 4054–4072.
  47. Mohamed, B. T., Awad, S., & Reda, M. (2025). A comparative analysis of GARCH-family models for Bitcoin volatility. Journal of Financial Innovation.
  48. Nelson, D. B. (1991). Conditional heteroskedasticity in asset returns: A new approach. Econometrica, 59(2), 347–370.
  49. Ngunyi, A., Mundia, S., & Omari, C. (2019). Modelling volatility dynamics of cryptocurrencies using GARCH models. Journal of Mathematical Finance, 9(4), 591–615.
  50. Ormaniec, W., Pitera, M., Safarveisi, S., & Schmidt, T. (2024). Estimating value-at-risk using quantile regression and implied volatilities. Journal of Risk, 26(4), 23–58.
  51. Patton, A. J. (2011). Volatility forecast comparison using imperfect volatility proxies. Journal of Econometrics, 160(1), 246–256.
  52. Petnehazi, G. (2021). Quantile convolutional neural networks for value-at-risk forecasting. Machine Learning with Applications, 6, 100096.
  53. Phan Huy, T., Pham, K. H., Nguyen Thi Thuy, L., Nguyen Nhat, N., & Le Danh, B. (2025). Volatility dynamics in cryptocurrencies: Comparing traditional and hybrid deep learning predictive techniques. Journal of Dynamics and Control.
  54. Politis, D. N., & Romano, J. P. (1994). The stationary bootstrap. Journal of the American Statistical Association, 89(428), 1303–1313.
  55. Pratas, T. E., Ramos, F. R., & Rubio, L. (2023). Forecasting Bitcoin volatility. Eurasian Economic Review, 13(2), 285–305.
  56. Quaedvlieg, R. (2021). Multi-horizon forecast comparison. Journal of Business & Economic Statistics, 39(1), 40–53.
  57. Rodikov, G., & Antulov-Fantulin, N. (2022). Can LSTM outperform volatility-econometric models? arXiv:2202.11581.
  58. Seo, M., & Kim, G. (2020). Hybrid forecasting models based on the neural networks for the volatility of Bitcoin. Applied Sciences, 10(14), 4768.
  59. Sözen, E. (2025). Volatility dynamics of cryptocurrencies: A comparative analysis using GARCH-family models. Future Business Journal, 11, 166.
  60. Staudemeyer, R. C., & Morris, E. R. (2019). Understanding LSTM: A tutorial. arXiv:1909.09586.
  61. Trucíos, C. (2019). Forecasting Bitcoin risk measures: A robust approach. International Journal of Forecasting, 35(3), 836–847.
  62. Trucíos, C., & Taylor, J. W. (2023). A comparison of methods for forecasting value-at-risk and expected shortfall of cryptocurrencies. Journal of Forecasting, 42(4), 989–1007.
  63. Walther, T., Klein, T., & Bouri, E. (2019). Exogenous drivers of Bitcoin and cryptocurrency volatility. SSRN Working Paper No. 3192474.
  64. Whaley, R. E., & Hornback, A. M. (2025). Spot Bitcoin ETFs: The struggle was worth it. Financial Analysts Journal, 81(2), 39–50.
  65. Zahid, M., Iqbal, F., & Koutmos, D. (2022). Forecasting Bitcoin volatility using hybrid GARCH models with machine learning. Risks, 10(12), 237.
  66. Zhao, Y., Wu, B., & Yin, L. (2025). Integrated GARCH-GRU in financial volatility forecasting. arXiv:2504.09380.

This paper compares econometric and deep learning approaches to one-day-ahead volatility forecasting for Bitcoin (BTC/USD) and Ethereum (ETH/USD) over the period from January 2016 to December 2025. Seven models are evaluated: a symmetric GARCH(1,1) with Student-t innovations, an asymmetric EGARCH(1,1) with Student-t innovations, a Long Short-Term Memory neural network, two hybrid LSTM specifications, a constant unconditional-variance benchmark, and a naive random walk. Hybrid-A augments the LSTM input with the EGARCH conditional variance forecast. Hybrid-B applies the LSTM to the standardised residuals of the fitted EGARCH model. To guard against the welldocumented sensitivity of recurrent neural networks to random initialisation, the LSTM-based models are estimated under ten different random seeds and the median forecast is reported. Squared daily log returns serve as the primary volatility proxy, with the Parkinson and Garman-Klass range-based estimators used as robustness benchmarks. Forecasts are produced through an expanding-window rolling scheme with monthly refits over the test period from January 2023 to December 2025, yielding 1,096 out-of-sample forecasts per model and asset. Statistical inference is based on the DieboldMariano test with the Harvey-Leybourne-Newbold correction and the Model Confidence Set procedure of Hansen, Lunde, and Nason (2011). The principal finding is that the relative ranking of cryptocurrency volatility forecasts is highly sensitive to the choice of loss function but stable across alternative volatility proxies. Under MSE, the LSTM achieves the lowest error and is the only model retained in the 90% Model Confidence Set for Ethereum, while several models survive jointly for Bitcoin. Under QLIKE, which is robust to noise in the squared-return proxy and penalises variance underestimation, both GARCH and EGARCH survive the Model Confidence Set jointly with the LSTM and Hybrid-B for Ethereum, and GARCH retains a place for Bitcoin. The LSTM-based models beat the constant unconditional-variance benchmark by 37% to 72% across proxies, confirming that the LSTM captures genuine volatility dynamics rather than merely predicting the mean. The variance-as-feature hybrid is dominated by the pure LSTM and exhibits substantially higher seed-to-seed variability than the other deep learning specifications, while the residual-based hybrid effectively collapses to the pure LSTM. Subperiod analysis confirms that these patterns are stable across the 2023, 2024, and post-spot-ETF 2025 windows. The findings indicate that the apparent superiority of LSTM models in cryptocurrency volatility forecasting is partly an artefact of evaluation under squared error and that simple parametric models retain a useful role under loss-consistent inference.

Keywords : Cryptocurrency Volatility; EGARCH; Long Short-Term Memory; Hybrid Forecasting.

Paper Submission Last Date
31 - October - 2026

SUBMIT YOUR PAPER CALL FOR PAPERS
Video Explanation for Published paper

Never miss an update from Papermashup

Get notified about the latest tutorials and downloads.

Subscribe by Email

Get alerts directly into your inbox after each post and stay updated.
Subscribe
OR

Subscribe by RSS

Add our RSS to your feedreader to get regular updates from us.
Subscribe