×
Buehler, H.; Gonon, L.; Teichmann, J.; Wood, B.

Deep hedging. (English) Zbl 1420.91450

Quant. Finance 19, No. 8, 1271-1291 (2019).
Summary: We present a framework for hedging a portfolio of derivatives in the presence of market frictions such as transaction costs, liquidity constraints or risk limits using modern deep reinforcement machine learning methods. We discuss how standard reinforcement learning methods can be applied to non-linear reward structures, i.e. in our case convex risk measures. As a general contribution to the use of deep learning for stochastic processes, we also show in Section 4 that the set of constrained trading strategies used by our algorithm is large enough to \(\epsilon\)-approximate any optimal solution. Our algorithm can be implemented efficiently even in high-dimensional situations using modern machine learning tools. Its structure does not depend on specific market dynamics, and generalizes across hedging instruments including the use of liquid derivatives. Its computational performance is largely invariant in the size of the portfolio as it depends mainly on the number of hedging instruments available. We illustrate our approach by an experiment on the S&P500 index and by showing the effect on hedging under transaction costs in a synthetic market driven by the Heston model, where we outperform the standard ‘complete-market’ solution.

Cited in 2 Reviews
Cited in 41 Documents

MSC:

91G20 Derivative securities (option pricing, hedging, etc.)
91G10 Portfolio theory

Keywords:

reinforcement learning; machine learning; market frictions; transaction costs; hedging; risk management; portfolio optimization

Software:

Adam
PDF BibTeX XML Cite
\textit{H. Buehler} et al., Quant. Finance 19, No. 8, 1271--1291 (2019; Zbl 1420.91450)
Full Text: DOI arXiv

References:

[1] Alexander, C. and Nogueira, L.M., Model-free hedge ratios and scale-invariant models. J. Bank. Finance, 2007, 31, 1839-1861. doi: 10.1016/j.jbankfin.2006.11.011
[2] Andersen, L.B.G., Jäckel, P. and Kahl, C., Simulation of square-root processes. In Encyclopedia of Quantitative Finance, 2010 (John Wiley and Sons).
[3] Bank, P., Soner, H.M. and Voß, M., Hedging with temporary price impact. Math. Financ. Econ., 2017, 11, 215-239. doi: 10.1007/s11579-016-0178-4 · Zbl 1409.91226
[4] Barles, G. and Soner, H.M., Option pricing with transaction costs and a nonlinear Black-Scholes equation. Finance Stoch., 1998, 2, 369-397. doi: 10.1007/s007800050046 · Zbl 0915.35051
[5] Bates, D.S., Hedging the smirk. Finance Res. Lett., 2005, 2, 195-200. doi: 10.1016/j.frl.2005.08.004
[6] Ben-Tal, A. and Teboulle, M., An old-new concept of convex risk measures: The optimized certainty equivalent. Math. Finance, 2007, 17, 449-476. doi: 10.1111/j.1467-9965.2007.00311.x · Zbl 1186.91116
[7] Bölcskei, H., Grohs, P., Kutyniok, G. and Petersen, P., Optimal approximation with sparsely connected deep neural networks. Preprint arXiv:1705.01714, 2017. · Zbl 1499.41029
[8] Bouchard, B., Moreau, L. and Soner, H.M., Hedging under an expected loss constraint with small transaction costs. SIAM J. Financ. Math., 2016, 7, 508-551. doi: 10.1137/15M1006787 · Zbl 1345.60030
[9] Broadie, M. and Kaya, O., Exact simulation of Stochastic volatility and other affine jump diffusion processes. Oper. Res., 2006, 54, 217-231. doi: 10.1287/opre.1050.0247 · Zbl 1167.91363
[10] Burgert, C. and Rüschendorf, L., Consistent risk measures for portfolio vectors. Insurance: Math. Econ., 2006, 38, 289-297. · Zbl 1138.91490
[11] Carmona, R., Indifference Pricing: Theory and Applications, 2009 (Princeton University Press: Princeton, NJ). · Zbl 1155.91008
[12] Crépey, S., Delta-hedging vega risk?Quant. Finance, 2004, 4, 559-579. doi: 10.1080/14697680400000038 · Zbl 1405.91604
[13] Davis, M.H.A., Panas, V.G. and Zariphopoulou, T., European option pricing with transaction costs. SIAM J. Control Optim., 1993, 31, 470-493. doi: 10.1137/0331022 · Zbl 0779.90011
[14] Du, X., Zhai, J. and Lv, K., Algorithm trading using q-learning and recurrent reinforcement learning. arxiv, 2009. Available online at: https://arxiv.org/pdf/1707.07338.pdf.
[15] Dufresne, D., The integrated square-root process, Centre for Actuarial Studies, Research Paper no. 90, University of Melbourne, 2001. · Zbl 0987.60026
[16] Dupire, B., Pricing with a smile. Risk, 1994, 7, 18-20.
[17] Föllmer, H. and Leukert, P., Efficient hedging: Cost versus shortfall risk. Finance Stoch., 2000, 4, 117-146. doi: 10.1007/s007800050008 · Zbl 0956.60074
[18] Föllmer, H. and Schied, A., Stochastic Finance: An Introduction in Discrete Time, 2016 (De Gruyter: Berlin, Germany). · Zbl 1343.91001
[19] Gatheral, J. and Schied, A., Dynamical models of market impact and algorithms for order execution. In Handbook on Systemic Risk, pp. 579-599, 2013.
[20] Glasserman, P., Monte Carlo Methods in Financial Engineering, Applications of Mathematics: Stochastic Modelling and Applied Probability, 2004 (Springer: New York). · Zbl 1038.91045
[21] Goodfellow, I., Bengio, Y. and Courville, A., Deep Learning, 2016 (MIT Press). Available online at: http://www.deeplearningbook.org.
[22] Halperin, I., QLBS: Q-learner in the Black-Scholes (-Merton) Worlds. arxiv, 2017. Available online at: https://arxiv.org/abs/1712.04609.
[23] Hodges, S. and Neuberger, A., Optimal replication of contingent claims under transaction costs. Rev. Futures Markets, 1989, 8, 222-239.
[24] Hornik, K., Approximation capabilities of multilayer feedforward networks. Neural. Netw., 1991, 4, 251-257. doi: 10.1016/0893-6080(91)90009-T
[25] Hull, J. and White, A., Optimal delta hedging for options. J. Bank. Finance, 2017, 82, 180-190. doi: 10.1016/j.jbankfin.2017.05.006
[26] Hutchinson, J.M., Lo, A.W. and Poggio, T., A Nonparametric approach to pricing and hedging derivative securities via learning networks. J. Finance., 1994, 49, 851-889. doi: 10.1111/j.1540-6261.1994.tb00081.x
[27] İlhan, A., Jonsson, M. and Sircar, R., Optimal static-dynamic hedges for exotic options under convex risk measures. Stoch. Process. Their. Appl., 2009, 119, 3608-3632. doi: 10.1016/j.spa.2009.06.009 · Zbl 1204.91120
[28] Ioffe, S. and Szegedy, C., Batch normalization: Accelerating deep network training by reducing internal covariate shift. In Proceedings of the 32nd International Conference on Machine Learning, pp. 448-456, 2015.
[29] Jiang, Z., Xu, D. and Liang, J., A deep reinforcement learning framework for the financial portfolio management problem. arxiv, 2017. Available online at: https://arxiv.org/abs/1706.10059.
[30] Kabanov, Y. and Safarian, M., Markets with Transaction Costs: Mathematical Theory, 2009 (Springer: Berlin, Germany). · Zbl 1186.91006
[31] Kallsen, J. and Muhle-Karbe, J., Option pricing and hedging with small transaction costs. Math. Finance, 2015, 25, 702-723. doi: 10.1111/mafi.12035 · Zbl 1347.91231
[32] Kingma, D.P. and Ba, J., Adam: A method for stochastic optimization. Proceedings of the International Conference on Learning Representations (ICLR), 2015.
[33] Klöppel, S. and Schweizer, M., Dynamic indifference valuation via convex risk measures. Math. Finance, 2007, 17, 599-627. doi: 10.1111/j.1467-9965.2007.00317.x · Zbl 1138.91502
[34] Lu, D., Agent inspired trading using recurrent reinforcement learning and LSTM neural networks. arxiv, 2017. Available online at: https://arxiv.org/pdf/1707.07338.pdf.
[35] Moody, J. and Wu, L., Optimization of trading systems and portfolios. Proceedings of the IEEE/IAFE 1997 Computational Intelligence for Financial Engineering (CIFEr), 1997, pp. 300-307.
[36] Muhle-Karbe, J., Reppen, M. and Soner, H.M., A primer on portfolio choice with small transaction costs. Annu. Rev. Financ. Econ., 2017, 9, 301-331. doi: 10.1146/annurev-financial-110716-032445
[37] Rogers, L.C.G., Why is the effect of proportional transaction costs ##img####img####img##O(δ2/3). In Mathematics of Finance, edited by G. Yin and Q. Zhang, pp. 303-308, 2004 (American Mathematical Society: Providence, RI). · Zbl 1101.91046
[38] Rogers, L.C.G. and Singh, S., The cost of illiquidity and its effects on hedging. Math. Finance, 2010, 20, 597-615. doi: 10.1111/j.1467-9965.2010.00413.x · Zbl 1232.91635
[39] Sepp, A., An approximate distribution of delta-hedging errors in a jump-diffusion model with discrete trading and transaction costs. Quant. Finance, 2012, 12, 1119-1141. doi: 10.1080/14697688.2010.494613 · Zbl 1279.91167
[40] Shaham, U., Cloninger, A. and Coifman, R.R., Provable approximation properties for deep neural networks. Appl. Comput. Harmon. Anal., 2018, 44, 537-557. doi: 10.1016/j.acha.2016.04.003 · Zbl 1390.68553
[41] Soner, H.M., Shreve, S.E. and Cvitanić, J., There is no nontrivial hedging portfolio for option pricing with transaction costs. Ann. Appl. Probab., 1995, 5, 327-355. doi: 10.1214/aoap/1177004767 · Zbl 0837.90012
[42] Whalley, A.E. and Wilmott, P., An asymptotic analysis of an optimal hedging model for option pricing with transaction costs. Math. Finance, 1997, 7, 307-324. doi: 10.1111/1467-9965.00034 · Zbl 0885.90019
[43] Xu, M., Risk measure pricing and hedging in incomplete markets. Ann. Finance, 2006, 2, 51-71. doi: 10.1007/s10436-005-0023-x · Zbl 1233.91291
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.