Skip to main content
SpringerLink
Log in

Adaptive online portfolio strategy based on exponential gradient updates

  • Published:
Journal of Combinatorial Optimization Aims and scope Submit manuscript

Abstract

Based on the momentum principle and adaptive learning mechanism, we design online portfolio selection strategies, which are suitable for nonstationary financial market. Firstly, we propose a Moving-window-based Adaptive Exponential Gradient (MAEG) strategy, which updates the learning rate of the EG algorithm by maximizing the recent cumulative return using the price data in a fixed length moving window. Secondly, we consider a special case where all-historical price data is used to design another strategy named Adaptive Exponential Gradient (AEG). Finally, we conduct an extensive numerical analysis using real price data and the empirical results show that the performance of the proposed strategies is steadily superior to some online strategies. In addition, MAEG and AEG are able to withstand certain transaction costs, which further supports their practical applicability in trading applications.

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

Notes

  1. Comb. 1 contains the following stocks: Alcoa, GM, IBM, Kin Ark, Kodak; Comb. 2 contains the following stocks: Alcoa, American Brands, GM, IBM, Kin Ark, Kodak; Comb. 3 contains the following stocks: Dupont, Ford, GE, GM, IBM, Kin Ark; Comb. 4 contains the following stocks: AHP, Dupont, Ford, GE, GM, IBM, JNJ, Kin Ark, MMM, Sherman Williams; Comb. 5 contains the following stocks: 600757, 600171, 000060, 000830, 000750; Comb. 6 contains the following stocks: 600757, 600171, 000060, 000830, 000750, 000930; Comb. 7 contains the following stocks: 600757, 600171, 000060, 000830, 000750, 000930, 000062, 600060; Comb. 8 contains the following stocks: 000858, 000725, 600585, 600028, 600741, 000898, 600104, 600115, 600027, 600016, 000625, 000069, 600674, 600036, 600663, 000063, 000157, 600066, 000415, 000627, 600271, 600118, 600886, 000768, 600018, 600170, 600068, 600177, 002044.

References

  • Agarwal A, Hazan E, Kale S, Schapire RE (2006) Algorithms for portfolio management based on the Newton method. In: ACM: Proceedings of the 23rd international conference on machine learning, pp 9–16

  • Bean AJ, Singer AC (2012) Universal switching and side information portfolios under transaction costs using factor graphs. IEEE J Select Topics Sig Process 6(4):351–365

    Article  Google Scholar 

  • Blum A, Kalai A (1999) Universal portfolios with and without transaction costs. Mach Learn 35(3):193–205

    Article  Google Scholar 

  • Borodin A, El-Yaniv R, Gogan V (2004) Can we learn to beat the best stock. J Artif Intell Res 21:579–594

    Article  MathSciNet  Google Scholar 

  • Cai X, Ye ZK (2019) Gaussian weighting reversion strategy for accurate online portfolio selection. IEEE Trans Sig Process 67(21):5558–5570

    Article  Google Scholar 

  • Chai X, Li WH, Yuan H, Wang LB (2020) Online scheduling on a single machine with linear deteriorating processing times and delivery times. J Combin Optim. https://doi.org/10.1007/s10878-020-00557-5

  • Chu G, Zhang W, Sun GF, Zhang XT (2019) A new online portfolio selection algorithm based on Kalman Filter and anti-correlation. Physi A Stat Mech Appl 536(15):Article 120949

  • Cover TM (1991) Universal portfolios. Math Finance 1(1):1–29

    Article  MathSciNet  Google Scholar 

  • Cover TM, Ordentlich E (1996) Universal portfolios with side information. IEEE Trans Inform Theory 42(2):348–363

    Article  MathSciNet  Google Scholar 

  • Dai WQ, Zheng M, Chen X, Yang ZL (2020) Online economic ordering problem for deteriorating items with limited price information. J Combin Optim. https://doi.org/10.1007/s10878-020-00603-2

  • Feng X, Xu YF, Ni GQ, Dai YW (2018) Online leasing problem with price fluctuations under the consumer price index. J Combin Optim 36:493–507

    Article  MathSciNet  Google Scholar 

  • Gaivoronski AA, Stella F (2000) Stochastic nonstationary optimization for finding universal portfolios. Ann Oper Res 100(1–4):165–188

    Article  MathSciNet  Google Scholar 

  • Guan H, An ZY (2019) A local adaptive learning system for online portfolio selection. Knowl Based Syst 186:Article 104958

  • Helmbold DP, Schapire RE, Singer Y, Warmuth MK (1998) On-line portfolio selection using multiplicative updates. Math Finance 8(4):325–347

    Article  Google Scholar 

  • Huang DJ, Zhou JL, Li B, Hoi SCH, Zhou SG (2016) Robust median reversion strategy for online portfolio selection. IEEE Trans Knowl Data Eng 28(9):2480–2493

    Article  Google Scholar 

  • Huang DJ, Yu SC, Li B, Hoi SCH, Zhou SG (2018) Combination forecasting reversion strategy for online portfolio selection. ACM Trans Intell Syst Technol 9(5):Article 58

  • Jegadeesh N (1990) Evidence of predictable behavior of security returns. J Finance 45(3):881–898

    Article  Google Scholar 

  • Kelly JL (1956) A new interpretation of information rate. Bell Syst Tech J 35(4):917–926

    Article  MathSciNet  Google Scholar 

  • Khedmati M, Azin P (2020) An online portfolio selection algorithm using clustering approaches and considering transaction costs. Exp Syst Appl 159(30):Article 113546

  • Lai ZR, Dai DQ, Ren CX, Huang KK (2018a) A peak price tracking-based learning system for portfolio selection. IEEE Trans Neural Netw Learning Syst 29(7):2823–2832

    MathSciNet  Google Scholar 

  • Lai ZR, Dai DQ, Ren CX, Huang KK (2018b) Radial basis functions with adaptive input and composite trend representation for portfolio selection. IEEE Trans Neural Netw Learn Syst 29(12):6214–6226

    Article  Google Scholar 

  • Lai ZR, Yang PY, Fang LD, Wu XT (2018c) Short-term sparse portfolio optimization based on alternating direction method of multipliers. J Mach Learn Res 19(1):1–28

    MathSciNet  MATH  Google Scholar 

  • Lai ZR, Yang PY, Wu XT, Fang LD (2018d) A kernel-based trend pattern tracking system for portfolio optimization. Data Mining Knowl Discov 32:1708–1734

    Article  MathSciNet  Google Scholar 

  • Lai ZR, Yang PY, Fang LD, Wu XT (2020) Reweighted price relative tracking system for automatic portfolio optimization. IEEE Trans Syst Man Cybernet Syst 50(11):4349–4361

    Article  Google Scholar 

  • Li B, Hoi SCH, Gopalkrishnan V (2011) CORN: correlation-driven nonparametric learning approach for portfolio selection. ACM Trans Intell Syst Technol 2(3):Article 21

  • Li B, Zhao PL, Hoi SCH, Gopalkrishnan V (2012) PAMR: passive aggressive mean reversion strategy for portfolio selection. Mach Learn 87(2):221–258

    Article  MathSciNet  Google Scholar 

  • Li B, Hoi SCH, Sahoo D, Liu ZY (2015) Moving average reversion strategy for on-line portfolio selection. Artif Intell 222:104–123

    Article  MathSciNet  Google Scholar 

  • Li B, Wang JL, Huang DJ, Hoi SCH (2018) Transaction cost optimization for online portfolio selection. Quant Finance 18(8):1411–1424

    Article  MathSciNet  Google Scholar 

  • Li TF, Chen K, Feng Y, Ying ZL (2017) Binary switch portfolio. Quant Finance 17(5):763–780

    Article  MathSciNet  Google Scholar 

  • Lin X, Zhang M, Zhang YF, Gu ZQ, Liu YQ, Ma SP (2017) Boosting moving average reversion strategy for online portfolio selection: a meta-learning approach. In: International conference on database systems for advanced applications (DASFAA 2017), pp 494–510

  • Lo AW, MacKinlay AC (1990) When are contrarian profits due to stock market overreaction? Rev Finan Stud 3(2):175–205

    Article  Google Scholar 

  • Ma YL, Han RZ, Wang WZ (2021) Portfolio optimization with return prediction using deep learning and machine learning. Exp Syst Appl 165(1):Article 113973

  • Markowitz H (1952) Portfolio selection. J Finance 7(1):77–91

    Google Scholar 

  • O’Sullivan P, Edelman D (2015) Adaptive universal portfolios. Eur J Finance 21(4):337–351

  • Sezer OB, Gudelek MU, Ozbayoglu AM (2020) Financial time series forecasting with deep learning: a systematic literature review: 2005–2019. Appl Soft Comput 90:Article 106181

  • Singer Y (1997) Switching portfolios. Int J Neural Syst 8(4):445–455

    Article  Google Scholar 

  • Vovk VG, Watkins C (1998) Universal portfolio selection. In: Proceedings of the eleventh annual conference on computational learning theory, COLT 1998. Madison, Wisconsin, USA, July 24-26, 1998, pp 12–23

  • Yang XY, He JA, Lin H, Zhang Y (2020a) Boosting exponential gradient strategy for online portfolio selection: an aggregating experts’ advice method. Comput Econ 55:231–251

  • Yang XY, He JA, Xian JY, Lin H, Zhang Y (2020b) Aggregating expert advice strategy for online portfolio selection with side information. Soft Comput 24(3):2067–2081

    Article  Google Scholar 

  • Yu JR, Paul Chiou WJ, Lee WY, Lin SJ (2020) Portfolio models with return forecasting and transaction costs. Int Rev Econ Finance 66:118–130

    Article  Google Scholar 

Download references

Acknowledgements

This research was supported by the National Natural Science Foundation of China (No. 71501049), the Humanities and Social Science Foundation of the Ministry of Education of China (No. 21YJA630117) and the Philosophy and Social Sciences Planning Project of Guangdong Province (GD19CGL06).

Author information

Authors and Affiliations

  1. School of Management, Guangdong University of Technology, Guangzhou, 510520, Guangdong, People’s Republic of China

    Yong Zhang, Hong Lin, Lina Zheng & Xingyu Yang

Authors
  1. Yong Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Hong Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Lina Zheng
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xingyu Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xingyu Yang.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict 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

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhang, Y., Lin, H., Zheng, L. et al. Adaptive online portfolio strategy based on exponential gradient updates. J Comb Optim 43, 672–696 (2022). https://doi.org/10.1007/s10878-021-00800-7

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10878-021-00800-7

Keywords

Mathematics Subject Classification

Search

Navigation