Abstract
Financial markets have always been a point of interest for automated systems. Due to their complex nature, financial algorithms and fintech frameworks require vast amounts of data to accurately respond to market fluctuations. This data availability is tied to the daily market evolution, so it is impossible to accelerate its acquisition. In this article, we discuss several solutions for augmenting financial datasets via synthesizing realistic time-series with the help of generative models. This problem is complex, since financial time series present very specific properties, e.g., fat-tail distribution, cross-correlation between different stocks, specific autocorrelation, cluster volatility and so on. In particular, we propose solutions for capturing cross-correlations between different stocks and for transitioning from fixed to variable length time-series without resorting to sequence modeling networks, and adapt various network architectures, e.g., fully connected and convolutional GANs, variational autoencoders, and generative moment matching networks. Finally, we tackle the problem of evaluating the quality of synthetic financial time-series. We introduce qualitative and quantitative metrics, along with a portfolio trend prediction framework that validates our generative models’ performance. We carry out experiments on real-world financial data extracted from the US stock market, proving the benefits of these techniques.
- [1] . 2019. From handcrafted to deep-learning-based cancer radiomics: Challenges and opportunities. IEEE Sig. Process. Mag. 36, 4 (2019), 132–160.Google Scholar
Cross Ref
- [2] . 2017. Wasserstein generative adversarial networks. In Proceedings of the International Conference on Machine Learning. PMLR, 214–223.Google Scholar
- [3] . 2020. Wav2vec 2.0: A framework for self-supervised learning of speech representations. Adv. Neural Inf. Process. Syst. 33 (2020).Google Scholar
- [4] . 2018. Zero-shot object detection. In Proceedings of the European Conference on Computer Vision (ECCV). 384–400.Google Scholar
Cross Ref
- [5] . 1986. Generalized autoregressive conditional heteroskedasticity. J. Economet. 31, 3 (1986), 307–327.Google Scholar
Cross Ref
- [6] . 2011. Econophysics review: I. empirical facts. Quantit. Fin. 11, 7 (2011), 991–1012.Google Scholar
Cross Ref
- [7] . 1997. Emergence of cooperation and organization in an evolutionary game. Phys. A: Statist. Mech. Applic. 246, 3-4 (1997), 407–418.Google Scholar
Cross Ref
- [8] . 2001. Empirical properties of asset returns: Stylized facts and statistical issues. Quantit. Fin. 1, 2 (2001), 223–236.
DOI: Google ScholarCross Ref
- [9] . 2021. Towards realistic financial time series generationvia generative adversarial learning. In Proceedings of the 29th European Signal Processing Conference (EUSIPCO).Google Scholar
- [10] . 2020. Omni-sourced webly-supervised learning for video recognition. In Computer Vision – ECCV 2020, , , , and (Eds.). Springer International Publishing, Cham, 670–688. Google Scholar
- [11] . 2015. Training generative neural networks via maximum mean discrepancy optimization. In Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence. 258–267.Google Scholar
- [12] . 2018. GANSynth: Adversarial neural audio synthesis. In Proceedings of the International Conference on Learning Representations.Google Scholar
- [13] . 1982. Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Economet.: J. Economet. Societ. 50, 4 (1982), 987–1007.Google Scholar
Cross Ref
- [14] . 2018. A variational U-Net for conditional appearance and shape generation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 8857–8866.Google Scholar
Cross Ref
- [15] . 2019. Enhancing stock movement prediction with adversarial training. In Proceedings of the 28th International Joint Conference on Artificial Intelligence. International Joint Conferences on Artificial Intelligence Organization, 5843–5849.
DOI: Google ScholarCross Ref
- [16] . 2020. Time series simulation by conditional generative adversarial net. Int. J. Mech. Industr. Eng. 14, 6 (2020), 463–476.Google Scholar
- [17] . 2020. Forecasting directional movements of stock prices for intraday trading using LSTM and random forests. arXiv preprint arXiv:2004.10178 (2020).Google Scholar
- [18] . 2018. NiftyNet: A deep-learning platform for medical imaging. Comput. Meth.ods and Progr. Biomed. 158 (2018), 113–122.Google Scholar
Cross Ref
- [19] . 2016. Deep Learning. The MIT Press. Retrieved from http://www.deeplearningbook.org.Google Scholar
Digital Library
- [20] . 2014. Generative adversarial nets. In Proceedings of the Conference on Advances in Neural Information Processing Systems. 2672–2680.Google Scholar
- [21] . 2006. A kernel method for the two-sample-problem. Adv. Neural Inf. Process. Syst. 19 (2006), 513–520.Google Scholar
- [22] . 2012. A kernel two-sample test. J. Mach. Learn. Res. 13, 1 (2012), 723–773.Google Scholar
Digital Library
- [23] . 2020. Predicting the trend of stock market index using the hybrid neural network based on multiple time scale feature learning. Appl. Sci. 10, 11 (2020), 3961.Google Scholar
Cross Ref
- [24] . 2016. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 770–778.Google Scholar
Cross Ref
- [25] . 2017. GANs trained by a two time-scale update rule converge to a local Nash equilibrium. In Proceedings of the Conference on Advances in Neural Information Processing Systems. 6626–6637.Google Scholar
- [26] . 2018. NSE stock market prediction using deep-learning models. Proced. Comput. Sci. 132 (2018), 1351–1362.Google Scholar
Cross Ref
- [27] . 2018. Listening to chaotic whispers: A deep learning framework for news-oriented stock trend prediction. In Proceedings of the 11th ACM International Conference on Web Search and Data Mining. 261–269.Google Scholar
Digital Library
- [28] . 2019. Timeception for complex action recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Google Scholar
Cross Ref
- [29] . 2017. Image-to-image translation with conditional adversarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Google Scholar
Cross Ref
- [30] . 2021. EnlightenGAN: Deep light enhancement without paired supervision. IEEE Trans. Image Process. 30 (2021), 2340–2349.Google Scholar
Digital Library
- [31] . 2016. Statistical industry classification. J. Risk Contr. 3, 1 (2016), 17–65.Google Scholar
- [32] . 1960. Mathematical methods of organizing and planning production. Manag. Sci. 6, 4 (1960), 366–422.Google Scholar
Digital Library
- [33] . 2019. A style-based generator architecture for generative adversarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 4401–4410.Google Scholar
Cross Ref
- [34] . 2019. HATS: A hierarchical graph attention network for stock movement prediction. arXiv preprint arXiv:1908.07999.Google Scholar
- [35] . 2019. Financial series prediction using attention LSTM. arXiv preprint arXiv:1902.10877.Google Scholar
- [36] . 2014. Auto-encoding variational bayes. In Proceedings of the International Conference on Learning Representation (ICLR).Google Scholar
- [37] . 2011. Can machine learning techniques be used to predict market direction?—The 1,000,000 model test. Retrieved on 21 July, 2021 from http://www.jonathankinlay.com/Articles/ONE%20MILLION%20MODELS.pdf.Google Scholar
- [38] . 2021. Generative adversarial networks for financial trading strategies fine-tuning and combination. Quantit. Fin. 21, 5 (2021), 797–813.Google Scholar
Cross Ref
- [39] . 1951. On information and sufficiency. Ann. Math. Statist. 22, 1 (1951), 79–86.Google Scholar
Cross Ref
- [40] . 2019. DP-LSTM: Differential privacy-inspired LSTM for stock prediction using financial news. arXiv preprint arXiv:1912.10806 (2019).Google Scholar
- [41] . 2015. Generative moment matching networks. In Proceedings of the International Conference on Machine Learning. PMLR, 1718–1727.Google Scholar
- [42] . 1991. Divergence measures based on the Shannon entropy. IEEE Trans. Inf. Theor. 37, 1 (1991), 145–151.Google Scholar
Digital Library
- [43] . 2018. Leveraging financial news for stock trend prediction with attention-based recurrent neural network. arXiv preprint arXiv:1811.06173.Google Scholar
- [44] . 2018. Are GANs created equal? A large-scale study. In Proceedings of the Conference on Advances in Neural Information Processing Systems. 700–709.Google Scholar
- [45] . 1999. Scaling and criticality in a stochastic multi-agent model of a financial market. Nature 397, 6719 (1999), 498–500.Google Scholar
Cross Ref
- [46] . 1951. The Kolmogorov-Smirnov test for goodness of fit. J. Amer. Statist. Assoc. 46, 253 (1951), 68–78.Google Scholar
Cross Ref
- [47] . 2018. Spectral normalization for generative adversarial networks. In Proceedings of the International Conference on Learning Representations.Google Scholar
- [48] . 2018. Generalised spatial and spatiotemporal autoregressive conditional heteroscedasticity. Spatial Statist. 26 (2018), 125–145.Google Scholar
Cross Ref
- [49] . 2020. Filterbank design for end-to-end speech separation. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 6364–6368.Google Scholar
Cross Ref
- [50] . 2019. PyTorch: An imperative style, high-performance deep learning library. Adv. Neural Inf. Process. Syst. 32 (2019), 8026–8037.Google Scholar
- [51] . 2020. Overview of the ImageCLEFmed 2020 concept prediction task: Medical image understanding. In CLEF2020 Working Notes(
CEUR Workshop Proceedings , Vol. 2696). CEUR-WS.org.Google Scholar - [52] . 2016. Unsupervised representation learning with deep convolutional generative adversarial networks. In Proceedings of the 4th International Conference on Learning Representations, ICLR, and (Eds.). Retrieved from http://arxiv.org/abs/1511.06434.Google Scholar
- [53] . 2018. Ensemble of technical analysis and machine learning for market trend prediction. In Proceedings of the IEEE Symposium Series on Computational Intelligence (SSCI). IEEE, 2090–2096.Google Scholar
Cross Ref
- [54] . 2014. Stochastic backpropagation and approximate inference in deep generative models. In Proceedings of the International Conference on Machine Learning. PMLR, 1278–1286.Google Scholar
- [55] . 2016. Improved techniques for training GANs. In Proceedings of the 30th International Conference on Neural Information Processing Systems. 2234–2242.Google Scholar
Digital Library
- [56] . 2020. Improving S&P stock prediction with time series stock similarity. arXiv preprint arXiv:2002.05784.Google Scholar
- [57] . 2019. Speaker recognition for multi-speaker conversations using x-vectors. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 5796–5800.Google Scholar
Cross Ref
- [58] . 2019. Modeling financial time-series with generative adversarial networks. Phys. A: Statist. Mech. Applic. 527 (2019), 121261.Google Scholar
Cross Ref
- [59] . 2016. A note on the evaluation of generative models. In Proceedings of the International Conference on Learning Representations (ICLR’16). 1–10.Google Scholar
- [60] . 2012. Forecasting stock performance in Indian market using multinomial logistic regression. J. Bus. Stud. Quart. 3, 3 (2012), 16.Google Scholar
- [61] . 2017. Strategic agent-based modeling of financial markets. RSF: Russell Sage Found. J. Soc. Sci. 3, 1 (2017), 104–119.Google Scholar
- [62] . 2020. Quant GANs: Deep generation of financial time series. Quantit. Fin.ance 20, 9 (2020), 1419–1440.Google Scholar
Cross Ref
- [63] . 2019. Time-series generative adversarial networks. In Proceedings of the Conference on Advances in Neural Information Processing Systems. 5509–5519.Google Scholar
- [64] . 2017. Adversarial feature matching for text generation. In Proceedings of the 34th International Conference on Machine Learning. JMLR. org, 4006–4015.Google Scholar
Digital Library
- [65] . 2020. Pushing the limits of semi-supervised learning for automatic speech recognition. In Neural Information Processing Systems Workshop on Self-supervised Learning for Speech and Audio Processing Workshop.Google Scholar
- [66] . 2018. Stock market prediction on high-frequency data using generative adversarial nets. Math. Prob. Eng. (2018).Google Scholar
Index Terms
Generation of Realistic Synthetic Financial Time-series
Recommendations
Transformation of traditional financial companies into FinTech
DTMIS '20: Proceedings of the International Scientific Conference - Digital Transformation on Manufacturing, Infrastructure and ServiceThe modern trend of the economy is the rapid evolution of the traditional business of financial companies in terms of expanding the list of services provided and changing the algorithm for communicating with customers with the introduction of modern ...
The Impact of Digital Transformation on the Process of Russian Enterprises Entering International Financial Markets
DEFIN-2021: IV International Scientific and Practical ConferenceThis article discusses issues related to the impact of digitalization on various spheres of and economic relations; in particular, financial markets and issues related to the entry of Russian companies into the international arena in the field of ...
Machine Learning vs. Economic Restrictions: Evidence from Stock Return Predictability
This paper shows that investments based on deep learning signals extract profitability from difficult-to-arbitrage stocks and during high limits-to-arbitrage market states. In particular, excluding microcaps, distressed stocks, or episodes of high market ...






Comments