Research Papers to Read on Computational Finance

Share:
Photo by micheile dot com on Unsplash
  1. Deep Stochastic Optimization in Finance(arXiv)

Author : A. Max Reppen, H. Mete Soner, Valentin Tissot-Daguette

Abstract : This paper outlines, and through stylized examples evaluates a novel and highly effective computational technique in quantitative finance. Empirical Risk Minimization (ERM) and neural networks are key to this approach. Powerful open source optimization libraries allow for efficient implementations of this algorithm making it viable in high-dimensional structures. The free-boundary problems related to American and Bermudan options showcase both the power and the potential difficulties that specific applications may face. The impact of the size of the training data is studied in a simplified Merton type problem. The classical option hedging problem exemplifies the need of market generators or large number of simulations.

2.Some applications of TDA on financial markets(arXiv)

Author : Miguel Angel Ruiz-Ortiz, José Carlos Gómez-Larrañaga, Jesús Rodríguez-Viorato

Abstract : The Topological Data Analysis (TDA) has had many applications. However, financial markets has been studied slightly through TDA. Here we present a quick review of some recent applications of TDA on financial markets and propose a new turbulence index based on persistent homology — the fundamental tool for TDA — that seems to capture critical transi- tions on financial data, based on our experiment with SP500 data before 2020 stock market crash in February 20, 2020, due to the COVID-19 pan- demic. We review applications in the early detection of turbulence periods in financial markets and how TDA can help to get new insights while in- vesting and obtain superior risk-adjusted returns compared with investing strategies using classical turbulence indices as VIX and the Chow’s index based on the Mahalanobis distance. Furthermore, we include an intro- duction to persistent homology so the reader could be able to understand this paper without knowing TDA.

3.Time-Varying Causality Between Bitcoin and Attention to COVID-19 News: Cultural Grouping(arXiv)

Author : Huaxin Wang-Lu

Abstract : The pandemic and people’s concern over it are associated with the Bitcoin market, while the extent of individualism differentiates how individuals regard and react to the viral spread and corresponding measures. This paper examines if real-time attention to country-specific COVID-19 news Granger causes daily Bitcoin returns and trading volumes between February 13, 2020 and April 04, 2022, and whether the causal relationship varies time-wise between the collectivistic and individualistic country group. Results show different timing and spans of the causality. In general, attention to COVID-19 news of the individualistic cluster presents stronger evidence of causal effects on both Bitcoin returns and trading volumes.

4.Neural Optimal Stopping Boundary(arXiv)

Author : A. Max Reppen, H. Mete Soner, Valentin Tissot-Daguette

Abstract : A method based on deep artificial neural networks and empirical risk minimization is developed to calculate the boundary separating the stopping and continuation regions in optimal stopping. The algorithm parameterizes the stopping boundary as the graph of a function and introduces relaxed stopping rules based on fuzzy boundaries to facilitate efficient optimization. Several financial instruments, some in high dimensions, are analyzed through this method, demonstrating its effectiveness. The existence of the stopping boundary is also proved under natural structural assumptions.

5.Forecasting foreign exchange rates with regression networks tuned by Bayesian optimization(arXiv)

Author : Linwei Li, Paul-Amaury Matt, Christian Heumann

Abstract : The article is concerned with the problem of multi-step financial time series forecasting of For- eign Exchange (FX) rates. To address this problem, we introduce a parameter-free regression network termed RegPred Net. The exchange rate to forecast is treated as a stochastic process. It is assumed to follow a generalization of Brownian motion and the mean-reverting process referred to as generalized Ornstein-Uhlenbeck (OU) process, with time-dependent coefficients. Using past observed values of the input time series, these coefficients can be regressed online by the cells of the first half of the network (Reg). The regressed coefficients depend only on — but are very sensitive to — a small number of hyperparameters required to be set by a global optimization procedure for which, Bayesian optimization is an adequate heuristic. Thanks to its multi-layered architecture, the second half of the regression network (Pred) can project time-dependent values for the OU process coefficients and generate realistic trajectories of the time series. Predictions can be easily derived in the form of expected values estimated by averaging values obtained by Monte Carlo simulation. The forecasting accuracy on a 100 days horizon is evaluated for sev- eral of the most important FX rates such as EUR/USD, EUR/CNY and EUR/GBP. Our experi- mental results show that the RegPred Net significantly outperforms ARMA, ARIMA, LSTMs, and Autoencoder-LSTM models in terms of metrics measuring the absolute error (RMSE) and correlation between predicted and actual values (Pearson’s R, R-squared, MDA). Compared to black-box deep learning models such as LSTM, RegPred Net has better interpretability, simpler structure, and fewer parameters. In addition, it can predict dynamic parameters that reflect trends in exchange rates over time, which provides decision-makers with important information when dealing with sequential decision-making tasks.