Gradient boosting for quantitative finance

Davis, Jesse; Devos, Laurens; Reyners, Sofie

doi:10.21314/jcf.2020.403

Cited by 5 publications

(3 citation statements)

References 25 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Any other regressor can be added to our initial list. We chose them because they have already been successfully applied in the literature for return predictions [34][35][36].…”

Section: The Proposed Ensemble For Stock Return Predictionmentioning

confidence: 99%

Leveraging Return Prediction Approaches for Improved Value-at-Risk Estimation

Bagheri,

Reforgiato Recupero,

Sirnes

2023

Data

View full text Add to dashboard Cite

Value at risk is a statistic used to anticipate the largest possible losses over a specific time frame and within some level of confidence, usually 95% or 99%. For risk management and regulators, it offers a solution for trustworthy quantitative risk management tools. VaR has become the most widely used and accepted indicator of downside risk. Today, commercial banks and financial institutions utilize it as a tool to estimate the size and probability of upcoming losses in portfolios and, as a result, to estimate and manage the degree of risk exposure. The goal is to obtain the average number of VaR “failures” or “breaches” (losses that are more than the VaR) as near to the target rate as possible. It is also desired that the losses be evenly distributed as possible. VaR can be modeled in a variety of ways. The simplest method is to estimate volatility based on prior returns according to the assumption that volatility is constant. Otherwise, the volatility process can be modeled using the GARCH model. Machine learning techniques have been used in recent years to carry out stock market forecasts based on historical time series. A machine learning system is often trained on an in-sample dataset, where it can adjust and improve specific hyperparameters in accordance with the underlying metric. The trained model is tested on an out-of-sample dataset. We compared the baselines for the VaR estimation of a day (d) according to different metrics (i) to their respective variants that included stock return forecast information of d and stock return data of the days before d and (ii) to a GARCH model that included return prediction information of d and stock return data of the days before d. Various strategies such as ARIMA and a proposed ensemble of regressors have been employed to predict stock returns. We observed that the versions of the univariate techniques and GARCH integrated with return predictions outperformed the baselines in four different marketplaces.

show abstract

“…Any other regressor can be added to our initial list. We chose them because they have already been successfully applied in the literature for return predictions [34][35][36].…”

Section: The Proposed Ensemble For Stock Return Predictionmentioning

confidence: 99%

Leveraging Return Prediction Approaches for Improved Value-at-Risk Estimation

Bagheri,

Reforgiato Recupero,

Sirnes

2023

Data

View full text Add to dashboard Cite

show abstract

“…Tree ensembles such as gradient boosted trees and random forests are a powerful model class. They are frequently used in practice because they can tackle a variety of important real-world problems in diverse areas such as medicine (Ghiasi and Zendehboudi 2021), finance (Davis et al 2020) and sports analytics (Decroos et al 2019). Moreover, there are many easy to use and performant public implementations such as XGBoost (Chen and Guestrin 2016), LightGBM (Ke et al 2017), CatBoost (Prokhorenkova et al 2018) and Bit-Boost (Devos, Meert, and Davis 2020).…”

Section: Introductionmentioning

confidence: 99%

Robustness Verification of Multi-Class Tree Ensembles

Devos,

Cascioli,

Davis

2024

AAAI

View full text Add to dashboard Cite

Tree ensembles are one of the most widely used model classes. However, these models are susceptible to adversarial examples, which are slightly perturbed examples that elicit a misprediction. There has been significant research on designing approaches to verify the robustness of tree ensembles to such attacks. However, existing verification algorithms for tree ensembles are only able to analyze binary classifiers and hence address multiclass problems by reducing them to binary ones using a one-versus-other strategy. In this paper, we show that naively applying this strategy can yield incorrect results in certain situations. We address this shortcoming by proposing a novel approximate heuristic approach to verification for multiclass tree ensembles. Our approach is based on a novel generalization of the verification task, which we show emits other relevant verification queries.

show abstract

“…In this context, the market size of quantitative investment funds has been increasing, and the total share of the overall financial market has been expanding [5]. At the same time, with the resurgence of AI technology in recent years, concepts such as "deep learning," "big data," and "data mining" have emerged, and various AI-based data processing technologies, financial forecasting models, and high-performance computers have also become more and more closely integrated with quantitative investment [6][7][8]. The concepts and body of knowledge related to it have gradually become the focus of the investment industry and scholars [9][10].…”

Section: Introductionmentioning

confidence: 99%

Fusion of multifactor modeling and supervised learning algorithms in quantitative finance: a comparative analysis of predictive and explanatory power

2024

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

Finance serves as the lifeblood of the real economy’s development, fundamentally aimed at bolstering the real economy and mitigating financial risks. This paper investigates the application of machine learning technologies in finance, organizing a detailed classification and analysis of associated risks within the sector. Specifically, the study employs the Conditional Value at Risk (CoVaR) to quantify systemic financial risks. It introduces both the Fama-French three-factor and five-factor models for constructing multifactor financial models. It assesses the significance and regression coefficients of risk factors within these models to provide a comparative analysis. Moreover, this research develops an early warning model for financial risk by incorporating financial risk measurement indices into the XGBoost machine learning framework. It utilizes the SHAP (Shapley Additive exPlanations) explanatory framework to elucidate the key features influencing financial risk. This comprehensive approach not only enhances the understanding of financial risk dynamics but also advances the predictive capabilities of financial risk management. It was found that the modified R² increased from 36.52% to 56.25% after adding the IMU factor to the three-factor model, and the coefficient of SMB decreased from 0.135 to −0.215 the larger the size of the enterprise. The accuracy of the financial risk early warning model was 96.35%, and the probability of being predicted by the model to be a high-risk sample was lower when the value of the characteristic INDI was taken as [0,0.125]. Financial institutions can enhance their risk prevention ability by combining the financial risk factors obtained from the multifactor model with the supervised learning algorithm.

show abstract

Gradient boosting for quantitative finance

Cited by 5 publications

References 25 publications

Leveraging Return Prediction Approaches for Improved Value-at-Risk Estimation

Leveraging Return Prediction Approaches for Improved Value-at-Risk Estimation

Robustness Verification of Multi-Class Tree Ensembles

Fusion of multifactor modeling and supervised learning algorithms in quantitative finance: a comparative analysis of predictive and explanatory power

Contact Info

Product

Resources

About