Statistical arbitrage with vine copulas

Stübinger, Johannes; Mangold, Benedikt; Krauß, Christopher

doi:10.1080/14697688.2018.1438642

Cited by 31 publications

(13 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As expected, the average daily returns after transaction costs amount -0.0010 compared to 0.0018 for OCP. This finding is well in line with Gatev et al (2006) and Stübinger et al (2016).…”

Section: Robustness Checkssupporting

confidence: 88%

Statistical arbitrage with optimal causal paths on high-frequency data of the S&P 500

Stübinger

2018

Quantitative Finance

Self Cite

View full text Add to dashboard Cite

This paper develops the optimal causal path algorithm and applies it within a fully-fledged statistical arbitrage framework to minute-by-minute data of the S&P 500 constituents from 1998 to 2015. Specifically, the algorithm efficiently determines the optimal non-linear mapping and the corresponding lead-lag structure between two time series. Afterwards, this study explores the use of optimal causal paths as a means for identifying promising stock pairs and for generating buy and sell signals. For this purpose, the established trading strategy exploits information about the leading stock to predict future returns of the following stock. The value-add of the proposed framework is assessed by benchmarking it with variants relying on classic similarity measures and a buy-and-hold investment in the S&P 500 index. In the empirical back-testing study, the trading algorithm generates statistically and economically significant returns of 54.98 percent p.a. and an annualized Sharpe ratio of 3.57 after transaction costs. Returns are well superior to the benchmark approaches and do not load on any common sources of systematic risk. The strategy outperforms in the context of cryptocurrencies even in recent times due to the fact that stock returns contain substantial information about the future bitcoin returns.

show abstract

“…As expected, the average daily returns after transaction costs amount -0.0010 compared to 0.0018 for OCP. This finding is well in line with Gatev et al (2006) and Stübinger et al (2016).…”

Section: Robustness Checkssupporting

confidence: 88%

Statistical arbitrage with optimal causal paths on high-frequency data of the S&P 500

Stübinger

2018

Quantitative Finance

Self Cite

View full text Add to dashboard Cite

show abstract

“…To conclude, strategies that rely on tree-based ML methods outperform strategies based on less sophisticated linear approaches (SVM, LIR) and strategies that hardly use any information at all (BET, HOM) which is clearly reflected in the vast majority of return/risk statistics-the latter is especially true for the strategy ALL. Following [45][46][47], we further analyze each strategies performance over time: the cumulative returns of RAF, BOO, SVM, LIR, and ALL are reported in the upper, the remaining three strategies BET, HOM, RAN in the lower graph of Figure 4 (football season 2006/2007 to 2017/2018). The time series are sorted top down in order of their cumulative return at the latest point in time series.…”

Section: Financial Analysismentioning

confidence: 99%

Machine Learning in Football Betting: Prediction of Match Results Based on Player Characteristics

2019

Self Cite

View full text Add to dashboard Cite

In recent times, football (soccer) has aroused an increasing amount of attention across continents and entered unexpected dimensions. In this course, the number of bookmakers, who offer the opportunity to bet on the outcome of football games, expanded enormously, which was further strengthened by the development of the world wide web. In this context, one could generate positive returns over time by betting based on a strategy which successfully identifies overvalued betting odds. Due to the large number of matches around the globe, football matches in particular have great potential for such a betting strategy. This paper utilizes machine learning to forecast the outcome of football games based on match and player attributes. A simulation study which includes all matches of the five greatest European football leagues and the corresponding second leagues between 2006 and 2018 revealed that an ensemble strategy achieves statistically and economically significant returns of 1.58% per match. Furthermore, the combination of different machine learning algorithms could neither be outperformed by the individual machine learning approaches nor by a linear regression model or naive betting strategies, such as always betting on the victory of the home team.

show abstract

“…We transfer the top 10 pairs (Miao 2014, Stübinger et al 2016 exhibiting both a high mean-reversion speed θ and a high volatility σ to the trading period. A larger θ leads to a higher trading frequency, and a larger σ leads to a bigger fluctuation of the process, both resulting in a higher profit in each trade (Zeng and Lee 2014).…”

Section: Formation Periodmentioning

confidence: 99%

“…The second benchmark is given by Zeng and Lee (2014) and represents a classic asymmetric strategy based on Brownian motion-driven OU models (CBM). The strategy takes positions at two-standard deviations and clears positions when the spread reverts back to the mean (Bollinger 1992, Avellaneda and Lee 2010, Clegg and Krauss 2016, Stübinger et al 2016. In stark contrast to the optimal trading rules of OLM and OBM, the entry and exit signals in CBM are model-free.…”

Section: Assessment Of the Developed Strategymentioning

confidence: 99%

Optimal trading strategies for Lévy-driven Ornstein–Uhlenbeck processes

Endres

Stübinger

2019

Applied Economics

View full text Add to dashboard Cite

This study derives an optimal pairs trading strategy based on a Lévy-driven Ornstein-Uhlenbeck process and applies it to high-frequency data of the S&P 500 constituents from 1998 to 2015. Our model provides optimal entry and exit signals by maximizing the expected return expressed in terms of the first-passage time of the spread process. An explicit representation of the strategy's objective function allows for direct optimization without Monte Carlo methods. Categorizing the data sample into 10 economic sectors, we depict both the performance of each sector and the efficiency of the strategy in general. Results from empirical back-testing show strong support for the profitability of the model with returns after transaction costs ranging from 31.90 percent p.a. for the sector "Consumer Staples" to 278.61 percent p.a. for the sector "Financials". We find that the remarkable returns across all economic sectors are strongly driven by model parameters and sector size. Jump intensity decreases over time with strong outliers in times of high market turmoils. The value-add of our Lévy-based model is demonstrated by benchmarking it with quantitative strategies based on Brownian motion-driven processes.

show abstract

Statistical arbitrage with vine copulas

Cited by 31 publications

References 58 publications

Statistical arbitrage with optimal causal paths on high-frequency data of the S&P 500

Statistical arbitrage with optimal causal paths on high-frequency data of the S&P 500

Machine Learning in Football Betting: Prediction of Match Results Based on Player Characteristics

Optimal trading strategies for Lévy-driven Ornstein–Uhlenbeck processes

Contact Info

Product

Resources

About