Threshold-based portfolio: the role of the threshold and its applications
- PDF / 1,451,103 Bytes
- 18 Pages / 439.37 x 666.142 pts Page_size
- 19 Downloads / 236 Views
Threshold-based portfolio: the role of the threshold and its applications Sang Il Lee1
· Seong Joon Yoo1
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract This paper aims at developing a new method by which to build a data-driven portfolio featuring a target risk–return. We first present a comparative study of recurrent neural network models (RNNs), including a simple RNN, long short-term memory (LSTM), and gated recurrent unit. The models are applied to the investment universe consisted of 10 stocks in the S&P500. The experimental results show that the LSTMbased prediction model outperforms the others in terms of hit ratio of 1-month-ahead forecasts. We then build predictive threshold-based portfolios (TBPs) that are subsets of the universe satisfying given threshold criteria for the LSTM-based return forecasts. The TBPs are rebalanced monthly to restore equal weight to the constituents of the TBPs. We find that the risk and return profile of the realized TBP represents a monotonically increasing frontier on the risk–return plane, where the equally weighted universe portfolio plays a role in the lower bound of TBPs. This shows the availability of TBPs in targeting specific risk–return levels, and the EWP of an universe plays a role in the reference portfolio of the TBPs. In the process, thresholds play dominant roles in characterizing risk, return, and the prediction accuracy of the TBPs. The TBP is more data-driven in designing portfolio return and risk than existing ones, in the sense that it requires no prior knowledge of finance such as financial assumptions, financial mathematics, or expert insights. For practical uses, we present a multiperiod TBP management method and also discuss the application of TBP to mean–variance portfolios to reduce estimation risk. Keywords Portfolio management · Recurrent neural networks · Efficient frontier · Financial time series
B
Seong Joon Yoo [email protected] Sang Il Lee [email protected]; [email protected]
1
Department of Computer Engineering, Sejong University, Seoul 05006, Republic of Korea
123
S. I. Lee, S. J. Yoo
1 Introduction Today, machine learning has come to play an integral role in many parts of the financial ecosystem, from portfolio management and algorithmic trading, to fraud detection and loan/insurance underwriting. Time series are one of the most common data types encountered in finance, and so time-series analysis based on econometrics has been widely used in finance and economics. The development of machine learning algorithms has opened a new vista for modeling the complexity of financial time series as an alternative to the traditional econometric models, by effectively combining diverse data and capturing nonlinear behavior. For this reason, financial time-series modeling has been one of the most interesting topics that has arisen in the application of machine learning to finance. Researchers have successfully modeled financial time series by focusing primarily on prediction accuracy or automatic trading rules [1
Data Loading...
