GP algorithm versus hybrid and mixed neural networks

Abstract

In the current paper we present an integrated genetic programming environment, called java GP Modelling. The java GP Modelling environment is an implementation of the steady-state genetic programming algorithm. That algorithm evolves tree based structures that represent models of inputoutput relation of a system. The motivation of this paper is to compare the GP algorithm with neural network architectures when applied to the task of forecasting and trading the ASE 20 Greek Index using only autoregressive terms as inputs. This is done by benchmarking the forecasting performance of the GP algorithm and 6 different ARMA-Neural Network combination designs representing a Hybrid, Mixed Higher Order Neural Network (HONN), a Hybrid, Mixed Recurrent Network (RNN), a Hybrid, Mixed classic Multilayer Perceptron (MLP) with some traditional techniques, either statistical such as a an autoregressive moving average model (ARMA), or technical such as a moving average convergence/divergence model (MACD), plus a naïve trading strategy. More specifically, the trading performance of all models is investigated in a forecast and trading simulation on ASE 20 time series closing prices over the period 2001-2008 using the last one and a half years for out-of-sample testing. We use the ASE 20 daily series as many financial institutions are ready to trade at this level and it is therefore possible to leave orders with a bank for business to be transacted on that basis.As it turns out, the GP model does remarkably well and outperforms all other models in a simple trading simulation exercise. This is also the case when more sophisticated trading strategies using confirmation filters and leverage are applied, as the GP model still produces better results and outperforms all other neural network and traditional statistical models in terms of annualised return.

The role of attribute selection in Deep ANNs learning framework for high‐frequency financial trading

“…The empirical works in the financial literature show a number of ML-based models that are developed for classification and regression tasks to discover hidden patterns in large amounts of HF financial data, which may facilitate the portfolio selection problem, arbitrage opportunities, risk management, and financial forecasting (Aloud, 2017a;Gerlein et al, 2016). In the financial market literature, numerous trading models have been developed, ranging from simple ML-based models (Gerlein et al, 2016) to complex models such as artificial neural networks (ANNs; Holland, 1975;Zimmermann, Neuneier, & Grothmann, 2001a, 2001bLeigh et al, 2002;Dunis, Laws, & Karathanasopoulos, 2013), support vector machines (SVMs; Kim, 2003), genetic algorithms (GAs; Holland, 1975;Bauer, 1994), and genetic programming (GP; Becker & Seshadri, 2003;Aloud, 2017b) owing to their potential for identifying hidden patterns in HF financial time series. The effectiveness of complex models such as ANNs, GP, GAs, and hybrid models (Cai, Hu, & Lin, 2012) have been explored, and they have shown promising results.…”

Designing Strategies for Autonomous Stock Trading Agents using a Random Forest Approach

“…Machine learning techniques identify the patterns in financial time series using a data driven approach. Common machine learning techniques for forecasting financial time series are decision trees [14], random forest (RF) [15][16][17], artificial neural networks (ANNs) [18][19][20][21][22] and evolutionary algorithms, such as genetic algorithms [23][24][25][26][27] and genetic programming (GP) [28][29][30][31][32][33]. Machine learning applications of financial time-series forecasting are reviewed in [34][35].…”

Can High‐frequency Trading Strategies Constantly Beat the Market?

“…The STGP (described in Appendix A) represents sophisticated trading algorithm that successfully replicates real‐life HFT strategies. According to Dunis et al (), genetic programming models perform remarkably well even in simple trading exercises. Moreover, Wah and Wellman () argue that questions about HFT implications are inherently computational in nature because the speed of trading reveals details of internal market activities and the structure of communication channels.…”

“…We also use combinations of forecasting techniques as benchmarks to demonstrate that HFT scalping strategies anticipate the trading order flow and constantly beat the market.HFTrs) and two indices -Russell 1000 (large-cap stocks) and Russell 2000 (small-cap stocks)to determine the level of profit HFTrs generate.The STGP (described in Appendix A) represents sophisticated trading algorithm that successfully replicates real-life HFT strategies. According to Dunis et al (2013), genetic programming models perform remarkably well even in simple trading exercises. Moreover, Wah and Wellman (2013) argue that questions about HFT implications are inherently computational in nature because the speed of trading reveals details of internal market activities and the structure of communication channels.We reproduce the HFT strategies in an artificial stock market environment where the exact level of profitability can be measured.…”