Development of a polynomial model of algorithmic trading strategies for prop traders: an economic security policy framework

Abstract

The object of this research is the economic security policy controlling algorithmic trading strategies for prop traders based on Polynomial Moving Regression Bands (MRB). The problem addressed is the absence of a functional conceptual framework combining polynomial regression with prop trading strategy infrastructure under conditions of high cryptocurrency market volatility. The study establishes that polynomial models, specifically Legendre, Chebyshev, Laguerre, Hermite and their hybrid combinations are effective tools for smoothing financial time series, identifying nonlinear trends and generating trading signals. A Polynomial Autoregressive (PAR) model was applied to intraday price data of four major cryptocurrencies (Bitcoin, Ethereum, Cardano and a fourth asset). Six-month live trading tests revealed that the buy-and-hold approach outperformed the proposed automated system for each of the four cryptocurrencies examined, with the best relative PAR model performance recorded for Ethereum and Bitcoin and the weakest for Cardano. The findings are recommended for practical application in prop trading firms as a basis for building automated trading systems with integrated risk management mechanisms and as a tool for hyperparameter optimization of trading strategies including through Bayesian optimization under volatile cryptocurrency market conditions.

Source: semanticscholar · PDF

Read the AI summary, key takeaways and discussion on WOBR Quant Research.