Genetic Algorithms in Quantitative Trading

What Is a Genetic Algorithm?

• Genetic algorithms use the concept of natural selection to find the optimal solution to a problem.
• They are typically used as optimizers, tuning parameters to maximize an objective.
• They can be used standalone or as part of building artificial neural networks.

For example, a trading rule might use indicators such as MACD and RSI. A genetic algorithm feeds values into these parameters with the goal of maximizing profit. Over time, mutations occur, and beneficial mutations are carried over to the next generation.

There are three types of genetic operations:

1. Crossover represents biological reproduction and recombination — offspring inherit certain traits from their parents
2. Mutation represents biological mutation, maintaining genetic diversity from one generation to the next by introducing small random changes.
3. Selection is the stage where individual genomes are chosen from the population for later breeding (recombination or crossover).

Implementation Steps for Genetic Algorithms

1. Initialize a random population where each chromosome has length n, where n is the number of parameters. In other words, create a random set of parameters, each with n elements.
2. Select the chromosomes or parameters that improve the desired outcome (typically net profit).
3. Apply mutation or crossover operators to the selected parents to generate offspring.
4. Recombine the offspring with the current population using a selection operator to form a new population.
5. Repeat steps 2 through 4

Python Libraries for Genetic Algorithms

You can use the geneticalgorithm Python library to apply genetic algorithms in quant strategies.

1pip install geneticalgorithm

123456789101112import numpy as np
from geneticalgorithm import geneticalgorithm as ga

def f(X):
    # X为因子。在本方法中，根据因子寻找股票，计算收益率并返回

varbound=np.array([[0,10]]*3)

model=ga(function=f,dimension=3,variable_type='real',
        variable_boundaries=varbound)

model.run()

Genetic algorithms are a type of optimization algorithm — if anything is unclear, it helps to review general optimization concepts.