Backtester engine

Blueprint Finance quant backtester

Introduction

The backtester engine is a key component and is used as a validation tool in the quantitative trading strategies R&D process. It allows visualization and analyze of a quantitative trading strategy, given a set of parameters.

Description

The backtester engine runs trading simulations using historic data over a multi years history.

At each reset (rebalance) date, the new target positions are generated by the portfolio optimizer.

Profit and loss are derived from target position and transactions cost estimates.

Different metrics are calculated in order to evaluate the quant trading strategy behaviour and performance.

Many visualization are generated, including and not limited to: equity lines (at the portfolio level and breakdown by coin), drawdown charts, risk, allocation by coin at each reset date, volatility, volumes, benchmark performance, fraction of Median Daily Volume (MDV) traded, etc.

Example

The backtester engine has the option of generating a simulated fake quant trading strategy by assuming 'perfect forecasts', i.e., the forecasts are always perfect.

This is equivalent (but not equal) to the Bayes error rate in statistics, which is the lowest possible error rate for any classifier/predictor of a random outcome.

It is a good way of testing the pipeline and provides an estimate of the best possible, although not achievable, quant strategy possible.

Example below.