yarro-s/stratotester

Strategy Tester is a C++ header-only library for testing trading and investing strategies.

StratoTester

Strategy Tester is a C++ header-only library for testing trading and investing strategies.

Quick Start

Include StratoTester.hpp as follows

#define STRATOTESTER_IMPL  // specified only once 
#include <StratoTester.hpp>

using namespace st;  // for convenience 

Here are some QQQ monthly quotes, starting from Jan 01, 2019 ending Dec 01, 2019

auto qqq_hist = st::prices {168.16, 173.19, 179.66, 189.54, 173.95, 186.74,
                            191.10, 187.47, 188.81, 197.08, 205.10, 212.61};

Let's implement a simple momentum trading rule:

auto last_close_up = [=](prices const &price_hist) {
    auto const last_close = price_hist.back();
    auto const first_close = price_hist.front();

    // buy when last close is higer than the first, hold otherwise
    return last_close > first_close ? 1.0 : 0.0;
};

A trading rule takes a rolling price history (represented internally as an std::vector<double>) and calculates weight in the range from -1.0 to 1.0. For instance, returning 0.5 means that the currently required share of the asset in the portfolio is 50% while returning -0.2 means that 20% of the portfolio is used for shorting the asset.

Next, let's create a strategy based on that rule

auto strat = strategy::with(last_close_up)
                .look_back(3)         // 3 samples rolling window
                .rebalance_every(2);  // rebalanace every 2 samples

We specified a three-sample rolling window with look_back(3). Asset rebalancing, specified as rebalance_every(2), is done every two samples. The samples here hold the monthly closing prices of QQQ for the year 2019.

Let's run this strategy with an initial deposit of $10000

auto init_deposit = 10000;
auto test = single_asset(strat, init_deposit).run(qqq_hist);

and log some of the results

std::cout << "Single asset portfolio value history: " << std::endl
          << str_rep(test.results().value_history())
          << std::endl;
std::cout << "Single asset portfolio total return: "
          << test.results().total_return() << std::endl;

After requesting the backtesting results with results(), you can access all the historical values of the portfolio with value_history() and calculate its total return with total_return().