Questions tagged [back-testing]

^{( in Quantitative Finance, AI, ML, Technical Cybernetics, Modelling ... )}

Back-testing

is a systematic approach to put a System-under-Test [ SuT ] into a state, where historical data ( as a known part of both the evolution of inputs and the eco-system responses ) are used & fed into a SuT so as to review it's behaviour in-vitro ( as opposed to a forward-testing ).

Strengths & Weaknesses

The concept of back-testing relies on an a-priori belief ( a proof of which is left on the reader ) that the SuT does not intervene with the environment's generalised dynamics.

In other words, back-testing assumes that the SuT does not influence the future evolution moves ( the very steps that were recorded in the Historical Data, that are being fed into the SuT during the flow of back-testing ) that the outer ecosystem, surrounding the SuT, actually undertook in the past.

Tools

Strategy Tester, Quantopian, Quant Modeller, Quant Strat, AmiBroker, TradingView, MultiCharts
_{and many more}

Historical Data

Quandl, Olsen Data, Dukas Copy, Yahoo! Finance, Google
_{and many more}