Backtesting

Back testing and rear refers to the comparison process, a strategy theory or model to evaluate by the strategy and the theory and the model is applied on historical data. Typical applications for backtesting are for example the examination of issues:

• What results would have brought a trading strategy in the past on a stock market?
• How well matched the predictions of climate models with the actual changes in weather together?

A major difference between backtesting and other historical test is that is calculated for back-testing as a strategy would have behaved if they had actually been carried out. Therefore, it is necessary to replicate properly the relevant historical conditions for proper backtesting result. Backtesting is a common and methodically recognized approach for scientific research. However, a high correlation between successful or a verified strategy and historical results no proof of the truth of a theory. This is because you can not necessarily indicative of future events from past events. Assuming, however, that future events have a great similarity to past events, backtesting is a useful tool for analysis and forecasting. Backtesting is used primarily in the social and natural sciences. This is because that processes provide measurable data in this area, through a relatively long period of time and the data are most suitable for statistical analysis.

Backtesting in finance and economics

Backtesting trading strategies in

New investment and trading strategies on capital or equity markets will be tested by observing what results they have delivered in the past, if they had been actually applied. In this type of backtesting is trying to learn from the " mistakes of the past ", without losing money. Since real data serve as a basis, the weakness of a strategy is hoped that better seen than in the use of synthetic data. Louis B. Mendelsohn used backtesting in 1983 as first in a commercial software for the PC. The software allows trAAAde an automated backtesting with iteration of the simulation parameters.

Back-testing for quality assurance in risk management

Backtesting is used in the risk management of banks in order to check the quality of the risk measure Value at Risk. This is compared to the number of times the predicted loss is exceeded. In many so-called outliers, the value at risk model must be modified.

Also in rating models banks must conduct regular evaluations. Here, the back-testing of default probabilities plays an important role. It is checking that the predicted number of failures that actually occurred with the number of failures matches. In practice, the binomial test is used for backtesting often.

Backtesting in climate models

Back-testing plays a major role in the evaluation of weather models and climate models. It is tested whether new models can also reproduce the historical weather and climate with historical input data.