ForeStock plugin uses its own forecasting engine to produce fast and reliable forecasts. Functionality of this inner engine is exposed through
external EEMetaSt.dll dynamic link library module. Data from Metastock(R) are conveyed through this module to the ForeStock inner solver engine, processed there and returned back as trading advises and backtesting reviews which you finally watch on screen.
To access all functionality of ForeStock engine only two universal functions are used:
AuraEngine("AlgorithmName");
AuraEngineEx("AlgorithmName");
"AlgorithmName" is one of predefined strings meaning the algorithm name to be used in engine calls. For now it can be
ARIMA
CLOSESTPATTERNTREND
HISTRYPROPHET
NAIVEPREDICTOR
STEPWISEBESTREGRESSIONMVAR
UNIVARIATELINEARREGRESSION
Other algorithms are going to be added soon.
Here is simple example on how formula can be used to formulate trading strategy:
ExtFml( "EEMetaSt.AuraEngine", ARIMA ) > CLOSE
This formula is used for buy signal using ARIMA forecasting algorithm from ForeStock. It has very simple meaning. AuraEngine function returns estimate of forecast for the next bar. Thus we issue buy signal if the forecast for the next period is higher than the current stock price.
In very same way functions can be used in other formulas.
Function AuraEngineEx makes exactly same as AuraEngine but ensures guaranteed forecast backhistory by recalculating whole model on each backhistory point. This provides extreme reliability of ForeStock backtesting but also slows down calculation several hundred times. We advise to use AuraEngineEx only on system tester to find real performance and use AuraEngine on any regular trading tasks.