DECOMPOSITION-BASED APPROACHES FOR FORECASTING STOCK EXCHANGE INDICES

Forecasting stock exchange index values is a fundamental pursuit in financial analysis, traditionally achieved by modeling past index values. However, an alternative approach involves separately forecasting prices for each individual stock comprising the index and then aggregating these forecasts to predict the index value, considering corresponding weights. This study explores the efficacy and utility of such separate forecasting methods in comparison to conventional approaches. Drawing upon the concept of market efficiency posited by Fama (1970), which suggests that direct forecasting may be futile, let alone the indirect method of separate forecasting, there arises a critical inquiry into the rationale and practicality of adopting such techniques. This paper delves into the implications of both methods, considering their theoretical underpinnings and empirical performance.t.