An Application of the Arbitrage Pricing Theory Using Canonical Correlation Analysis

This paper demonstrates an application of the Arbitrage Pricing Theory using canonical analysis as an alternative to the conventional factor analysis. Following the traditional view that asset prices are influenced by unanticipated economic events, the systematic effects of the major composite economic indices on a wide spectrum of industry returns are explored. The main conclusion is that profitability may be considered as the single most important factor that influences security returns. Also, the composite lagging economic indicators appear to be more useful to investors in forming market expectations than the composite leading economic indicators. Finally, it is argued that the composite index of coincident economic indicators do not exhibit any significant influence in the pricing of capital assets.