Digital Commons Network^™

Technical Analysis Under Knightian Uncertainty, Andre Mouton

Student Theses and Dissertations

Technical analysis, or the forecasting of asset price movements using past prices, is commonly practiced in financial markets but poorly explained by mainstream economic theory. I show that a technical rule can have predictive power when an asset’s payoffs are subject to Knightian uncertainty, defined as variation that cannot be described probabilistically (Knight, 1921). I present an asset-pricing model in which asset payoffs undergo periodic shifts in trend, and agents form expectations about these payoffs using a constant gain least squares (CGLS) rule. I investigate whether a second CGLS rule, operating on price, can provide a more accurate forecast of …