Abstract: A large part of economies around the world rely on stock markets. To predict stock prices or commodities, econometric techniques are used. Analysts choose the suitable econometric technique according to error measures which may create confusion, especially in the case where there is no preference amongst error measures. Therefore, there is no unique score to rank econometric techniques based on multiple error measures. To bridge this gap, we propose, a MCDM like methodology [geometrical Benefit of Doubt (BoD)] that considers econometric techniques as alternatives and error measures as criteria. A real application with 194 econometric techniques and 8 error measures is presented. The efficiency scores derived from geometrical BoD model provide better discrimination and experiemental results indicate that ARMA and ARCH models are ranked higher. The proposed geometrical BoD model is compared to similar geometrical MCDM formulations.