The article explores how using a substitute for the actual market portfolio can affect the calculation of beta in the CAPM model. By comparing different portfolios to the true market portfolio, the researchers found that betas calculated using these substitutes can be significantly different. This suggests that using an inaccurate proxy for the market portfolio can lead to misleading results when estimating beta.