Do analysts predict market returns?

Fabian Scheler

July 6, 2023

Data providers like Bloomberg aggregate the target prices published by sell-side analysts to derive a marketwide consensus price target. Does the wisdom of this crowd tell us something about future returns?

5 Minutes Read

Sell-side analysts working for investment banks routinely issue forward-looking price targets for the stocks they cover, which can be summed and weighed for broad market indices. The difference between the contemporary index level and these price targets can be considered the implied expected price return. We have asked ourselves whether the return outlook given by analysts on aggregate (consensus estimate) does indeed provide information about future market performance. The answer is simple and disappointing but yields some intriguing insights into the sell-side analyst’s reaction function.

A simple linear regression model using past price performance, earnings yield, dividend yield, and benchmark interest rates as independent variables explains most of the variation in marketwide consensus price targets (R* of 0.80 for the S&P 500 and even 0.84 for the SMI). While sell-side analysts show some sensitivity to valuation levels, reducing return expectations when earnings multiples are high, realized performance over the most recent semester is the most critical driver of implied expected returns, given considerable inertia in analysts' estimates.
Key Insights

• Sell-side analysts from investment banks continuously follow a set of stocks and regularly publish recommendations and price targets founded on their fundamental and quantitative research.

• These target prices are aggregated for indices like the S&P 500 by financial data providers such as Bloomberg to derive a marketwide implied expected return, also dubbed the 'Return Potential'.

• We have tested this metric's predictive power for several markets and found no correlation between its level and realized price performance over the subsequent one to 36 months.

• What's more, a simple linear regression model almost perfectly explains the spread between current prices and consensus price targets as a function of past performance, valuation levels, and interest rates.

Access the paperGreifen Sie auf das Papier zuAccéder au papier