Who Says There is a High Consensus Among Analysts When Market Uncertainty is High? Some New Evidence from the Commercial Real Estate Market

This paper seeks to determine whether analyst consensus is a function of the level of price informativeness, and therefore attempts to gauge the extent to which analyst consensus is high when market uncertainty is high. There is the view that a small or declining volume of trading implies less information and lower quality information, and that a lower level of price informativeness will lead analysts to put more weight on their private information at the time of forecasting. We add to this view the notion that when uncertainty is high there might be a positive option value of waiting. In this case, it appears reasonable to conclude that analysts may actually lack consensus when uncertainty is high. If this hypothesis is correct, then one would expect the level of price informativeness to override the Keynesian incentive "to fail conventionally than to succeed unconventionally" or at least reduce its magnitude in periods of declining trading volume and low price discovery. Tests of this hypothesis produce favorable results. The tests are conducted using forecasts of market rent changes and expected returns on investment, all for office buildings. Focus in this paper is on out-of-sample forecast results for 18 quarters from 2007Q1 to 2011Q2. In the out-of-sample period, transaction volume and price informativeness plummeted as the demand for commercial office buildings dropped off. At the same time, analyst consensus is much lower than predicted from comparisons with a vector error correction model fitted to data before the advent of the financial crisis of 2007-2009. We also find that the level of consensus is much higher among public analysts as compared with private analysts between 2009Q2 and 2010Q2. The results can be rationalized on the basis that amid heavier trading in the public market (owing to greater liquidity and lower transaction costs) and lackluster trading in the private market consensus should be higher in the public market than in the private market. We also examine whether the high level of consensus implies greater forecasting accuracy