Working Paper: CEPR ID: DP15786
Authors: Olivier Dessaint; Thierry Foucault; Laurent Fresard
Abstract: We analyze the effect of alternative data on the informativeness of financial forecasts. Our starting hypothesis is that the emergence of alternative data reduces the cost of obtaining information about firms' short-term cash-flows more than their long-term cash-flows. If correct, and forecasting short-term and long-term cash-flows are distinct tasks, analysts will reduce effort to process long-term information when alternative data become available. Alternative data thus makes long-term forecasts less informative, while increasing the informativeness of short-term forecasts. We confirm this prediction using variations in analysts' exposure to social media data and a new measure of forecast informativeness at various horizons.
Keywords: alternative data; security analysts; forecasting; horizon; forecasts; informativeness; social media
JEL Codes: D84; G14; G17; M41
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Exposure to alternative data (C81) | Informativeness of short-term forecasts (C53) |
| Exposure to alternative data (C81) | Informativeness of long-term forecasts (C53) |
| Decrease in cost of obtaining short-term information (G14) | Informativeness of short-term forecasts (C53) |
| Decrease in cost of obtaining short-term information (G14) | Informativeness of long-term forecasts (C53) |