Given the vast amount of information available, investors must filter source data ruthlessly. In their May 2005 paper entitled “Investor Attention, Overconfidence and Category Learning”, Lin Peng and Wei Xiong model investor allocation of attention to markets/sectors and stocks and examine how attention allocation process affects asset prices. Emphasizing that attention is a scarce cognitive resource, they find that:
- Investors generally tend to allocate a disproportionate level of attention to market-level and sector-level factors at the expense of company-specific data. Under severe overload, they allocate all attention to market and sector information and ignore company data.
- This investor attention allocation tendency causes co-movement of stock prices within sectors and across markets to a greater degree than justified by individual company fundamentals.
- Advances in information technology have allowed investors to devote an increasing proportion of attention to company-specific data in recent years, somewhat reducing unwarranted co-movement of stock prices.
- A news-driven increase in investor attention to company-specific information reduces co-movement of the stock with its sector and the market and increases the predictability of price overreaction (short-term price momentum and long-term price reversal).
In summary, company-specific news tends to decouple stock price behavior from the market, while the absence of news promotes co-movement.
