In Search of Attention

The authors address a core challenge in behavioral finance: measuring investor attention. Traditional proxies—like trading volume, media coverage, or extreme returns—are indirect and may not accurately reflect the attention investors pay to specific stocks. In a digital age overwhelmed by information, the real scarcity lies in human attention, not data. Recognizing this gap, the paper proposes using Google search frequency, specifically the Search Volume Index (SVI), as a direct, real-time, and objective indicator of investors’ active interest in individual stocks.

The study constructs weekly SVI data for each stock in the Russell 3000 index (from January 2004 to mid-2008). This data is then compared with traditional attention proxies to assess its accuracy and timeliness. The researchers also examine the relationship between SVI changes and trading activity by individual investors, using retail order execution reports under SEC Rule 11Ac1-5. Furthermore, they analyze how shifts in SVI relate to short-term price pressure, IPO performance, and momentum effects.

Key findings reveal that SVI is correlated with—but distinct from—existing attention proxies, and it provides faster and more accurate insights into investor interest. Notably, increased SVI predicts a stock’s price rise over the following two weeks, often followed by a price reversal within the year. Regarding IPOs, heightened SVI during the offering period is linked to large first-day returns coupled with long-term underperformance. In addition, they detect strong linkages between SVI spikes and trading volume by retail investors, showing that heightened search activity coincides with increased attention and trading by less sophisticated participants.

The authors interpret these patterns as evidence that active investor attention—captured through search behaviors—can drive significant short-term price dynamics, such as temporary overpricing and momentum effects. The findings align with behavioral models like Daniel–Hirshleifer–Subrahmanyam (1998), where investor overreaction to private signals can generate momentum, and with models like Hong–Stein (1999), where information diffuses slowly when attention is limited. Essentially, attention itself becomes a force shaping asset price behavior.

This paper makes three pivotal contributions:

1. It introduces a novel, real-time, and direct measure of investor attention using Google search data (SVI), filling a major gap in behavioral finance research.
2. It provides empirical confirmation that attention spikes predict short-term price pressure and price reversals, especially in IPO contexts.
3. It establishes a causal link between retail investor attention (as measured by SVI) and trading behavior, reinforcing the role of individual investors in market dynamics.

Overall, by unveiling how search behavior influences stock prices, this study opens new pathways for incorporating attention metrics into asset pricing models, trading strategies, and regulatory insights.

Listen via Podcast