Insights from Google Insights?

Google Insights for Search enables users to “compare [normalized] search volume patterns across specific regions, categories, time frames and properties.” Might search volume patterns for a stock or fund symbol reveal an investor/trader level of interest that later emerges predictably in price movements? For example, might search volume patterns for “XLF” reliably indicate future price movements for the Financial Select Sector SPDR (XLF)? Using weekly normalized U.S. “XLF” search volumes and weekly prices for XLF over the period 12/30/07 through 3/14/09 (64 weeks), we find that:

The following chart compares the weekly price for XLF with the same-week normalized U.S. “XLF” search volume (search intensity) over the entire sample period. Note that the maximum search intensity over any selected sample period is 100 because of the normalization approach. There is no obvious systematic relationship between the two series.

For a more precise test of the relationship, we look at weekly XLF returns versus same-week search intensity.

The following scatter plot relates weekly XLF returns to same-week U.S. “XLF” search intensity over the entire sample period. The Pearson correlation for these two series is 0.16, and the R-squared statistic is 0.02, indicating that weekly U.S. “XLF” search intensity explains 2% of same-week returns. If we use worldwide “XLF” search intensity rather than U.S. search intensity, the R-squared statistic decreases to 0.01.

Is there any systematic lead-lag relationship between search intensity and returns?

The final chart shows Pearson correlations between weekly XLF returns and weekly “XLF” search intensify for various lead-lag relationships. The generally weak correlations suggest that both a decline in XLF last week an advance in XLF this week has a slight tendency to boost interest in “XLF” searches this week. The boost in search interest might carry over to a rise in XLF next week, but the effect is negligible (an R-squared statistic of 0.00 for U.S. searches and 0.01 for worldwide searches).

In summary, evidence from simple tests does not support a belief that investor/trader interest in stocks and funds as measured by normalized Google search volumes can help predict future returns.

You May Also Enjoy...

• Sentiment from Google Insights and Return Continuation
• Google Search Data as a Measure of Investor Attention
• Option Expiration Week Stock Return Drill-down