
Digital advertising activity may provide useful signals for equity investors, but its relationship with subsequent stock returns varies significantly across geographical markets, according to new research from quantitative investment research firm Brain.
The study, “Advertising Activity as an Alternative Data Signal for Equity Investing,” conducted using Oxford Data Plan’s Advertiser Insights dataset, identifies different patterns across US, European and Asia-Pacific equities, suggesting that advertising-derived investment signals require region-specific analysis rather than a uniform global approach.
In the US, the strongest historical relationships were negative, with lower advertising intensity and a narrower range of advertising platforms associated with better subsequent stock performance. European equities exhibited a contrasting pattern, with stronger advertising growth, momentum and unexpected changes in spending generally linked to higher forward returns. Europe also demonstrated the greatest stability in portfolio-level results across different testing periods.Asia-Pacific results were more dependent on the particular advertising measure examined, with growth and platform-breadth indicators producing different relationships with stock performance. The aggregated global universe showed weaker differentiation between equity portfolios, suggesting that combining regional markets can obscure potentially useful signals.
From Advertising Estimates to Investment Signals
The research, undertaken by Brain executive chairman and head of research Matteo Campellone and CEO and CTO Francesco Cricchio, examines whether digital advertising intelligence can be converted into quantitative features suitable for systematic investment research.
Oxford Data Plan’s dataset provides weekly estimates of digital advertising spending, derived from impressions observed through a predominantly US-based panel of consenting users and modelled advertising costs. According to the provider, it covers more than 3,000 listed companies, approximately 60,000 parent companies and 148,000 brands.
Brain mapped the advertising observations to publicly traded companies and developed measures covering spending intensity, growth, acceleration, momentum, unexpected changes in expenditure, and the number of brands and advertising platforms used.These features were tested against subsequent equity returns using point-in-time data, portfolio rankings, statistical correlations and stability analysis. The researchers also examined exposure to established equity factors to assess whether the advertising measures captured information beyond conventional financial characteristics.
The paper includes an illustrative adaptive approach that combines advertising features selected according to their historical effectiveness, although the authors caution that further testing would be required before applying such signals in live portfolios.
Testing the Investment Case
The study provides a practical example of Brain’s expanding work in third-party alternative data validation. In a July interview with Market & Alt Data Insight, Cricchio described how the firm had extended its internal cross-sectional research platform to evaluate external datasets, helping investment managers assess their potential usefulness before committing resources to integration.
The findings of the research, commissioned by Oxford Data Plan, are based on historical backtests rather than live investment performance, and the authors acknowledge that testing numerous potential signals creates a risk of selection bias.
The advertising expenditure figures are themselves modelled estimates rather than directly observed corporate spending. Nevertheless, the findings suggest that changes in advertising behaviour may offer additional information for quantitative investors, particularly when evaluated alongside existing market, fundamental and alternative data signals.
Subscribe to our newsletter


