Broadband internet and the stock market investments of individual investors

In this paper, we study the effects of the roll-out of broadband internet on the portfo- lio choices of individual investors in Norway in the 2000s. [...] For each transaction made by an individual investor, the data contain the anonymized ID of the individual, the date of transaction, the ticker of the security, and the number of shares purchased or sold. [...] In what follows, we report estimates of the effect of broadband coverage on stock market participation (δ from the reduced form model), the effect of broadband coverage on broadband internet use (φ from the first stage model), and the effect of broadband internet use on stock market participation (ω from the IV model). [...] Effect heterogeneity Following Imbens and Angrist (1994), we interpret the main IV estimates in Table II as the effect of broadband internet use on stock market participation for the part of the population that is induced to use broadband internet as a direct result of the broadband reform.12 It is possible that the complier group consists mostly of young males, as in Barber and Odean (2002) and C. [...] To assess the robustness of this result, in Panel B of Figure 3 we re-estimate the ventile-specific δ’s using the log of the mean number of stock trades as the outcome, and 14In practice, all of the ventile-specific effects are estimated in the same regression model by pooling all the ventile-municipality-year observations and interacting all the terms in equation (2) with ventile fixed effects.