The votes of legislators who have a vested interest in an industry that is affected by a particular bill may suggest whether the bill will be positive or negative for the affected industry. The authors argue that this information, even though it is publicly available, is largely ignored by market participants until the bill is passed. They study the stock performance of affected industries as well as their longer-term earnings revisions and sales growth after the passage of relevant bills, and they find statistically significant results that confirm that voting by interested legislators is a reliable predictor of the impact of a bill on stock price returns as well as on real fundamentals of the industry or firms affected by the bill.

The authors consider the information content in the voting records of legislators, which they claim remains unnoticed by the market at large until after a bill is passed. Focusing on the economic incentives of these legislators as expressed through their votes is very informative in determining the likely impact of legislation on industries and firms. The authors provide evidence by creating a long–short portfolio strategy based on the votes of interested legislators. Going long (short) industries on which the interested legislators vote positively (negatively) produces statistically significant abnormal returns.

The abnormal returns are not realized if the stocks are bought or sold prior to passage of the bill—that is, if there is no run-up in stock prices in anticipation of its passage. In fact, the stock prices have a delayed response to bill passage: They continue to rise or fall for approximately 60 trading days after the bill is passed. This drift in prices and the accompanying abnormal returns are more pronounced in cases of complex bills, which are harder for the market to decipher, and statistically insignificant for routine or noncomplex bills. The authors find that these stock price changes do not reverse, implying that they signal changes in real fundamentals of the affected industries.

The industries for which their methodology assigns bills as positive (negative) show longer-term positive (negative) changes in their fundamentals as well, which the authors measure in terms of changes in earnings per share (EPS) estimates, earnings surprises, and long-term profitability and sales growth.

The authors observe the raw voting records (including “yes” votes, “no” votes, and abstentions) of legislators on all congressional bills for a period of 20 years (1989–2008). They analyze the text of each bill and assign the bills keywords—1 of the 49 Fama–French industry classifications. This procedure establishes which industry or industries are affected by each bill. The authors then determine “interested” senators as those for whom these industries rank among the top three—in terms of economic activity as defined by aggregate sales—in their home states. The authors propose that interested legislators’ voting behavior will be different from that of indifferent senators with no direct stake in the industry.

Once the interested senators are isolated, the authors use the senators’ voting records to calculate a measure of “economic interest.” The authors base their analysis of whether a bill will affect an industry positively or negatively on the economic interest ratio.

They call this procedure of assigning a positive or negative impact “signing” a bill and try to prove the predictive ability of their measure of economic interest and their procedure of signing bills by creating a long–short portfolio strategy. They buy (sell) a value-weighted industry portfolio when the economic interest signing measure for a bill is positive (negative). The portfolio is rebalanced monthly, and stocks enter the portfolio one month after passage of the bill. Using various statistical methods—including the capital asset pricing model, three-factor alphas, the Fama–French three-factor alphas, and the Carhart four-factor alphas—the authors determine that the strategy earns abnormal returns of 76–92 bps, hence confirming that the voting of incented legislators indeed has an impact on stock returns. Most of the abnormal return comes from the short side of the strategy. The authors also show that there is no run-up in the stock price before passage of the bill and that cumulative abnormal returns drift upward for up to 60 trading days after bill passage.

To further refine their measure of economic interest to obtain a continuous measure of interested voting, the authors calculate the difference between the percentage of interested senators voting in favor of the passed bill and that of uninterested senators voting in favor of the bill. They call this the “interested vote.” They also regress the interested vote against certain variables that represent the real effects on the fundamentals of the industry.

Finally, the authors perform robustness and additional tests. To test the robustness of their methodology, the authors change the scope of their measure of economic interest or interested vote. Running their additional tests, they find that when a senator’s personal stockholdings in industries affected by a particular bill fall, that senator’s vested interest in the bill increases, but it does not add any more value to the economic interest signal and is, therefore, only secondary in importance. Similarly, the authors find that greater lobbying by industry groups in favor of or against a particular bill weakens their economic interest signal, perhaps because lobbying narrows the gap between interested and uninterested legislators.

The authors add a new dimension to the power of legislation in affecting stock market prices and real fundamentals of firms. The authors have used robust statistical analysis to prove that there is important information embedded in the way legislators vote and that this information is not realized immediately by the market. In a way, their research implies that law-savvy investors might have an advantage over the average investor when it comes to preempting legislative action.