One of the most striking and persistent anomalies observed in U.S. financial markets is the apparent linkage between stock market returns and the four year cycles associated with the presidential election political process.
In particular, stock market returns are much higher on average during Democratic presidencies than during Republican administrations. Figure 1 illustrates that, over the 1948-2009 time period, the average of annual excess returns for large company stocks is 11.82 percent during Democratic administrations compared to only 5.05 percent during Republican administrations. For small company stocks the difference in average excess returns is even higher, averaging 17.64 and 6.79 percent during Democratic and Republican administrations, respectively. This partisan effect is quite remarkable given the depth, liquidity, and presumed efficiency of the U.S. equity markets.
Political scientists and economists have studied the effects of the election cycle and political competition on the economy since the early 1970s. Alesina and Rosenthal (1995), Alesina, Roubini, and Cohen (1997) and Drazen (2000a, 2000b) provide thorough surveys of this research and provide empirical evidence in support of political impacts on macroeconomic variables such as GDP growth and employment.
Academic research interest in the potential links between the election process and the stock market, however, has been more recent. Allvine and OlNeill (1980) test a trading strategy that designed to exploit pre-election strength in stock market returns. Riley and Luksetich (1980) examine the movement of stock prices during election years. Herbst and Slinkman (1984) report evidence of a four year stock market cycle that is closely associated with presidential elections. Huang (1985) documents the existence of both partisan and year of term effects in large company stock market returns. Hensel and Ziemba (1995) report a partisan effect and year of term effect for large and small company stocks.
In addition, they document a partisan effect in which bond and cash returns are higher in Republican administrations. Johnson, Chittenden, and Jensen (1999) consider large and small company stocks as well as cash and bond returns. They consider both nominal and real returns. They find no partisan effect for large company stocks, a significant partisan effect for small company stocks, confirm the Republican partisan effect in bond returns, and document a year of term effect in stock market returns. Siegel (1998) documents the existence of both partisan effects on returns and presidential election cycle seasonal effects.
Santa Clara and Valkanov (2003) provide a thorough examination of the evidence of partisan impacts on U.S. equity returns. Using monthly returns data from 1927 o1998, they find that partisan differences in excess returns are statistically significant and robust to choice of sample period. Additionally, they break down excess returns into expected and unexpected components and find that it is the unexpected component that accounts for most of the variation in partisan differences. They conclude that the market is systematically positively surprised by Democratic policies. But how can systematically higher returns that depend on the party of the president be reconciled with a liquid and efficient market populated with rational, forward looking agents?
Download
PDF Ebook Stock Market Returns and Political Business Cycles
