Tactical Asset Allocation and Presidential Elections
Grant, James L., Trahan, Emery A., Financial Services Review
Over the past 75 years, common stocks performed better under Democrats, while U.S. government bonds and Treasury (T) bills performed better under Republicans. Using a mean-variance framework, we find that Democrats provide better risk-reward opportunities for portfolios weighted toward stocks, while Republicans provide better tradeoffs for portfolios weighted toward government bonds and T-bills. More recently, Republicans provide better portfolio opportunities than Democrats for a bond-stock allocation range typical of diversified investors. Moreover, when segmenting the value stock (style) premium by political party, we find that Republicans provide better risk-reward tradeoffs than Democrats for portfolios of value stocks, bonds, and bills. © 2006 Academy of Financial Services. All rights reserved.
JEL classification: D14; G11; G12
Keywords: Presidential elections; Tactical asset allocation; Calendar investing
The issue of tactical asset allocation (TAA) around calendar events, such as U.S. presidential elections, is a controversial one for investors.1 At the heart of the matter is whether or not the capital market is efficient in the sense that security prices fully reflect the information content of known events. If so, then calendar events, such as presidential elections, are irrelevant to current investment decision making because security prices already reflect the information content of any perceived patterns or cyclicality. Conversely, if investors evaluate the investment consequences of calendar events in a somewhat inefficient market, or if the outcomes of presidential elections impact the returns on various asset classes, then a series of questions emerge that are relevant to tactical investing.
Applied to U.S. presidential elections, a prominent four-year calendar event, these active investing questions are as follows: Are asset prices impacted by a four-year presidential election cycle? If so, what are the effects on different asset classes (stocks, bonds, bills, etc.) according to the political party elected into office? More important, as presidential elections come and go, should investors depart from their long-term or strategic asset allocation to pursue a TAA posture?
Our initial focus is on whether asset returns vary by the political party in office. If asset prices are related to presidential elections, then investors will want to consider information pertaining to election outcomes in making asset allocation decisions. Tactical investing around a four-year election calendar would hold the possibility of earning superior returns (alpha). Anecdotal evidence suggests that many investors follow expected election outcomes closely.
We update the results of prior studies from 1929 through the 2004 election. Consistent with prior evidence on movements in asset prices around presidential elections, we find that returns on large- and small-company common stocks are higher under Democratic administrations, although the results are statistically different only for small-company stocks. The returns on long-term government bonds and Treasury (T) bills are higher under Republican administrations. Post-1960, only the results for T-bonds are statistically significant.
We then argue that the TAA decision around presidential elections should be addressed in the context of an efficient frontier analysis of portfolio opportunities rather than the traditional stock-only or bond-only allocations studied in prior literature. To our knowledge, this is the first paper in the literature that addresses asset returns around presidential elections in a mean-variance efficient frontier framework. We find that the efficient frontier is sensitive to presidential time periods, with Democrats providing a broader (to equal) set of risk-reward opportunities over the long term, while Republicans provide better opportunities over the past quarter century when considering bond-stock allocations typical of diversified investors. …