ملاقات واشنگتن با وال استریت: بررسی دقیق تر پازل چرخه ریاست جمهوری
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|12996||2014||20 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of International Money and Finance, Volume 43, May 2014, Pages 50–69
We show that the annual excess return of the S&P 500 is almost 10 percent higher during the last two years of the presidential cycle than during the first two years. This pattern cannot be explained by business-cycle variables capturing time-varying risk premia, differences in risk levels, or by consumer and investor sentiment. We formally test the presidential election cycle (PEC) hypothesis as an alternative to explain the presidential cycle anomaly. The PEC states that incumbent parties and presidents have an incentive to manipulate the economy (via budget expansions and taxes) to remain in power. We formulate eight testable propositions relating to the fiscal, monetary, tax, unexpected inflation and political implications of the PEC hypothesis. We do not find statistically significant evidence confirming the PEC hypothesis as a plausible explanation for the presidential cycle effect. The presidential cycle effect in U.S. financial markets thus remains a puzzle that cannot be easily explained by politicians employing their economic influence to remain in power, as is often believed.
The presidential cycle effect in U.S. stock market returns consists of higher stock market returns during the second half of a presidential term compared to the first. This phenomenon first appeared in Hirsch's Stock Trader's Almanac in 1967 and has returned in the yearly Almanac ever since. The first academic interest dates from at least the 1980s. Huang (1985) reports that trading strategies based on the presidential cycle effect produce returns superior to a traditional buy-and-hold strategy. Foerster and Schmitz (1997) examine each year of the presidential cycle individually and conclude that U.S. stock market returns are significantly lower in the second year of the presidential term, compared to the other three years. The question why there is a relation between the presidential cycle and stock market returns has puzzled academics for years. Several market efficiency explanations have been put forward. First, the presidential cycle might merely proxy for variations in expected returns due to business cycle fluctuations. Booth and Booth (2003), however, find that this is not the case. Second, the relationship between the presidential cycle and stock market returns could be concentrated around and limited to election dates. However, Santa-Clara and Valkanov (2003) find no significant evidence of stock price changes immediately before, during, or immediately after presidential elections. Third, the difference in returns during the presidential cycle might be a compensation for risk. Market volatility could simply be higher in the second half of the cycle, thereby explaining the higher returns. Campbell and Li (2004), however, indicate that the differences in returns cannot be explained by differences in market volatility. Finally, the presidential cycle effect might be driven by the impact of outliers. Gärtner and Wellershoff (1995) as well as Foerster and Schmitz (1997) find that the effect is not driven by individual outliers in the data, such as the October 1987 stock market crash. Since most rational explanations fail to provide an adequate answer, we formulate a testable framework of the presidential election cycle theory as an alternative. We label this framework the presidential election cycle (PEC) hypothesis. It is based on the macroeconomic political business cycle (PBC) theory by Nordhaus (1975) and MacRae (1977) and states that political parties that try to win elections often manipulate business conditions. Nordhaus (1975) argues that presidential administrations have an incentive to stimulate the economy prior to the elections in order to increase the probability of their electoral success. Fair (1982), who develops a model for voting behavior, indicates that voters do not look back more than a year or two in judging the economic performance of an administration. His result might give presidents an incentive to manipulate the economy prior to the elections, since the myopic electorate only judges the administration on its last years. Rogoff (1990) provides rational underpinnings for the PBC hypothesis by introducing the assumption of information asymmetry whereby policy makers are better informed than voters about their competence. The significant interactions between macroeconomic outcomes and presidential administrations are also confirmed by Chappell and Keech (1986) and Alesina and Sachs (1988), while Tufte, 1978 and Grier, 1987 and Haynes and Stone (1988) find empirical evidence in favor of an American PBC. Under the PBC theory, an incumbent president would impose stimulative economic measures and corporate friendly policies to create a favorable voting environment close to elections. In order to create this PBC, some economic policy instruments must be manipulated. For electorally-inspired policy making to have macroeconomic consequences, incumbent presidents must either manipulate fiscal or monetary policy. However, the independence of the Federal Reserve System should guard against political pressures, and little empirical evidence for the existence of a political monetary cycle has been uncovered for the United States (Golden and Poterba, 1980 and Beck, 1987).2 Researchers have therefore concluded that fiscal policy, a macroeconomic tool more directly under the control of the president, is the tool of choice and opportunity (Abrams and Iossofov, 2006). Rogoff (1990) suggests that, in equilibrium, political budget cycles developed by incumbent governments tend to increase consumption spending, lower taxes, and raise transfers prior to and during election years. This would result in an increase in federal spending, less regulation, lower taxes and a mounting growth of money supply prior to elections. Investors would become confident and optimistic about the upcoming election. This would lead to a bullish stock market in the second half of a presidential term and should work independently from variation in expected returns due to business cycle variation. After the election, the sentiment of anticipation would decrease as investors become disappointed with the presidential administration for not realizing their campaign promises. These sentiments, coupled with the fact that the price for the stimulative policies conducted prior to the elections have to be countered with post-election deflationary measures, would lead to a bearish stock market in the first half of the cycle. The popular press routinely points at the presidential election cycle hypothesis as an explanation for the stock market performance. This study formally tests the presidential cycle effect and examines a broad range of possible explanations for this effect in U.S. stock and credit markets. This is the first study that not only analyzes rationally motivated explanations, such as business cycle fluctuations and time-varying risk, but also the implications of the presidential election cycle (PEC) hypothesis and investor and consumer sentiment. We furthermore subject our results to a range of robustness checks, including robust and bootstrapped standard errors and an analysis of sub-samples. We also examine whether the presidential cycle pattern shows up in expected and unexpected returns. Using data over the sample period 1948–2008, we find a clear presidential cycle effect in both U.S. stock and credit markets. During this period, stock returns in excess of the risk-free rate are on average −2.02, −3.45, +12.78, and +1.27 percent in year one to four of the cycle, respectively. The pattern in real stock returns is comparable: −0.55, −2.45, +13.59, and +2.01 percent, respectively. We find a similar pattern in the credit spread, especially in the second and the third year, where the credit spread moves by +34 and −28 basis points, respectively. These findings are statistically significant and the null hypothesis of equal returns (spread changes) across the years is strongly rejected. We show that our findings are not attributable to outliers. We then turn to potential explanations for this statistically strong pattern in returns. First, the political term may coincide with changes in the business cycle that have been shown to track variation in expected returns (see, e.g., Fama and French, 1989). When we control for the dividend yield, default spread, term spread, and relative interest rate, the results do not change. Business cycle variation, therefore, does not explain the presidential cycle in stock and credit markets. As a second potential explanation, we consider time-varying market risk. If the last years of the presidential cycle are more risky than the first, rational investors demand a higher expected return. However, stock market volatility is lower, rather than higher, in the third year compared to the second year of the cycle. Furthermore, we do not find significant differences in stock market risk across the four years of the presidential term. As a third potential explanation we investigate the presidential election cycle (PEC) hypothesis and formulate eight testable implications of the theory. In this way, we extend the existing literature by empirically examining the presidential cycle pattern through an analysis of the full scale of policy tools available for economic manipulation by an incumbent president. We consider fiscal, macroeconomic and political variables. Our first five propositions have a financial, fiscal or macroeconomic character and focus on any potential economic manipulation by an incumbent president. Accelerating growth of money supply, increasing unexpected inflation, lowering U.S. income tax levels or raising U.S. federal spending are examples of popular presidential manipulations suggested by the PEC hypothesis. We find, however, little evidence in favor of any of these presidential manipulations of the economy. Our final three propositions address the political mechanisms behind the presidential cycle effect. The partisanship of the president appears to have no significant impact on the strength of the presidential cycle effect. This is supportive for the PEC hypothesis, which like Nordhaus' (1975) political business cycle (PBC) theory, states that presidential administrations have the same incentive when manipulating the economy (namely enhancing their chances of re-election), regardless of their political orientation. We analyze the impact of the partisanship of the majority of Congress, but find no congressional influence on the strength of the presidential cycle effect. This lack of influence is at odds with the PEC hypothesis and diminishes its political credibility. As a final check, we investigate whether the eligible status of an incumbent president has an impact on the presidential cycle effect. One would expect more economic manipulation when there is a re-eligible president in office. Although we find a clear presidential cycle effect in excess returns when there is a re-eligible president in office, it disappears when there is no re-eligible president in office. In summary, we find very limited empirical evidence in support of the PEC hypothesis, except for some of the political propositions. Fourth, we examine whether the presidential cycle is present in expected and unexpected returns. Interestingly, we find that the presidential cycle effect is prevalent in unexpected returns, but non-existent in expected returns. This result suggests that investors are systematically surprised during the second half of the presidential term. Given the absence of a presidential cycle effect in fiscal and monetary policy variables, it is unclear what can be the underlying cause of this persistent bias. As a final exercise, we therefore consider changes in consumer and investor sentiment as a potential explanation. There is no clear pattern over the years of the presidential term in consumer sentiment (measured by surveys from the Conference Board and University of Michigan) and investor sentiment (measured as the first principal component of a range of sentiment indicators from Baker and Wurgler, 2006). When controlling for sentiment, the presidential cycle pattern in stock market returns remains significant. Thus, we conclude that the existence of a persistent presidential cycle surprise in U.S. financial markets remains a puzzle that cannot be easily explained by politicians employing their economic influence to remain in power, as is often believed. The remainder of this paper is structured as follows. Section 2 introduces the data and variables used in this study. The empirical findings on the presence of the presidential cycle effect in U.S. stock and bond markets are presented in Section 3. Section 4 examines potential explanations related to the business cycle, time-varying risk, the PEC hypothesis, expected versus unexpected returns, and consumer and investor sentiment. Our main conclusions are presented in Section 5, which also sets out the agenda for future research.
نتیجه گیری انگلیسی
This paper documents the existence of the presidential cycle effect in U.S. stock and credit markets. The average excess return of the S&P 500 is a significant 10 percent higher during the first half of the presidential cycle than during the second. Furthermore, the average excess return is −5.6 percent in the second year and +10.6 percent during the third year. Although less strong than in the case of the U.S. stock market, we show that the presidential cycle effect also exists in the U.S. corporate bond market. Changes in the credit spread indicate a pattern influenced by the presidential cycle. On average, the spread widens by 27 basis points in the second year, while it contracts by 35 basis points in the third year of the cycle. The results are statistically and economically significant, stable over sub-samples, and robust to controlling for business cycle effects and time-varying risk. We conclude that the presidential cycle effect in U.S. stock and bond markets is a robust phenomenon. As a potential explanation for this phenomenon, we investigate the presidential election cycle hypothesis (PEC hypothesis) by designing eight empirically testable propositions. The popular press routinely points at the PEC hypothesis as an explanation for the presidential cycle effect. However, after a thorough empirical analysis, there is little to no financial, inflation, fiscal or macroeconomic evidence for any economic manipulation by an incumbent president. Neither the growth of money supply, U.S. income tax levels, U.S. federal spending, nor the U.S. budget indicate a statistically significant presidential cycle pattern. Furthermore, the political propositions we test fail as well in uncovering any significant evidence for the political background behind the presidential cycle effect. We argue that the credibility of the PEC hypothesis as an explanation for the presidential cycle is only limited. We finally turn to consumer and investor sentiment as explanations for the pattern. Although consumer sentiment shows a somewhat similar pattern, correcting for the business cycle and sentiment jointly does not eliminate the pattern observed in stock returns. We conclude that the existence of the presidential cycle effect in U.S. financial markets remains a puzzle and certainly deserves further academic attention. However, since most rational explanations as well as the PEC hypothesis fail to solve the puzzle, alternative explanations become scarce. The conventional wisdom that the presidential cycle effect is caused by politicians misusing their economic influence to remain in power is not supported empirically.