نقشه برداری از چرخه انتخابات ریاست جمهوری در بازارهای سهام ایالات متحده
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|15659||2009||11 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Mathematics and Computers in Simulation, Volume 79, Issue 11, July 2009, Pages 3267–3277
This paper shows that in the almost four decades from January 1965 through to December 2003, US stock prices closely followed the 4-year Presidential Election Cycle. In general, stock prices fell during the first half of a Presidency, reached a trough in the second year, rose during the second half of a Presidency, and reached a peak in the third or fourth year. This cyclical trend is found to hold for the greater part of the last ten administrations, starting from President Lyndon Johnson to the administration of President George W. Bush, particularly when the incumbent is a Republican. The empirical results suggest that the Republican Party may have greater cause to engage in active policy manipulation to win re-election than their Democratic counterparts. There is irony in that bullish runs in the stock market have tended to coincide with sub-periods under Democratic administrations. The existence of the Presidential Election Cycle shown in the paper may constitute an anomaly in the US stock market, which could be useful for investors.
Prices are the outcomes of volatile human expectations, shifting the supply and demand lines, and causing prices to oscillate. Fluctuations in prices are a natural process of changing expectations, thereby leading to cyclical patterns. There are many kinds of cycles, with the combined effect of driving movements in stock prices. Hurst  is a classic in the annals of cyclical analysis. Historically the stock market has had a role as one of the most sensitive indicators of the business cycle, and one of the most influential variables in the government's index of leading economic indicators , ,  and . Six interesting cyclical patterns are worthy of note: 28-day Trading Cycle (the Lunar Cycle), 10.5-month Futures Cycle, January Effect, 4-year Cycle (the Kitchin Cycle and the Presidential Election Cycle), 9.2-year Cycle (the Juglar Cycle), and 54-year Cycle (the Kondratieff Cycle). The 28-day cycle, which is also known as the Lunar Cycle, was found as early as the 1930s in the wheat market. Regardless of the actual causes, many markets, including stocks, appear to follow the 28-day cycle. Although individual commodities exhibit their own unique cycles, another cycle ranging between 9 and 12 months has been found in the Commodity Research Bureau Index, which is a good reflection of the overall behaviour of commodity prices. In addition, the stock market has shown an uncanny tendency to end the year higher (lower) if prices increase (decline) during the month of January, leading to the saying: “so goes January, so goes the rest of the year.” Between 1950 and 1993, the January Effect was correct on 38 of 44 occasions, with an accuracy rate of 86%. Financial astrology is one of the explanations for the cycle analysis, but they are not based on the study of market prices in the derivation of the cycles’ numbers. W.D. Gann, a well-known trader in the 1950s, is one of the pioneers in adopting financial astrology, namely the Jupiter–Saturn cycle, to facilitate trading activities. Since Jupiter and Saturn are the largest planets in our solar system, their gravitational pull when they are in line is believed to be one of the strongest. The effect of this gravitational pull is so strong that it causes the sun to shift periodically, based on their positions around the true center of the mass of the solar system. Such shifts will cause the weather to change substantially, which will eventually affect commodity prices. As commodity, bond and stock markets overlap closely, such linkages will spread with spillover effects into bond and stock markets. In 1860, Clemant Juglar found that a cycle lasting approximately 9 years existed in many areas of economic activity, including the stock market. Subsequent research found that this cycle had a significant presence during the period 1840–1940. Named after a Russian economist, Nicolai Kondratieff, the Kondratieff Wave is a long-term, 54-year cycle that has been identified in prices and economic activity. However, since this cycle is extremely long, to date it has only repeated itself three times in the stock market. The most popular cycle is the 4-year cycle, which is also known as the Kitchin Wave and the Presidential Election Cycle. The theory that seeks to explain the relationship between stock prices and Presidential elections is called the Theory of the Presidential Election Cycle. Every 4 years, American voters elect a President. Due to the President's overwhelming influence on both domestic and world affairs, the ripple effects of Presidential elections are staggering. Hence, it should not come as a surprise that stock prices, often called a leading indicator of the macroeconomy, are affected by Presidential elections. It is well known that business prospers in an environment of low taxes and stable government policies, which is precisely what policy-makers serve up in an election year when no public official wants to be seen as a big tax-and-spend proponent. Therefore, the stock market tends to prosper in an election year. Kitchin  found that a 40-month cycle existed in a variety of financial variables in both Great Britain and the USA between 1890 and 1922. The 4-year Presidential cycle was later found to have an extremely strong presence in the stock market between 1868 and 1945. Stovall  found a pattern of low returns in the first 2 years of a President's term and high returns in the last 2 years in the stock market. Although it is called a “4-year cycle”, the length of the cycle has been found to vary between 40 and 53 months. This paper explores the existence of the Presidential Election Cycle in the USA based on Presidential elections that occur every 4 years. One possible underlying reason is that stock prices may decline following a Presidential election as the newly elected President takes unpopular steps to make adjustments to the economy. In mid-term, stock prices may begin to rise in anticipation of a stronger Election Day economy. Schwert  and Fama  have confirmed that stock prices are correlated with future economic activity in statistical regressions. Confidence in the President may implicitly reflect the underlying economic conditions, which are important in determining stock prices. Recently, Decker and Wohar  found that the probability of the incumbent party losing a state previously carried increases with petroleum product prices, but only in those states that have primarily energy consuming economies. They also found that increases in the number of international conflicts, increases in real state per-capita income growth, and increases in state per-capita grants-in-aid all reduce the likelihood of losing previously carried states, while higher taxation growth increases this likelihood. The pattern in stock market prices related to the 4-year Presidential term has been the focus of Wall Street pundits for some time. Allivine and O’Neill , Gartner and Wellershoff , Hensel and Ziemba , Huang , and Booth and Booth  found that the difference between the returns in the first half and those in the second half of the Presidential term are economically and statistically significant. In addition, Huang  reported that such a Presidential cycle has persisted in both Democratic and Republican administrations, but is more pronounced in Democratic administrations. Examining each year's returns in the Presidential term, Foerster and Schmitz  found that both US and international stock returns were lower in the second year of the US Presidential term relative to those in years 1, 3, and 4. Finance academic researchers have tended to focus their attention on how monetary policies and business conditions might explain stock market movements as the reason for a Presidential Election Cycle in the stock market. In order to explain the causes of a Presidential Cycle, Nordhaus  commented that political business cycles are often manipulated by political parties in trying to win elections. He argued that Presidential administrations have an incentive to stimulate the economy prior to elections and to pursue deflationary policies after the elections, regardless of the political orientation of the incumbent. Alesina and Sachs  and Willet  suggested that the Presidential cycle in stock return coincided with the business cycle because inflationary and deflationary measures were often pursued through fiscal and monetary policies. Hence, political factors are conjectured to influence the economy through these economic policies, which affect the timing and severity of the inherent business conditions. Examining the role that monetary policy has played in explaining stock returns, Jensen et al.  and Thorbeche  discovered that monetary policy can help explain future stock returns. Rogoff  argued that political budget cycles developed as government tended to increase spending, especially on projects with high immediate visibility, cut taxes, and raise transfers prior to and during election years. Megna and Xu  used indicators to forecast the economy and cycles for the USA. Booth and Booth  confirmed the previous findings and found that the Presidential Cycle had explanatory power beyond the business condition proxies that have been shown to be important in explaining stock returns. However, Santa-Clara and Valkanov  found that the difference in returns could not be explained by business cycle variables that were related to expected returns, and were not concentrated around election dates. There was no difference in the riskiness of the stock market across Presidencies that could justify a risk premium. The differences in returns through the Presidential Cycle remain a puzzle. Several studies have shown the interactions between macroeconomic outcomes and Presidential administrations. Chappell and Keech  and Alesina and Sachs  found that macroeconomic outcomes during Democratic administrations differed from those during Republican administrations, with higher (lower) average rates of growth in GNP in the first half than in the second half of Democratic Republican administrations. In addition, numerous studies have examined the relationship between Presidential administrations and the stock market. In a comprehensive examination of historical stock returns over the period 1988–1992, Siegel  revealed that the stock market, as measured by the Dow Jones Industrial Average (DJIA), rose by an average of 0.81% on the day following Republican victories, and fell by 0.75% on the day following Democratic victories. An earlier study by Riley and Luksetich  also documented a favourable (unfavourable) market reaction following a Republican (Democratic) victory. However, Kansas  reported that the stock market generated an average of 11.1% in the fourth year of a presidential term, while the market provided a 15% return in the fourth year of Democratic Presidencies. These studies suggest that stock returns are closely related to Presidential administrations. However, politics and financial markets share a common truth that nothing lasts forever. In the year following an election, voters can be euphoric as they look forward with anticipation to the achievement that the President hopes to accomplish. Around this time, policy-makers may begin to feel less restrained about introducing new programs that will establish their agenda. This somewhat cynical view holds that the early days of a new administration are the best time to make big changes as the myopic electorate will forget about them over the next 3 years. These new programs and economic policies usually involve more taxes, more spending, and more regulations, which begin to eat into business profits. As a consequence, stock prices that follow corporate earnings will likely fall. This negative environment reaches a nadir during the second year of a Presidential term, following which policy-makers begin to change their ways to be more “voter friendly” before the next election. Once again, they become more conservative and less prone to new tax-and-spend policies, such that markets begin to recover before the next election. In the literature, a Presidential Cycle is hypothesized and a t-test or simple regression analysis is typically used to test the hypothesis. In this paper, we contribute to the literature by applying spectral analysis to estimate the periods of the cycles, which verifies a formal existence of the 4-year Presidential Election Cycle in the US stock market. The limitations of applying a t-test or simple regression analysis are that they do not take into consideration the time dependence and conditional heteroskedasticity (or the GARCH effect) in stock returns. In order to circumvent these problems, we adopt the exponential GARCH Intervention model to examine the Presidential Cycle. It is found that, in general, stock prices fall during the first half of a Presidency, experiencing a trough in the second year, and rise during the second half of a Presidency, reaching a peak in the third or fourth year. Furthermore, on average the cyclical pattern during a Republican administration is much more significant than that of a Democratic administration, which suggests that the Republican Party may engage in policy manipulation to win elections more frequently than may their Democratic counterparts. The structure of the remainder of the paper is as follows. Section 2 discusses the data and methodology. Section 3 reports and explains the empirical findings from estimation. Section 4 provides some concluding comments.
نتیجه گیری انگلیسی
The political behaviour of governments may have an important effect not only on the macroeconomic performance of the whole economy, but also an important influence on the microeconomic behaviour of all individuals. More often than not, the incumbent government can try to manipulate economic policies in both the macro- and micro-spheres in order to achieve specific political goals, including re-election. This paper analyzed the impact of Presidential elections on stock prices with reference to the USA, which has not only a well-developed political electoral system but also a mature and sophisticated stock market. The empirical results using spectral analysis and the EGARCH Intervention model found that, since 1965, the US stock market has experienced several robust and quantitatively important Presidential Election Cycles. We first performed unit root tests to determine whether the stock index series was stationary, using the augmented Dickey-Fuller (ADF) test. Then we used spectral analysis to show the types and strength of the existing cyclical patterns in the US stock market, especially the Presidential Election Cycle. This was followed by using the EGARCH Intervention model to determine whether any Presidential administration might have tried to manipulate economic policies to win re-election. Finally, we stratified the data by political affiliation, specifically Republicans and Democrats, in order to differentiate their respective attitudes toward policy manipulation with a mind to winning elections. The result of the spectral analysis indicated that stock prices in the USA exhibited several cycles, with the 4-year Presidential Election Cycle being the most prominent cycle, with the strongest spikes being spaced at about 200 weeks (or around 4 years) apart. The results of the EGARCH Intervention analysis showed that there were statistically significant Presidential Election Cycles in the US stock market during the greater part of the past four decades, at both the overall as well as individual levels. In other words, during some of the ten 4-year Presidential Election Cycles, stock prices decreased with a statistically significant amount in the second year, and then increased by a statistically significant amount in the third year of the Presidential Election Cycle. Furthermore, on average the cyclical pattern during a Republican administration was much more significant than that during a Democratic administration, which suggests that the Republican Party may have dabbled in policy manipulation more actively to win re-election than did the Democratic Party. However, the efficacy of activist policy manipulation was thrown into some doubt by the irony that bull runs in the US stock market tended to coincide with sub-periods under Democratic rather than Republican administrations. The empirical results reported here provide some justification for investors to incorporate the Presidential Election Cycle in their investment decisions in order to obtain greater returns. For even better returns, investors could incorporate investments that are timed to the Presidential Election Cycle with other approaches, such as technical analysis ,  and  and fundamental analysis or strategies based on the economic situation and financial anomalies , , , ,  and . They could also enhance their investment decisions by applying other advanced forecasting and statistical techniques , ,  and . We note that the cyclical evidence could reflect risk premiums and market inefficiencies if such evidence were to be prolonged. These anomalies are of particular interest because their existence violates the weak form of market efficiency. In an efficient stock market, prices fully reflect all the available information, so that investors make only normal profits and cannot make abnormal returns by exploiting these anomalies. As these anomalies are relatively easy to exploit, they should weaken or disappear over time. If the market is efficient, cyclical evidence should not persist for lengthy periods because market forces would lead to adjustments and to the absence of cycles. For example, if the January Effect is found to exist, investors would purchase stocks in December and sell them at the end of January. This would continue, thereby driving up stock prices in December and down in January, until prices in January relative to their prices in December were low enough to make the marginal investor indifferent to trading in December or January. Therefore, it is not surprising that the recent literature has found evidence that the January Effect has been diminishing and/or disappearing  and .8 However, there are some financial anomalies that have not disappeared.9 The Presidential Election Cycle was discovered by Stovall in 1992, and the results presented here show that its effect is still significant. One possibility for the existence of the Presidential Cycle to hold for long periods is that many investors do not realize that such a cycle exists, or do not pay attention to capitalize on the effect. One might expect that such an effect would diminish or disappear over time. If it does not do so, then it could be considered to be a financial anomaly. It should be noted that cycles of different length and frequency might be present in stock markets, with the Presidential Election Cycle being one of the more important. Alternative shorter cycles, such as the Monday Effect and the January Effect, may cause prices to oscillate in stock markets. Alternative longer period cycles, such as the Juglar Cycle of 9.2 years and the Kondratieff Cycle of 54 years, may also lead to fluctuations in stock prices. The stock market crash in 1929 and the stock market meltdown in 2008–2009 may belong to an entirely separate long period cycle. As the purpose of the paper is to examine the effect of the relatively short 4-year Presidential Election Cycle, as distinct from a longer period cycle, we have not included data beyond 2003.