LONG TERM VOLATILITY ON S&P 500 INDEX AND ITS RELATION TO POLITICAL PARTIES

Financial markets have been badly affected by COVID-19. Especially market indexes plunged dramatically in a short while. S&P 500 dropped by 22.8% between February and April but it quickly started to recover through April and May. However, the index tumbled to its trough by approximately 85% in June 1932 from its peak in September 1929. Of course, this comparison does not seem so coherent before seeing index prices during the next two years but it can be easily said that the index has begun to rise shortly after its dip occurred because of COVID-19. As of June first, two months after its dip, the index rebounded more than two-thirds of its losses. Nevertheless, it could not rebound even half of its losses until March 1937 from its June 1932 dip and it took almost a quarter of a century to reach the same price it had in 1929. Some forecasts which I am planning to discuss in my next article show that the index will have recovered all its losses by March 2022.

This article intends to understand if political parties play a big role on the volatility of the S&P 500 index returns. As we know, volatility is a variance change on time series. In this article, I picked up the index S&P 500 indicating weighted average monthly stock prices of 500 biggest corporations in the U.S. as a time series from January 1928 to June 2020 in order to make a volatility analysis by using an exponential generalized autoregressive conditional variance model. I also try to explain how political parties have changed the volatility on the index over nearly 92 years. I make a comparison among the terms of democrats and the terms of republicans based on their impacts in the economy by observing the volatility on the returns of the biggest market index of the U.S., S&P 500. My criticism is that if volatility on a market index rapidly increases, it harms the economy. I test in which political party’s term the volatility on the index has increased or decreased more throughout almost a century. 

  

I would like to show two different graphs of the S&P 500 index before jumping to volatility analysis. The first one is the nominal (unadjusted with inflation) prices of the market index. 

This graph represents only pure index prices during the term. It merely helps us to get how prices have changed over the term but it cannot provide a comprehensive and right benchmark among crises. When looking at the graph, it may be said that the market index is practicing a price bubble now. Yet, it is not true because we cannot see the month by month percentage price change on this graph.

On the other hand, price change in the logarithmic base can be seen in the second graph and it gives a better presentation to interpret the duration. Since the logarithmic series has a steady trend, which is called trend stationary process, its trend line can be used basically as an average benchmark to determine whether or not a price bubble is currently being revealed. As long as the prices which are on the trend line does not deviate much from itself, a price bubble does not exist. Seemingly, the index prices have been tracking that trend line for almost the last two decades so the index is not practicing a price bubble at the moment. 

Furthermore, the logarithmic series can be used to compare the impacts of crises. The logarithmic series shows that there is no similarity between the great depression and COVID-19 economic crisis in terms of price changes on the S&P 500 index. As can be seen in the logarithmic series, price decline from September 1929 to June 1932 was immensely large compared to price decline from February 2020 to April 2020. Briefly, COVID-19 economic crisis is temporary and it seems like its harms would not affect the U.S. economy in the long run as it was after the great depression stealing the country’s 25 years.   

After touching on the importance of the logarithmic price series we continue to make volatility analysis on the returns of the S&P 500 index and test if there is a trade-off between high volatility and terms of political parties. As I mention in this article before, my standpoint is that high volatility harms economies and my purpose is to remark that the dimension of volatility has been larger in which political party’s terms.

In order to draw a picture regarding the volatility change, differences of logarithmic prices that indicate the percentage change of S&P 500 index prices are used. We call it returns series along the rest of this article. The returns series created by the difference of the logarithmic price series turns the analysis to a more comprehensive and serious one. First of all, the returns series has no trend since returns turn around zero mean (average) thus the returns series describes a difference stationary process. Hence returns of the index do not follow a specific path, they have been asymmetrically changing during these 92 years. Whenever a return is above 0, it is a positive return (profit); whenever a return is below 0, it is a negative return (loss). Yet it does not say enough to make comparisons among largeness of crises occurred over time therefore dimensions of returns should be measured to demonstrate the harmful impacts of distinct crises. Volatility on the returns series exhibits those impacts and allows us to observe how large they are.   

Depart from that, which political party’s terms have experienced more and larger volatility can be determined. How long political parties have kept the power during the period are seen on the third graph with red timelines for terms of republicans and blue timelines for terms of democrats. Democratic Party has ruled the country for 49 years as Republican Party has been ruling it for 43 years since 1928. The longest uninterrupted term during the period belongs to democrats with 20 years while the longest uninterrupted term of republicans is only 12 years. Apparently, terms of republicans are more volatile than terms of democrats on the returns series shown by the third graph. General volatility on returns during terms of republicans is larger particularly in 1928-1933 term, 1981-1993 term, 2001-2009, and current term. Conversely, returns do not practice large general volatility during terms of democrats except the 1933-1953 term. When watching the largeness of negative devastating volatility causing the crisis, Republican Party faces negative volatility on returns more than Democratic Party did. Throughout the period, republicans have been practicing 4 severe economic crisis including the great depression, oil crisis, mortgage financial crisis and COVID-19 economic crisis as democrats practiced only second world war economic crisis and dotcom bubble whose dimensions of its volatility are much smaller than other crisis occurred in Republican Party’s terms.    

The returns series on the third graph also indicates that COVID-19 economic crisis is as bad as the mortgage crisis but it is not even close to the great depression. Besides, oil crisis and crisis exposed owing to the second world war look similar.

Volatility on financial time series is affected by news and information announced to the public. According to the EGARCH model using the S&P 500 index returns during the period, the effect of incoming information to markets is persistent therefore volatility clusters are being revealed over time and there is leverage effect on returns of the index. The leverage effect states that negative news contributes to negative volatility more than positive news contributes to positive volatility so returns of the S&P 500 are more sensitive to bad news.

Believe or not it is just a prospect but the conspicuous thing in this work is that each time republicans come to power, bad news, and despair start to circulate and a sudden crisis pops up. And also, the power passes to Democratic Party notedly after a volatility fluctuation is practiced in Republican Party's terms during the period. Who knows this routine might perish in the coming election.