My Writing

Investing

The Election Year Cycle

Originally published at About.com - March 2, 2000

Ace stock picker Pat McKeough occasionally mentions the U.S. Presidential election cycle as a possible factor in stock market performance. Without elaboration he has noted that stock markets tend to do well in Presidential election years, poorly in the first year or two of a term and well in the last two years of a term.

But is there any basis in fact for the cycle argument? Intuitively it makes sense. After all, an incumbent President wants to get re-elected. And, as Dr. Bart DiLiddo, a writer for Vectorvest notes, "A prosperous economy and a booming stock market virtually guarantee re-election of an incumbent administration".

Fortunately the issue has been studied and the facts noted. An article from HR Consultants shows that since 1940, stock market averages have tended to oscillate in a four year cycle. A graphic representation of this phenomenon shows clearly defined V troughs every four years from 1942 on, though the troughs have virtually smoothed out in the bull market of the 1990s. The most striking example is shown in a graph for the years from 1965 - 1976 with the troughs falling at almost identical points in each Presidential year cycle.

Looking at a graph of the timing and extent of bear market declines in the cycle, HR Consultants shows that 9 of the 13 lows occurred in the second year of an administration, 3 in the first year, one in the third year and none in the fourth or election year. Except for the one anomalous year of the crash of 1987, the last two years of an administration have been good for the stock market.

The rationale is simple. As noted above, incumbent administrations want to look good in the last two years in office. They find a way "to prevent rising interest rates and tight money" notes HR Consultants, "thereby reducing the likelihood that the pain of a recession would incur voter displeasure during the election year". Conversely, they make all the so-called "hard" decisions, such as raising taxes or introducing new and possibly unpopular programs in the first two years in office.

More astounding than the cycles themselves is the analysis of gains and declines in the cycle. Yale Hirsch, the editor of the Stock Trader's Almanac, notes that since 1832, the last two years of each administration "produced a total net market gain of 557% compared with 81% gain of the first two years of these administrations". The average gains for each four year cycle were 13.6% for each of the last two years and only 2.0% for each of the first two years.

And HR Consultants shows with a graph of the composite averages for the S&P 500 and the NASDAQ Composite indices that almost the entire gain for the cycle can "be realized by buying on October 1 of the Year 2 of the Presidential Cycle and selling on December 31 of Year 4".

This has even continued on to the Clinton years. In 1993, the first year of his administration, the S&P 500 Composite rose 7.06%. In the second year it declined 1.54% for an average of 2.76%, a bit higher than the average of 2%. But 1995 and 1996 saw increases in the S&P 500 of 34.11% and 20.26% for an average of 27.18%, more than double the average of 13.6% but validating the general trend all the same.

Clinton's second term, however, is an anomaly. There were no intervening bad years. The gains for 1997, 1998 and 1999 were 31.01%, 26.67% and 19.53% respectively. And 2000, an election year, went completely contrary to the cycle theory as the S&P dropped 10.1%. Historically the cycle is correlated with short term interest rates. But the Clinton administration raised interest rates affecting the broad market as represented by the S&P 500.

These are interesting times. The historical trend of two fallow years followed by two boom years seems to have been broken in the current cycle. 2000 did not follow the pattern but 2001, the first year in the Bush administration, did act as the theory predicts (at least up to Nov. 16th as this is being written). The S&P is down considerably year-to-date, but could still move into positive territory by year's end.

There are some, like Mark Hulbert of the Hulbert Financial Digest, who argue that "the cycle has little intrinsic predictive power". He noted in an article in Forbes in 1996 that the better performing newsletters at the time pooh poohed the cycle and that whether the market were to decline in 1997 as per the theory, it would "not be because of the presidential election cycle". And Hulbert was right to a point. The market, as noted, went up in both 1997 and 1998. However, there was a sharp market drop in September 1998, a bear market trough as predicted by the cycle theory. The only difference was that the market recovered fully by year's end.

An interesting point to consider is that perhaps, with two consecutive negative years in 2000 and 2001 and with interest rates cut to the bone, the cycle will advance a year and 2002 will be a banner year. Interest rates have a stronger correlation than the year by itself.

Links of Interest

The Election Cycle - Dr. Bart A. DiLiddo examines the Election Year Cycle in this article at Vectorvest.

The Four Year Stock Market Cycle - An article on the Election Year Cycle from HR Consultants. (This article is no longer available online.)

It's the Economy, Stupid! - Mark Hulbert of The Hulbert Financial Digest takes a skeptical approach to the Election Year Cycle in this Forbes article from Dec. 2, 1996. (This article is no longer available online.)

The 4-Year Cycle in the Stock Market - A lengthy analysis from Australian Walter Bressert at the Adest Trader's Resources website. (This article is no longer available online.)

Postscript: The market crash of 2008 seems to refute the four year election cycle, however, this article from the Graziadio Business review sees it as an anomaly, not a refutation.

The Four-Year U.S. Presidential Cycle and the Stock Market: An Updated Analysis - from Graziadio Business Review, Vol. 15, # 2 - 2012