Elections As The Most Powerful Short-Term Cycle

Although much has been written about this phenomenon and opinions differ, one thing is clear. Whether the incumbent US President is a Democrat or a Republican, the period leading up to each election enjoys more economic stimulus than after the voters have picked the winner. So we tested this idea to see if there were any trading gems to be gleaned.

PRESIDENTIAL PLAYOFF
We first tested this idea on the Dow Jones Industrial Average (DJIA) a few years ago. Even though we expected lopsided results thanks to the habit of incumbent governments to put the voter in the best possible mood before each election, we were surprised by what we discovered.

We based our trading system on a composite chart put together by thechartstore.com of 19 election cycles from 1928 to 2004 that showed a definitive DJIA cycle low at the end of September each midterm year and a high at the end of December of each election year 26 months later. Our backtests compared DJIA returns for the pre-election versus post-election period between 1902 and 2006 in a total of 104 cycles.

If there was little or no election effect on the markets, returns for each period should have had roughly equal results. In our case, a neutral result would have been 56:44 (26 months versus 22 months) for a ratio of 1.2:1. It became clear, however, that the two periods were anything but similar (Figure 1).

FIGURE 1: Results of owning for the two years (26 months) leading up to each election vs. the two years (22 months) after compared to the neutral result assuming no election cycle impact. But this was just the beginning.

Next, we wondered if governments had gotten better at prepping markets for each election or if, as some have suggested, the Presidential cycle had lost its clout. How would our hypothetical trader have done trading this system in the last three election cycles? To make our tests more realistic, we employed a system to buy the basket of Dow 30 stocks at the beginning of each period and sell all stocks at the end of the period and include the cost of commissions. For our tests we used the VectorVest Simulator, which allowed us to utilize a complex set of inputs to give us realistic outcomes. Here's what we found.

FIGURE 2: Equity curve for buying the stocks in the Dow 30 at the midterm low (end of September) two years before each US election and selling the average high (end of December) of the election year 26 months later. An investment of $100,000 invested October 2, 1994, would have grown to $335,631.62 for a total return of 235.6% by the end of the test period of October 2, 2006. Chart by www.VectorVest.com Simulator.

BEST TWO YEARS OF THE ELECTION CYCLE
Our hypothetical election trader and buy & hold investor both bought all Dow 30 stocks on the last trading day of September 1994. Our election trader exited on the last trading day of December 1996, at which point both $100,000 portfolios had grown to $165,687.26 after commissions (Figure 2). While our election trader was out of the market over the next 22 months post-election from January 2, 1997, to September 28, 1998, the buy & hold position continued growing to $219,984.

Over the subsequent 26 pre-election months (October 2, 1998, through the end of December 2000), both were in the market. The buy & hold portfolio peaked at $357,903 on January 7, 2000 (Figure 3), and declined to $308,218 by the end of December. Meanwhile, the election system grew to $233,783 before the trader exited December 29, 2000, for the next post-election period.

FIGURE 3: Buy & hold performance from October 1, 1994, through October 2, 2006, showing the big drop from 2000 through 2002.

At this point, the election trader exited while the buy & holder stayed in the market. By the end of the test period on October 2, 2006, the buy & holder had narrowed the gap with a final portfolio value of $321,257.

As well as the fact that the election trader's portfolio was bigger at the end of the testing period, there were three major differences between the two approaches:

1 Risk exposure. While the buy & hold investor had his money in the market and exposed to risk for all 4,383 trading days in 13 years, the election trader was at risk for 2,463 days or 56% of the time. So in a little more than half the time and including the cost of commissions the election trader made more money.
2 Rates of return for actual time in trade. Over the 13 years, the buy & holder enjoyed a respectable 17% per annum. However, the election trader earned a more enviable 41.3% while invested.
3 The difference in drawdowns. This was the most important. The buy & holder suffered a crippling maximum drawdown of 45.6% during the meltdown from 2001 to 2002, while the election trader's worst drawdown was a much more manageable 17.4%.

We then tested the inverse election trading system in which the trader bought post-election and sold at the midterm low 22 months later. This system earned a grand total of $1,171 for a return over the period of 1.2% after commissions.

BEST YEAR OF THE ELECTION CYCLE
Next we looked at the best year to be in the market. From 1902 through 2006, the pre-election year was responsible for 60% of the DJIA gains. How would the hypothetical trader have done between 1994 and 2006 by only being invested in the pre-election calendar years of 1995, 1999, and 2003?

As we see from our results, buying the Dow 30 stocks for the calendar 12 months in pre-election years returned 104% in three cycles (see Figure 4). This was significantly better (42.1% more) than the trader would have earned by buying & holding in each of the other three years (total 61.6%) in the election cycle combined!

FIGURE 4: By far the best-performing year of the election cycle was the pre-election year, which returned an impressive 104% over the period ($100,000 to $203,665.51), which worked out to 47% of the total buy & hold gain over the period. The maximum drawdown in pre-election years was 15.6%.

The road to trading hell is littered with the broken dreams of those who made big profits but then lost them thanks to unwieldy drawdowns. Most of us have learned that a 50% loss requires a 100% win just to get even again. Keeping losses contained makes it easier to get back into the black when the rally resumes.

Traders who win consistently don't necessarily earn the biggest profits; they are able to hold on to a larger percentage of what they earn because they minimize losses. Survival in this labyrinthine jungle is all about risk management.

Staying out of the market during the two years post-election won't guarantee you won't lose money, but this work shows that average losses should be smaller.

Trading the pre-election period does not involve a large number of trades and it is a simple system to employ. Perhaps best of all, your maximum drawdowns should also be more manageable, leaving you with a lot more cash to play the pre-election party and greatly increase your chances of garnering you your fair share of the political bounty that begins to surge through Wall Street every four years.

The VectorVest Simulator background information http://www.vectorvest.com/products/simulator.htm