PRESIDENTIAL CYCLE INVESTING
After spending time going through the monthly seasonality strategies presented in the last chapter, you probably thought that’s about all you can squeeze out of the calendar. Guess again. There are even more profitable strategies in this chapter. If you are interested in pursuing seasonal strategies as a trader or investor, be sure to get a copy of Stock Trader’s Almanac 2011 (Wiley, 2011) at Amazon.com or your local book store. This book is updated yearly, so make sure to buy the latest edition.
Vakkur Tests Optimal Months Combined with Optimal Presidential Cycle Years
Mark Vakkur made another significant research contribution by publishing another seminal article in the October 1996 issue of Technical Analysis of Stocks & Commodities.1 In this nine-page article, Vakkur combined the optimal months with the optimal presidential cycle years. The results were outstanding, so let’s get right to them. The election cycle years are:
- Pre-election year. The year before the election year (e.g., 1987, 1991, 1995, 1999, 2003, 2007, and 2011).
- Election year. The year of the election (e.g., 1988, 1992, 1996, 2000, 2004, 2008, and 2012).
- Post-election year. The year after the election year (e.g., 1989, 1993, 1997, 2001, 2005, 2009, and 2013).
- Midterm election year. Two years after the last election year and also prior to the next election year (e.g., 1990, 1994, 1998, 2002, 2006, 2010, and 2014).
Historically, there is a four-year cycle in the stock market encompassing the presidential election years. Usually, the market rises more in the pre-election year than in the election year itself. The election year is the second best of those four years. And in the two years after the election, the market usually does not make much progress. Table 8-1 contains the data from 1950 through 1995 delineating the average monthly returns of the Standard & Poor’s (S&P) 500 Index in each of the four years of the presidential election cycle as compiled by Vakkur.
As the table indicates, the pre-election years’ stock market returns are by far superior to any of the other three years of the presidential election cycle. The average annual return since 1950 was 18.72 percent, and the market rose every pre-election year. The next-best-performing year was the election year, with an average annual return of 10.08 percent and a 91 percent success ratio.
Conversely, the worst-performing presidential cycle year was the post-election year, with an average annual return of only 2 percent and with the markets rising only 45 percent of the time in those years. Next-worst performance was the midterm year, which sported an average annual return of a meager 4.63 percent, with a positive market in 55 percent of the years.
Freeburg Tests Presidential Cycle Back to 1886
Nelson Freeburg tested the quadrennial presidential cycle back to 1886 through 2001 to double-check and expand on the data tested by Vakkur. In the May 31, 2002, issue of Formula Research, Freeburg provided the statistics for the Dow Jones Industrial Average (DJIA; see Table 8-2).2 Keep in mind that Vakkur used the S&P 500 Index (without dividends) in calculating his numbers (nominal return), whereas Freeburg used the DJIA (without dividends) because that index had a much longer price history available. Clearly, there has been a definite bias toward higher stock market returns in the preelection and election years during the 45 years observed in Vakkur’s analysis, as well as in Freeburg’s more up-to-date and comprehensive analysis.
How did the electoral cycle work out in recent years? Actually, it was a mixed bag. The election cycle performance was as follows:
- 2002, 2006, and 2010 (midterm): –16.8%, +16.3%, +6.8% (October 22, 2010 price)
- 2003 and 2007 (pre-election): +25.3%, +6.4%
- 2004 and 2008 (election): +3.1%, –33.8%
- 2005 and 2009 (post-election): +3.1%, +18.8%
Vakkur’s Leveraged Strategies Earn Millions
Table 8-3 shows how an investor fared by taking advantage of Vakkur’s several strategies from 1950 to 1995.3 First, let’s get the lowdown on the returns from buy-and-hold. This often-touted simple, no-decision strategy had an average annual return of 8.4 percent that turned a $10,000 initial investment into $372,388 with a 12-month maximum drawdown (or maximum loss) of 41.3 percent during the test period. This means that the buy-and-hold investor sat through this loss without flinching. In reality, the average investor would have a hard time doing this.
Now compare those results with the simple strategy of investing only during the best-performing two presidential cycle years— the pre-election and election years—and remaining in cash for the other two years. This strategy has an average annual return of 10.0 percent, which is 1.6 percentage points a year better than buy-andhold. However, over 45 years, that incremental difference resulted in growth of the $10,000 investment to $733,605 compared with $372,388 for buy-and-hold. This translates into a 97 percent improvement over buy-and-hold, with a standard deviation of only 8.3 percent versus 14 percent for buy-and-hold, and a maximum 12- month drawdown of only 18.7 percent. Thus, on all counts, this presidential cycle strategy provided superior results with 50 percent less risk because the investor was invested for only two years out of four. Higher return with less risk is always a win-win strategy.
You may be curious to see the results of investing only in the worst two performing presidential cycle years—the post-election and midterm election years. As expected, the performance results were much worse than buy-and-hold, with only a 3.9 percent average annual return and the same annual maximum drawdown of –41.3 percent. Even the lowly 100 percent cash portfolio (not investing at all) beat these two years’ performance by generating an average annual return of 5.5 percent with no negative drawdown. When cash is able to beat buy-and-hold (for the worst two presidential cycle years), you know that you should avoid investing at all. Vakkur tested a number of different strategies to determinetheir profitability. The first strategy used was with 2:1 leveraging. This strategy used 50 percent margin in the pre-election year, invested funds with no margin in the election year, and was 100 percent in cash for the other two years. As expected, this strategy increased the return by 3.41 percent over buy-and-hold.
With an average annual return of 11.4 percent and an ending balance of $1,272,369, this leveraged strategy provided an additional return of $538,764 over the nonleveraged pre-election and election year strategy. Using this more aggressive strategy also resulted in more than doubling the maximum annual drawdown (largest loss incurred) to –46 percent from –18.7 percent in the prior strategy. Also, the maximum annual gain jumped about 100 percent, to 86.3 percent from 43.4 percent. And the standard deviation (variability around the 11.4 percent return) rose to 14.8 percent from 8.3 percent.
As Milton Friedman, the Noble Laureate economist, said, “There is no such thing as a free lunch.” And, as the saying goes, “No pain, no gain.” As mentioned earlier, risk and reward go hand in hand. Investors who have a higher risk tolerance may want to consider using this leveraged strategy. However, they should use a stop-loss limit order to minimize the risk of a large intrayear drawdown and avoid getting clobbered when the market goes against them. And leveraged exchanged-traded funds (ETFs) can be used instead of using margin and paying for margin interest.
Vakkur tested another leveraged strategy called the leveraged best months (LBM) strategy. Here, the best-performing months (November to April) of the year were leveraged 2 to 1 (50 percent margin) in the pre-election and election years, 100 percent was invested during these same months in the post-election and midterm election years, and 100 percent was invested during the May–October period during the same two years. The only time that there would be a 100 percent cash allocation would be in the May–October post-election and midterm election years.
This strategy really hit pay dirt, with an annual return of 12.7 percent and total ending principal of $2,183,257. Moreover, both the maximum drawdown and the maximum yearly gain were reduced. And the standard deviation moved up only slightly to 15.2 percent from that of the previously reviewed 2:1 leveraged strategy of 14.8 percent. On a risk-adjusted basis, the LBM strategy is the best tested so far. This strategy was less risky than the 2:1 leveraged strategy while providing an additional $910,888, quite an astounding difference in the bottom line.
Before reviewing another highly profitable leveraged strategy that more than doubled the LBM returns, let’s look at Table 8-4 showing the monthly returns over 45 years in each of the four election cycle years. The best seven months in the pre-election and election years were January, February, March, April, July, November, and December. The worst months in all the four election cycle years were May and September. Two more worst months occurred in June and August of the post-election and midterm election years.
All the other months not mentioned—January, February, March, April, July, October, November, and December in the post-election/ midterm years and June and August in the pre-election years—were considered average months.
Using this information, Vakkur formulated another highly profitable strategy that he referred to as the optimal months strategy. Look at the footnotes in Table 8-3, which provide the exact optimal months strategy used. That strategy produced an annual return of 14.9 percent, or an astonishing $5,189,384, over the 45 years tested. With only a slightly higher 0.3 percent standard deviation than the LBM strategy and virtually equivalent drawdowns, this strategy added another $3 million to the bottom line. Also, on a risk-adjusted basis, this strategy had the highest annual return of all the strategies illustrated, at 13.7 percent.
Freeburg also tested the strategy of investing in the market (using the DJIA) only in the pre-election and election years from 1886 to 2001 and going into cash in the other two years.4 He found that this strategy resulted in an annual return of 5.9 percent compared with 5.2 percent for buy-and-hold. This 70-basis-point difference on a $10,000 initial investment was valued at $8.4 million at the end of the test period compared with $3.7 million for buy-andhold. The opposite strategy of investing in the two weakest performing years and going to cash during the two strong years since 1886 has returned 3.6 percent per year, worth $607,000. This is 93 percent less than the optimal strategy.
Freeburg ran additional leveraged tests of the data using the strongest months in the strongest years, and they confirmed Vakkur’s results. Since Vakkur analyzed the data through 1995, the question is how has the four-year presidential cycle performed since then (in particular, using the strongest months in the strongest years)? Freeburg, who updated the data from 1995 through April 2002, found that using Vakkur’s methodology produced an annual gain of 17.4 percent per year compared with 10.7 percent for buy-and-hold with the S&P 500 Index. Even since 1886, use of the same methodology has given an annual return of 8.3 percent compared with 5.2 percent for the DJIA.
In conclusion, focusing on the optimal months in the presidential election cycle years really brings home the bacon. This strategy is far superior to just investing during the entire pre-election and election years with outstanding return and risk parameters.
MIDTERM TO ELECTION YEAR PHENOMENA
Over the years, a number of analysts have scrutinized the presidential cycle years to look for patterns. One very consistent pattern that has been uncovered occurs from the lows of the midterm election cycle year to the highs of the following election year. These results are shown in Table 8-5, which was prepared by the Hays Advisory Group.5 Of the 23 midterm years until the election years, 22 have produced gains. Although the S&P 500 Index gains have ranged from 18.4 to 121.9 percent, the average gain has been 55 percent. Excluding the only loss of 43.6 percent from 1930 to 1932, the average gain for the period has been 59.44 percent.
Table 8-6, also provided by the Hays Advisory Group, provides the S&P 500 Index data not only for the midterm year but also for the other surrounding years.6 Just examining the midterm year November to April performance since 1950, we see that in every instance there was a positive return. The mean return was 16.4 percent, with the lowest return being 3.52 percent in 2002 and the highest being 24.86 percent in 1970. The next-best year was the pre-election year, with a mean return of 5.1 percent, with three years of small negative returns.
The election year November–April period returned an average 2.8 percent with four negative years. Lastly, the post-election year performance was the third best, with a mean return of 3.7 percent with seven down periods.
MINTON’S PRESIDENTIAL CYCLE STRATEGY
Dr. Jerry Minton is president of Alpha Investment Management, a money management firm using asset-allocation strategies that are
based on long-term seasonal factors. The research in this section was published under the title, “The Election Cycle: How to Use the Political Class for Investment Gains.” The complete research paper can be found on his Web site www.alphaim.net (select the “Alpha Research” tab, then click on “Recent Articles”).
Since 1933, the DJIA (with dividends) has advanced an average of 17.1 percent during every pre-election year (year three of the presidential term), about triple the average return of the other three years. Every bear market with the exception of the 1987 collapse has occurred outside the pre-election year over this 78-year period. As you can see in Table 8-7, the best quarters of the presidential election cycle years are the fourth quarter of the midterm year and all the quarters of the pre-election year.
Table 8-8 provides the percentage return during this consecutive 15-month cycle period (dubbed the “power zone” by Minton) since 1934, showing an average return of 25.6 percent, which is an annualized return of 19.9 percent without dividends.
Over the past 50 years (Table 8-9), the performance of the S&P 500 over the power-zone periods has averaged 30.3 percent.
In comparison, Table 8-10 shows the Nasdaq’s remarkable performance of the power-zone periods since 1962. Clearly, the
average performance of 47 percent is substantial by any measure. Moreover, there was only one losing period of –5.9 percent in 1986–1987.
Minton’s Strategy Description
The strategy is to invest in the stock market only during every power-zone period, as described previously, with the portfolio split evenly between the S&P 500 Index and the Nasdaq 100 Composite. This portfolio contains large-cap stocks with an overweight to growth and technology companies. During the remaining periods, the funds are invested in the Barclay’s Capital 1–3 Year Treasury Index.
Figure 8-1 shows the performance of this approach. Since 1989, $1 invested in the five power-zone periods grew to $19.50. The same dollar, invested continuously in the S&P 500, with dividends, grew to $4.20. Invested just 30 percent of the time in equities, this simple election cycle strategy produced a 20-year average return of 16.1 percent, with no down years. The S&P 500, over the same period, returned 7.43 percent, with dividends, with multiple down years and more volatility.
Over the past decade, the stock market has destroyed capital and reduced purchasing power at an alarming rate. For the 10-year period ending March 31, 2009, the total return of the S&P 500 Index was –26.3 percent. In real, inflation-adjusted terms, the total return was –65.3 percent, a real loss of 6.35 percent per year. To recover this loss will require about a 200 percent real increase. Minton indicated that he has no doubt that the recovery period will exceed 10 years, which means at least a 20-year dead zone for the market. Pity the poor indexer who staked it all on the market in the late 1990s.
If the past decade has taught us anything, it is that the doctrine of being continuously invested in the market is bankrupt. Any investment strategy that results in a 65 percent real loss over a decade cannot be a sound choice for serious investors. It certainly cannot be a strategy for retirees who are drawing down their savings constantly.
The simple election cycle strategy described in this section is a good alternative to the conventional buy-and-hold doctrine. By accepting market risk only during the power-zone period, the entire political establishment is on our side. This is a powerful form of portfolio insurance. It shows up in the real-return comparisons for the past decade. Look at Figure 8-2 to see the exceptional return of this strategy compared with buy-and-hold with the S&P 500 Index.
Based on the research by Freeburg, Vakkur, the Hays Advisory Group, and Dr. Jerry Minton,7 the intelligent investor should carefully consider the potential financial benefits of combining the strongest seasonal monthly and presidential cycle years.
With seasonality investing, there is always something happening, either on the long or the short side of the market, and you can invest on either side of the market if you know what you are doing. Buy-and-hold is certainly not a recommended strategy in any circumstance, especially in light of the ability to use calendar-based investing strategies to limit losses during the weak presidential cycle years. The recommended investment vehicles to use for both the seasonal and presidential cycle strategies are either index funds or regular ETFs (such as the SPY and QQQQs), as well as leveraged ETFs (such as QLD and QID). Of course, you will be paying capital gains taxes if you are investing in nonretirement funds, and you won’t be if you are investing your retirement funds. Therefore, these strategies are particularly attractive for your retirement accounts, where your profits compound year over year tax-free.