You are probably aware that we use charts as the core for our analysis of the markets. Commonly known as technical analysis, there are many forms and combinations of indicators that professionals use in this endeavor. There are many who are skeptical about this strategy to manage portfolios, but it is difficult to argue with results that some of these chartists produce.
We thought it might be interesting today to share with you some of the data points that we use that are not necessarily chart related, but we do integrate them with technical analysis to attempt to manage portfolios more effectively. If you are enamored with the stock market or are an information junkie, “The Stock Traders Almanac” is a must for your book shelf. Most of the information we are about to share comes from that resource.
The most impactful of these stats is the seasonal measure. Going back to 1950 and using the Dow Jones Industrial as the investment index, the outcome of this seasonal strategy is fairly remarkable. The scenario looks like this; you take $10,000 and put it to work in the Dow on May 1st and on October 31st you remove it and put it under your mattress through the end of April. You repeat the process so that every year your funds are exposed to the market only for the six months of May through October. Fifty seven years later your $10,000 would have grown to a whopping $11,021! You made just north of a grand…for 57 years! When we were twelve we had a paper route for six months that made more than that.
What is equally amazing is if you reversed the six month process and invested in the Dow for the period November to the end of April and used the mattress for the May-October time frame. The net result there was a total of $531,444! When we are giving a presentation this stat invariably causes the audience to stop the presentation and have us repeat to them what we had just shared. After confirming the numbers they request the source and ask how they can verify for themselves this anomaly.
What staggers them further is when the indicator MACD is employed with a simple crossing strategy around the four dates used to enter and exit the Dow. The numbers as you can see on the table are even more dramatic. You actually lost over 60% of your money in the May-October time frame. This is the origin of the statement “Sell in May and go away.” Now you can ask macd histogram the person who confidently tells you “No one can time the market” to explain this seasonal phenomenon. The answers we have gotten in the past from those who make that statement reveal to us more their level of research and credibility than the merits of the data. It often amazes me how we find some people would rather be “right” than make money
The next stat we find almost as interesting is the first day of the month phenomenon, especially of late. Going back to September 2nd of 1997 and ending in May 1st of last year the Dow Jones Industrials netted 5,388 points during the period. The first days of the month during that period gained 5,193 points, over 96% of the gains! Using this stat the first day of the month averaged about 40 points while the rest of the days of the month averaged .08 points. This is a dynamic we employ regularly in our work. Month end contributions to retirement plans are a reason many give to explain this imbalance.
Last week we mentioned in our missive one of the reasons for my current concern with the stock market is that the December low for the Dow was taken out in January. Going back to 1952 the Dow’s December low was taken out in the subsequent year’s first quarter thirty-one times. Twenty-nine of those occurrences led to a continued correction for the Dow that averaged 9.8%. You have to admit those percentages/probabilities are worth paying attention to, especially at this time.
The last stat we use is the market action around option expiration. This is a dynamic we have noticed just from paying attention to the stock market. What we have found is that often the market action the week before option expiration, the third Friday of every month, is inverse to the action the week of option expiration. So if the market is lower the week prior to option expiration, the week of option expiration tends to end higher than the Friday before and vice versa. We are actually compiling the stats around this thesis and will let you know if it has any empirical validity in the future.
We find these type of statistics fascinating and attempt to use them on an ongoing basis to improve our performance. What we have found is that most mutual fund and managed account managers are not aware of these dynamics. The few that we have found are aware cannot use them because of their always fully invested charter. We have noticed that even when it is not part of their charter they tend to choose a consistent fully invested position. We think it is obvious, and hopefully you see too, that having the flexibility to act on these data points is a decided advantage when making portfolio decisions.
Did You Know…
The S&P 500 had 72 trading days in 2008 (i.e., 2 out of every 7 days) where the stock index finished with a gain or a loss of at least 2% (i.e., change in the value of the raw index over consecutive trading days), the greatest number of “2% change” days in any calendar year in the last 50 years. The second most “2% change” days (52 such days) took place in 2002. In 2004 and 2005 there were no trading days when the index moved by at least 2% on any single day (source: BTN Research).