A Look at the Seasonality Index

I was born and raised in a (very) small town in rural Wisconsin, and never had much access to, or need for, the "finer" things. Something that's carried with me through the years is a deep appreciation for the base things in life.

Much wiser men than me discovered similar paths to happiness and success, and expounded on the benefits of ignoring the noisiness of life. "Keeping it simple" was a formula for success hundreds of years ago, and I've found that it's just about right for my nature, too.

There are an innumerable number of ways to complicate our personal and professional lives, but I keep finding that the simplest solutions are often the best. So today we're going to take a simple look at an effective way to forecast the likelihood of equity market gains or losses in the form of the Seasonality Index.

What is it?

The in-depth analysis of seasonal factors in the trends of stock prices has been going on for decades, with the most well-known proponents being the Hirsch's of Stock Trader's Almanac fame. Others have taken the concepts and gone much further with it, drilling down to sector- and stock-specific tendencies with success. See some of Phil Erlanger's excellent work for examples.

In the December 2002 issue of Technical Analysis of Stocks & Commodities, Jay Kaeppel presented an alternate method for scoring each day's tendency to rise or fall. He based the score on four seasonal factors, those being the end-of-month phenomenon, holiday biases, the "sell in May and go away" theory and the Presidential election cycle.

If none of those factors are positive for a particular day, then it gets a score of "0"; if all four are positive, then that day gets a "4". The score can range anywhere between those two boundaries.

Why should we follow it?

The Seasonality Index is simple but effective, and we have the ability to go back and test its efficacy through different market environments. While not perfect of course, I have found that over the past 56 years, using the Index as a general guide would have helped with identifying those periods of low and high risk.

As one example, if we had been long the S&P 500 on any day that scored a "0", we would have showed a gain 50% of the time, with an overall average return of -0.1%. If we had been invested on a "3" day, however, we would have had a gain 62% of the time with an average return of +0.23%. That's over a very, very large sample.

Since 1950, if we were out of the market on all "0" days, and margined ourselves 2-to-1 long on any "3" or "4" day, then $10,000 would have grown to over $13,000,000 by 2005, beating a buy-and-hold return many times over, and with less risk.

The trading system above ignored commissions, which certainly would have eaten a good chunk of the returns. But that's not the point – the point is that for the most part, we very much have the wind at our backs on the long side when the Seasonality Index is high, and quite the opposite when it is not.

To try to get a handle on what's coming up, I sum up the Seasonality Index score for each of the next 21 days. That's what's depicted in the chart below – high readings indicate a good probability that we'll see a rising market or a bottom form, while low readings suggest that we're going to be in for a spot of trouble.

What are the challenges in using it?

As some of these tendencies get broad public acceptance, there is a chance that more people will anticipate and act on them, thus either destroying their value or shifting the pattern.

We've seen that to some degree with the end-of-month pattern, which is no surprise since there are some foreign and even U.S. funds that use that pattern exclusively as a basis for their trades.

It can also be difficult to implement depending on your time frame. For short-term traders, the utility is clear – be more willing to go short on a "0" day, and press any long-side setups aggressively on a "3" or "4" day. For those with a very long-term time frame, however, these factors can be moot except for perhaps fine-tuning entries and exits.

What does it look like?

What's it suggesting now?

You can see from the chart that the Index did a pretty good job at highlighting the difficult times that were about to hit in 2001 and 2002, then the likely low that was going to form in late 2002. It helpfully suggested being heavily long in late 2003, and being more defensive around the middle of 2004.

Where the price chart of the S&P ends is where we are now. We can see that the Index is going to scrape along the lowest levels that it gets for the next couple of months before turning positive later this year.

"Turning positive" is putting it mildly – according to the Seasonality Index, we will have the most positive market environment possible as we approach the fall, and historically it has been rare to see any meaningful weakness in stocks when this is the case. But in order to get there, we might have some more trouble.

I do not trade this Index mechanically and would not recommend that anyone else do so, either. But as a way to get a general handle on the market's prospects going forward that is different from our fundamental and technical analysis, this is a good tool to investigate.

A Look at the Seasonality Index

You mean I can come back in October?!

WHAT'S POPULAR IN THE VILLE

A Look at the Seasonality Index

You mean I can come back in October?!

Stay Connected

Stay Connected

WHAT'S POPULAR IN THE VILLE