History is replete with strategies that worked for a time, then became ineffective, and were subsequently abandoned.
Examples can be found in sales and marketing, product design, and in all facets of entertainment from TV shows to stand-up comedy. But the easiest place to understand the phenomenon of once-dominant strategies losing their effectiveness might be in military battlefield tactics.
The bow and arrow, plus horses, were used in battles from ancient times and no significant technological advances came into play until gunpowder was first employed in the 14th century.
So, for thousands of years, military tactics (rather than technology) were of paramount importance on the battlefield. One of the most famous, effective, and enduring formations was the Greek phalanx. In this formation, soldiers packed densely together with interlocking shields and used long spears. The strength of the phalanx lies in its unified and seemingly impenetrable front.
Many variations were made to the original rectangular formation, and the phalanx as a dominant military formation reached its peak effectiveness under the command of Alexander of Macedon (better known as Alexander the Great). His use of unbalanced phalanx formations, along with his ability to move troops quickly and brilliant field leadership coalesced to bring this military strategy to its zenith.
Then, following closely behind the death of Alexander, was the death of the phalanx. The phalanx became ineffective in the face of the faster and more flexible three-line Roman Legion. And the Roman conquest of the Greek city-states, and the rest of the known world, followed.
Like the rise and fall of the phalanx as a military strategy, the famous January Barometer has come to the end of its useful life. Just as the agile Roman Legion overwhelmed the phalanx, quick and volatile markets have killed the January Barometer as well.
In its 51st year, the January Barometer has lost its predictive powers, as we shall explore below. But first, let’s admit that there are…
Too Many January Indicators
With so many different indicators flying around that have the word “January” in their name, let me clarify which one we’ll be looking at today.
- The January Barometer was devised in 1972 by Yale Hirsch of Stock Trader’s Almanac fame. So, this is the indicator’s 51st anniversary. It states that as the S&P 500 goes in January, so goes the year. We’ll look at some statistics for this once-impressive indicator below. And bid it a farewell…
- The First Five Days of January. This indicator (discussed just last week) holds that the first five trading days of the year predict the market action for the rest of the year. We busted that myth as well.
- The January Effect. This one is not an indicator but an observed phenomenon that small cap stocks outperform the big caps in January. We described the merits of this observation back in an article in December and noted that the effect still exists, but has moved to an earlier timeframe—beginning in the second half of December and lasting until the middle of February.
The January Barometer: Many Discuss It, Few Understand It. Don’t Trade It!
In the coming weeks, as we head into the end of January, pundits typically write about the January Barometer in glowing terms. Don’t be sucked in by the hype.
On the surface, the January Barometer has had an impressive track record with an 83.3% accuracy claim over the past 71 years. That’s the number you’ll hear quoted over and over again. It is calculated quite simply: If January presents a positive return for the S&P 500, the January Barometer predicts that the full calendar year will be positive, and vice versa for a down January.
The indicator makes an exception for any year that the market moves less than 5% in either direction and calls that year neutral or flat. The flat year is then not counted for or against the track record (which is a reasonable exception). Given that caveat, the January Barometer has only been “wrong” 12 times since 1951. The January Barometer used to do a credible job of predicting up and down years. For the last 2+ decades? Not so much…
If we break down the January Barometer signals into decades to see how many times out of 10 years it is correct for a given decade, we start to see a problem. Let’s look at how many misses (an up January as part of a down year or vice versa) and how many flat years there have been per decade. Once we subtract misses and flat years from 10, the remaining years in the decade are the ones that the indicator can claim to be helpful:
As you can see, the usefulness – or efficacy – of the indicator has dropped off drastically in the last two decades. In fact, the indicator has only been useful in nine years out of the last 22. Perhaps this is not a statistically significant sample, but it does raise the question as to whether a once interesting indicator has any effectiveness in anticipating market direction, especially since I’d propose that more recent decades are the most relevant. As we shall see, the problem is even deeper. But we do need to pause and mourn the fact that a small down January in 2021 (-1.1%) would have had you flat the market for the rest of the year; making the indicator miss out on a +28.3% up move for February through December!
Perhaps there has been a change in the investment behaviors that drive markets and the tax-selling of stocks in December – that turns into January buying – is no longer a significant factor. But regardless of whether underlying investing practices have changed, the structural nature of the markets has changed. The markets have become more volatile and large-scale direction changes occur more quickly and go further, starting in the late 1990s. Let’s dig one level deeper into the logic flaw that makes the January Barometer a broken indicator.
The January Barometer’s Flawed Logic
The Stock Trader’s Almanac judges the January Barometer by using calendar year results, when those gains (or losses) already contain the January returns. It’s a bit like saying that a football team that outscores its opponent in the first quarter is likely to win the game (or vice versa).
If you used the strategy to trade each year, you would have to buy on the first trading day of February and exit on the last day of December, sitting out the month of January each year. That strategy, over the last 22 years, results in a net loss of -5.99% per year (even though 2008 returned a whopping 40.1% return!).
The January Barometer may have had a good run “back in the day”, but in today’s faster news cycle-driven market, it’s not even useful for cocktail talk.
I’d love to hear your thoughts and feedback. Just send an email to drbarton “at” vantharp.com.
Great trading and God bless you,
D. R.