I've been chipping away at articles I've stored in my Pocket and just made it through a series of pieces on the "Sell in May" phenomenon.
If you're not familiar, the proclamation "Sell in May and Go Away" speaks to the perceived seasonal trend where the US stock market tends to be weaker in the summer-to-fall period than the winter-to-spring period.
The phrase is actually part of the longer and somewhat forgotten "Sell in May and Go Away, Buy Again on St. Leger's Day", referring to the horse race that happens every September.
Three thoughts on this somewhat controversial yet easy-to-remember slogan.
1) We love catchy phrases for market phenomena, regardless of their effectiveness.
You can bucket "Sell in May" with other catchy phrases such as, "When Santa Claus should fail to call, bears may come to Broad and Wall."
Source: Wikimedia Commons
Most articles you read will include some sort of backtest, designed to demonstrate why a) it works so well that you have to pay attention to it, or b) it clearly never works so anyone that does actually pay attention to it is a total idiot.
There's a problem with any sort of backtest that you'll see, and that brings us to point #2.
2) The start date on any backtest makes a huge difference.
If you start your backtest of the Sell in May phenomenon in 1988, or in 1950, or even 2012 (note to self: not long enough for a backtest of seasonal patterns), you may get dramatically different results.
Some market professionals feel that since we all have access to reams of financial data, we should backtest everything and assume the results are valid. However, you have to remember that when you're talking long-term financial data, there often aren't enough observations to draw results that are statistically significant.
To quote Einstein, "Not everything that counts can be counted, and not everything that can be counted counts."
Whether or not you've seen the pattern recently, you have to believe that its very existence is based on some truth at some point in market history.
As Jeff Hirsch from Stock Traders Almanac explained, "Aside from the technical or empirical evidence, there is the cultural evidence of how we behave as humans, and that's what creates these trends, these market patterns."
“Not everything that counts can be counted, and not everything that can be counted counts.”
Of course, the key question is whether it will work this year, and next year, and in subsequent years. That brings us to #3...
3) Price is king.
Regardless of the history of a seasonal pattern, it's best to wait until you actually see something happen. Simply buying or selling based on what has happened in previous years doesn't sound like a sound investment strategy!
Jeff Hirsch suggests using a trend-following mechanism like MACD to give the "all clear" that a seasonal downtrend is in play.
What's interesting is that in "normal" years, at this point in the cycle we're usually in some sort of uptrend. The corrective period we've seen in 2018 has forced the weekly MACD indicator negative and we're still waiting for a positive signal.
Market slogans aside, I plan to continue to watch the market itself for confirmation of either an upside breakout from this correction or a downside continuation of the selloff.
For now, the trend is decidedly sideways.
RR#6,
Dave
Disclaimer: This blog is for educational purposes only, and should not be construed as financial advice. Please see the Disclaimer page for full details.