Many global markets have made a strong start to 2019. At the time of writing the S&P 500 is up over 15% for 2019 so far. That’s a good outcome compared to history. It represents the S&P 500’s best Q1 performance since 1998. So 2019 for the stock market is off to a great start, relatively speaking.
In part, this is because the lows due to recent growth and interest rate concerns occurred just a week before 2018 ended, so calendar year 2019 has mapped almost perfectly to the market’s rebound. However, the old adage of ‘sell in May’ has a strong track record and the idea itself goes back decades, so, should you take follow the rule, sometimes called the Halloween indicator or Halloween rule, in 2019?
What ‘Sell In May’ Means
There is some confusion over what ‘sell in May and go away’ means. It does not only apply to the month of May. It means a strategy of moving to bonds from May to October, returning the to stock market from November to April. So essentially, when following the strategy, you are out of the stock market and in bonds during the summer months, and fully invested over the winter, including some of spring and fall.
Is sell in May just meaningless market folklore, or is that something behind it? Perhaps surprisingly the historic data is robust. Let’s start with how we should examine these sort of market phenomenon. If there’s something behind these rules, then they should hold up over a long run of history, over different countries and, ideally we should have a good sense of why they work. With sell in May it can appears we can fortunately check all three boxes.
Let’s start with the historical data. It’s pretty good. Bouman and Jacobsen in their academic research find that the sell in May effect holds in 36 out of 37 countries and over decades, which is to say that returns are, on average, higher for November to April than May to October. Zhang and Jacobsen update the analysis to 2017 here. However, there are two important things to note. Firstly the effect is fairly small, returns to implementing the Halloween strategy are about 1.5% better per year before taxes and fees, on average, than a basic buy and hold plan.
Secondly the Halloween strategy only works about 6 year out of 10. In investing, where gaining any sort of edge is challenging, these are good results. Even though it only works a little better than half the time, that average 1.5% performance improvement is still an annual average. Therefore, the Halloween indicator has the potential to be positive, but not transformative, for a portfolio. Ideally the sell in May rule should be tested over a period of around 7 years or more, since any single year has a relatively high chance of failure, but over a series of a years, a more robust edge can emerge.
There is criticism of the strategy. Some have argued that the strategy doesn’t overcome its trading costs. This may have been true in past decades, but is not true more recently as the costs of trading continue to fall. Trading costs are now extremely low even for retail investors turning over their portfolio twice a year as the strategy would recommend, especially as low-cost ETFs can help you track the market cheaply and various funds are offered commission-free. However, taxes may be a legitimate impediment. This is because gains on the strategy in a taxable account would likely be short-term when following the Halloween indicator, whereas with a buy-and-hold approach you can generally only incur long-term capital gains taxes, which are often taxed a lower rate.
It is also argued that the sell in May effect has disappeared in recent years. Here it is almost certainly too early to tell. Again, history suggests success in about 6 years in 10, so a few bad years for the strategy should not negate it. As with other investment strategies they have seen better and weaker periods and many strategies that have been proclaimed dead, only to subsequently rebound. As recently as 2012 the CFA Institute found the Halloween effect “stronger than ever”.
It is also argued that the effect is too simple and likely the result of data-mining. However, this particular trading strategy has existed going back to the 1930s, long before sophisticated statistical techniques and large datasets for computers to mine through. There is also some evidence that the strategy could be due to the impact of summer vacations on the markets, when portfolio managers are less engaged and delegating to subordinates who may be uncomfortable making big buy decisions. Even if the cause is unclear, its past robustness over time and geography is impressive. Nonetheless, one thing to note is that the ‘sell in May’ effect appears to have more impact on the European markets than in the U.S. based on history.
What To Do?
The evidence for sell in May does appear robust, especially so for many Europe stock markets. However, investors in taxable accounts should be wary of trading strategies that involve holding stocks for less than a year because this can lead to higher capital-gains taxes. If you are in a tax-sheltered investment account such as IRA, 401(k), 457(b) or 403(b), then taking some risk of the table between May and October has clear evidence behind it and the tax disadvantages associated with shorter term trades are unlikely to have an impact. However, given the strategy has historically worked about 6 times out of 10, it is a strategy to consider on a multi-year view because in any single year, the chances of ‘sell in May’ working are fairly close to even. It is over multiple years that the effect of the Halloween indicator has the potential to add up.
Most in the media are touting that the Dow Jones Industrial Average, S&P 500 and the Nasdaq Composite will soon be setting new all-time closing highs. This measure is meaningless as technicians draw trends through all-time intraday highs. It is so misleading in hindsight if an average sets a new closing high then fails to set a new intraday high.
I am not saying that closes are not important. They are used to track moving averages and weekly, monthly, quarterly, semiannual and annual closes are the inputs to my proprietary analytics.
Here’s an important reason for tracking both closing highs and lows and intraday highs and lows. Most think all five major averages set their lows on Christmas Eve. Not so, the Dow and S&P set their intraday lows on December 26. This is true for so many stocks! For almost all tickers that set their intraday lows on December 26 that day was a “key reversal,” which was a signal that a tradeable rally would begin. If you tracked only closing lows, you missed this important buy signal.
All five major equity averages are in bull market territory from their December 14 or Decenber 26 Christmas lows. The Dow Jones Industrial Average is up 21.6% from its December 26 intraday low of 21,712.53 and is just 2% below its all-time intraday high of 26,951.81 set on October 3. The S&P 500 is up 23.9% from its December 26 intraday low of 2,346.58 and is just 1.1% below its all-time intraday high of 2,940.91 set on September 21. The Nasdaq Composite is up 29% from its December 24 intraday low of 6,190.17 and is just 1.8% below its all-time intraday high of 8,133.30 set on August. 30.
To complete the analysis, the Dow Jones Transportation Average is up 26.3% above its December 24 intraday low of 8,636.79 but is 6.1% below its all-time intraday high of 11,623.58 set on September 14. The Russell 2000 is 25.1% above its December 24 intraday low of 1,266.92 but is 9.0% below its all-time intraday high of 1,742.09 set on August 31.
Here’s Last Week’s Scorecard
- The Dow Jones Industrial Average (26,412.30 on April 12) is above its annual pivot at 25.819 and above its 200-day simple moving average of 25,267.37. My monthly and semiannual value levels are 25,513 and 24,340, respectively, with its annual pivot at 25,819 and weekly pivot at 26,125. The all-time intraday high of 26,951.81 was set on October 3 and its second quarter risky level is 27,891. Its weekly chart is positive but overbought.
- The S&P 500 (2,907.41 on April 12) is above its 200-day simple moving average of 2,762.55 and above its annual pivot at 2,867.1. My monthly and semiannual value levels are 2,728.5 and 2,668.8, respectively, with the annual pivot at 2,867.1 and weekly pivot at 2,892.1. The all-time intraday high of 2,940.91 was set on September 21 with its second quarterly risky level at 2,985.1. Its weekly chart is positive but overbought and has become an “inflating parabolic bubble.”
- The Nasdaq Composite (7,984.16 on April 12) remains above its 200-day simple moving average of 7,501.17. My monthly, annual and semiannual value levels are 7,373, 7,370 and 7,274, respectively, with a weekly pivot at 7,972. The all-time intraday high of 8,133.30 was set on August 30 with the quarterly risky level at 8,367. Its weekly chart is positive but overbought and has become an “inflating parabolic bubble.”
- The Dow Transportation Average (10,912.19 on April 12) remains above its 200-day simple moving average of 10,499.16. My monthly and semiannual value levels are 9,858 and 8,858, respectively, with a weekly pivot at 10,433 and annual and quarterly risky levels at 10,976 and 11,372, respectively, with its all-time intraday high of 11,623.58 set on September 14. Its weekly chart is positive but overbought.
- The Russell 2000 (1,584.80 on April 12) is just above its 200-day simple moving average at 1,571.79 and a weekly pivot at 1,553.66. My semiannual and monthly value levels are 1,504.17 and 1,436.97, respectively, with annual and quarterly risky levels at 1,612.54 and 1,667.15, respectively, with its all-time intraday high of 1,742.09 set on August 31. Its weekly chart is positive, but its weekly stochastic reading is just below the overbought threshold of 80.00.
Value levels and risky levels are based upon the last nine weekly, monthly, quarterly, semiannual and annual closes. The first set of levels was based upon the closes on December 31. The original semiannual and annual levels remain in play. The weekly level changes each week; the monthly level was changed at the end of January, February and March. The quarterly level was changed at the end of March. My theory is that nine years of volatility between closes are enough to assume that all possible bullish or bearish events for the stock are factored in.
To capture share price volatility investors should buy on weakness to a value level and reduce holdings on strength to a risky level. A pivot is a value level or risky level that was violated within its time horizon. Pivots act as magnets that have a high probability of being tested again before its time horizon expires.
My choice of using 12x3x3 weekly slow stochastic readings was based upon back-testing many methods of reading share-price momentum with the objective of finding the combination that resulted in the fewest false signals. I did this following the stock market crash of 1987, so I have been happy with the results for more than 30 years. The stochastic reading covers the last 12 weeks of highs, lows and closes for the stock. There is a raw calculation of the differences between the highest high and lowest low versus the closes. These levels are modified to a fast reading and a slow reading and I found that the slow reading worked the best.
The stochastic reading scales between 00.00 and 100.00 with readings above 80.00 considered overbought and readings below 20.00 considered oversold. Recently I noted that stocks tend to peak and decline 10% to 20% and more shortly after a reading rises above 90.00, so I call that an “inflating parabolic bubble” as a bubble always pops. I also call a reading below 10.00 as being “too cheap to ignore.”
Since it’s National Pet Day, Universal/Comcast is finally offering a full-blown theatrical trailer for what will probably be their biggest-grossing movie of the summer. We’ve had months of character-centric mini-teasers, which allowed Universal to sell the Illumination sequel without giving away the story. And to the extent that this trailer is arguably less spoiler-y than the official synopsis, that’s still the case. That being said, the International trailer (released concurrently with the domestic one), is arguably intended to remind you not just of The Secret Life of Pets but also of Sing.
The domestic trailer, which is fine on its own merits, offers the expected mix of comedy, farce and the whole “Here’s what your pets do when you’re not around.” gimmicks. The notion of digging into their psyche is interesting enough, as long as all of the pets aren’t confronting the same issue. That being said, the international trailer (which is slightly plot-heavier) seems to imply that “fear” is the common, well, fear. We hear variations on “fear” or “afraid” so often that I thought I was watching Batman Begins again, but c’est la vie.
The second trailer offers what amounts to a light drama amid the comedy, complete with character arcs sets to The Killers’ “All These Things That I’ve Done.” It’s not an exact match, but I was reminded of the first full trailer for Sing (set to “Without You”), which sold the musical comedy as a more dramatic and more emotional fantasy than the likes of Despicable Me 2 or Secret Life of Pets. That was one of my favorite “big” trailers of 2016, and one that sold the film best among the various pieces of marketing. Watching the international trailer for this Illumination sequel, I’m inclined to think others at Universal agreed with me.
Yes, the international trailer also focuses on the whole “Snowball cosplays as a superhero” thing, but I don’t think Universal thinks audiences are dumb enough to think that it’s going to turn into an actual superhero movie. I’d be curious to see if we see both trailers in both respective marketplaces, as they are offering somewhat different pitches. Neither is “misleading,” (this isn’t the first super-serious teaser for Planes: Fire and Rescue), but frankly I think the melodramatic one better positions the film as somewhat different from its predecessor and perhaps a little meatier than the first Pets romp.
Granted, when the first film opens with $104 million (the biggest opening ever for an original movie) and grosses $875 million global on a $75 million, maybe “same” is good. Yes, this is one of those classic “can afford to earn a lot less than its predecessor and still be a hit” sequels. And with yesterday’s Avengers: Endgame post discussing comparative opening weekends of follow-ups to predecessors that scored super-huge debuts, it will be interesting to see if Secret Life of Pets 2 stays the course, opens a little higher like Hunger Games 2 or opens a little lower like Harry Potter 2.
To be fair, even a comparable drop on opening weekend to Jurassic World: Fallen Kingdom (-29%) or Fate of the Furious (-34%) gets this sequel an opening between $67 million and $73 million. Unless Fast and Furious Presents: Hobbs and Shaw really overperforms, or this sequel takes a big dive from the first film’s $368 million domestic/$875 million worldwide totals, this Illumination sequel will presumably be Universal’s biggest earner of the summer, if not the year. At the very least, we’ll get more cats this time, and that’s not even counting Universal’s Cats this Christmas.
The Brian Lynch-penned and Chris Renaud-directed Secret Life of Pets 2, starring
Louis C.K. Patton Oswalt, Kevin Hart, Ellie Kemper, Eric Stonestreet, Harrison Ford, Lake Bell, Jenny Slate and Tiffany Haddish, opens June 4, 2o19 in North America (and late May elsewhere).