Can You Find Superior Stock Returns by Just Checking Your Calendar? (Part 1)

Author's Avatar
Dec 04, 2014
Article's Main Image

Mark Twain was not shy from sharing his wisdom about the stock market. "OCTOBER: This is one of the peculiarly dangerous months to speculate in stocks," Twain said. “The others are July, January, September, April, November, May, March, June, December, August and February.”

Yes, we got it. Every month is dangerous. However, are they equally dangerous? Are they equally good/bad? As I am writing this on the first day of December, it turned out December has been the best month in history. The average of December return is 1.8 percent, the highest of all 12 calendar months. The conclusion is based on monthly data compiled by CRSP (the Center for Research in Security Price at the University of Chicago) from 1927 to 2012, and from Database compiled by Professor Eugene Fama and Kenneth French. Market return is equally weighted return of all stocks in CRSP database.

Based on those data, Mark Twain might want to change his famous quote to: “December: This is one of the best months to invest in stocks. The others are January, July, November, April and August.” The order matters, and all these months have an average return higher than 1 percent. Now here are the bad ones: “September: This is one of the worst months to invest in stocks. The others are October, May, February, March and June”. September is the only month with negative return (-0.7 percent), although other months (on this bad list) are positive, their average return is below 1 percent. The complete list of average monthly return is below. The average of all months is 0.9 percent, equivalent to about 10 percent annually.

Table 1: Average Monthly Returns from 1927 to 2012
Month Market Return (%)
Jan 1.5
Feb 0.5
Mar 0.8
Apr 1.3
May 0.4
Jun 0.9
Jul 1.5
Aug 1.2
Sep -0.7
Oct 0.3
Nov 1.5
Dec 1.8
Average 0.9

Be cautious

Calendar months and seasons are caused by the moon orbiting the earth and earth orbiting the sun. Should our stock market, driven by rational and irrational human behaviors and economic conditions, be independent of how stars and planets are aligned? And if there is a pattern discoverable by simply taking the average, should the pattern be traded away by smart investors, like you?

When you look at such information from an historical average, remember three things: 1) The pattern could simply be caused by chance. Statistically speaking, they might not be significant. 2) The pattern might have been discovered and trading based on these pattern could have destroyed the pattern gradually, and 3) the average is just an average; it is far from a guaranteed result for any single observation (like this coming December).

However, empirical researches have singled out two or three months that are statistically different than other months. December is one of them.

The December Effect

03May20171240501493833250.jpg

December has been the best month. Why?

Some suspect that the December effect is caused by mutual funds/ hedge funds adding winners to their portfolio before they report their year-end holdings to investors. If every fund manager needs to assure their investors that they are not stupid for not loading tons of Apple (AAPL, Financial), Walmart (WMT, Financial), Alibaba (BABA, Financial) or Starbucks (SBUX, Financial), this December would be the last month they can buy those winners (if they haven’t done so earlier) before the show time.

More important, there is tax effect. If investors sell their winners in December, they need to pay capital gains tax. Rational investors would consider at least postponing selling winners to January. The lack of sell orders should help the momentum of the winners to stretch further into December.

Table 2: Average December Monthly Return ( %) of Portfolios Ranked by Return of Previous 11 months, from 1927 to 2012
Portfolio formed by Prior Returns Decile1(Prior Loser) Decile2 Decile3 Decile4 Decile5 Decile6 Decile7 Decile8 Decile9 Decile10(Prior Winner) Ă‚
December Return -1.5 0.0 0.6 0.9 1.4 1.6 1.9 2.3 2.7 3.3 Ă‚

And, yes, December is the most generous for winners. From table 2, when stocks are ranked based on their returns in previous 11 months, stocks of the highest prior returns decile (winners), on average, earn a whopping 3.3 percent in December. As you might have guessed, December is indeed cold for losers. Stocks of the lowest prior returns decile loses -1.5 percent in December. Moving up along the prior return axis, your December return gets better and better, in general.

Table 3: Average December Monthly Return ( %) of Portfolios Ranked by Size (Market Capitalization), from 1927 to 2012
Portfolio formed by Size Decile1 (Smallest) Decile2 Decile3 Decile4 Decile5 Decile6 Decile7 Decile8 Decile9 Decile10 (Biggest)
December Return 0.0 0.9 1.5 1.9 1.7 2.1 2.0 1.9 1.9 1.7

Table 3 shows size (measured by market capitalization) also matters at the first glance. Is it because winners getting bigger from rising price, and losers losing weight? December is one of the worst month for smallest firms. Firms of the smallest decile (the bottom 10 percent by capitalization), on average, earns only 0 percent in December. However, size seems not as important as prior return. The biggest firms (top 10%) earn an average return of 1.7 percent, close to the average of all sizes.

Table 4: Average December Monthly Return ( %) of Winner Portfolios Ranked by Size, from 1927 to 2012
Portfolio formed by Size Quintile 1(Small) Quintile2 Quintile3 Quintile4 Quintile5 (Big)
December Return 2.0 3.0 3.4 3.4 2.8
Ă‚ Ă‚ Ă‚ Ă‚ Ă‚ Ă‚
Table 5: Average December Monthly Return ( %) of Loser Portfolios Ranked by Size, from 1927 to 2012 Ă‚
Portfolio formed by Size Quintile 1(Small) Quintile2 Quintile3 Quintile4 Quintile5 (Big)
December Return -1.2 0.1 0.8 0.6 0.4

Now, when you combine the size effect and winner effect together, size no longer matters in December. For example, when I rank the winners by size in table 4, there is no obvious relationship between size and return in December. When I rank the losers by size in table 5, size does not matter as long as you are not investing the smallest size quintile. (I use quintile instead of decile to make sure I have enough stocks in each portfolio when I slice the market into too many groups.)

Be cautious again

The wonderful December return does fluctuate. It was less pronounced in the 1930s and 1960s. Its absolute value is only 1.1 percent in the 2000s. Given the boring average return of the 2000s (average monthly return is only 0.1 percent), December of 2000s is relatively good. Table 6 shows December returns by decades.

Table6: Average December Return (%) by Decades Ă‚ Ă‚ Ă‚ Ă‚ Ă‚ Ă‚
Decades 30s 40s 50s 60s 70s 80s 90s 2000s
December Return -0.6 2.9 2.7 0.8 2.5 1.0 3.6 1.1
Average Return of All Months 0.5 0.9 1.5 0.7 0.6 1.4 1.5 0.1

Also, the conclusion of this article is based on the data from 1927 to 2012. I intentionally left the data un-updated to see how the pattern works out-of-sample. For December of 2013, S&P 500 has a good return of 2.4 percent. However, considering 2013 is a monster year with an average monthly return of 2.2 percent, last December does not stand out. It ranks only seventh among the 12 months in 2013.

Bottom line

There are actually advantages to being an individual investor. You don’t have conflicts with your own interest. You don’t buy expensive stocks just to show around. You can actually take advantage of professional investors when they make sub-optimal decisions for their customers. In general, buying winners (those that have performed well in the previous 11 months), avoiding losers and very small stocks in December seems to have worked in the past. Professor Honghui Chen and Vijay Singal published a very good article in the Financial Analyst Journal suggesting the December effect is also relatively easy to arbitrage.

If December is related with fund managers buying/holding winners and selling losers, it should have consequences for next month: January. Will the losers bounce back? Will the winners lose momentum?

As you can see, the December effect is somewhat related to the more well-known January effect. Yes, January will be my next topic as the earth approaches it.