February 3, 2015
The Calendar Doesn’t Lie… Or Does it?
Please read important disclosures HERE.
February 3, 2015
Please read important disclosures HERE.
There are few things that are consistently predicable in the world of investing. The media that covers investing, however, is extremely predictable. If you spend any time reading the financial press, you know that you will stumble on an article about the ‘January Effect’ early in the New Year, shortly followed by a story on the ‘January Barometer’ in early February. In the spring, you undoubtedly will run across an article about whether it makes sense to ‘Sell in May and Go Away’ only to be followed up by other stories about seasonal tendencies in the stock market later in the year.
The point of all of these articles is the same. In each case, they point out research that shows consistent patterns of stock prices for specific periods during the year. Financial academics call these periods ‘calendar anomalies’ because they believe that seasonality in stock prices should not occur in an efficient market. But do these anomalies exist? And if so, should an investor try to take advantage of them?
The January Effect – Here Today, Gone Tomorrow
One problem with trying to take advantage of historical seasonal patterns is that there is no guarantee that any pattern will continue in the future. In fact, if a pattern exists, the outperformance will usually not persist for long.
Nowhere has this been more evident than the earliest incarnation of the January Effect. The January Effect was first discovered in the 1940’s1 and was simply the observation that January historically had the best stock market returns of any calendar month. Researchers posited that the reason for the outperformance was the reinvestment of tax related sales in December and the investment of year-end bonuses in January.
As investors began to anticipate higher returns in January, they began buying in December to take advantage. This led to reduced returns in January and higher returns in December which basically negated the January Effect. In fact, ask a trader about the January Effect and they will jokingly tell you it now starts on Halloween! As you can see from the chart below, January is now the fifth best performing month when looking at S&P 500 Index returns.
So Goes January, So Goes the Year – The January Barometer2
The January Barometer is a theory stating that the return of the stock market in January determines the return for the rest of the year. For example, if you experience positive returns in January, you should expect positive returns for the remaining 11 months of the year. The Barometer is frequently mentioned when returns for January are negative as some people believe it is a sign to expect poor market returns for the year.
Using S&P 500 Index data from 1927 through 2014, this theory has been correct approximately 66% of the time. The problem with the strategy is its poor track record for predicting years with negative returns. For example, the data show that when January’s return was positive, returns for the rest of the year were positive 80% of the time. But when January’s return was negative, returns for the rest of the year were negative only 42% of the time. In fact, January has experienced a negative return for three of the last six years yet each time the market ended the year higher by more than 10%.
Sell In May and Go Astray3
‘Sell in May and Go Away’ (SMGA) is a well-known adage that is written about every spring. The strategy originated from the Stock Trader’s Almanac which noted that since 1950, the average returns on the Dow Jones Industrial Average (DJIA) are significantly better for the period from November to April than May through October . The recommended strategy is (not surprisingly) to reduce or sell stocks in May and then purchase them back in November.
The first problem with implementing a strategy like this is the transaction costs associated with following the strategy. The benefit from avoiding May-October needs to be greater than the costs of selling and then repurchasing your entire portfolio every year. Tax considerations in taxable accounts are another significant hurdle. But let’s assume for a moment that there are no additional costs or taxes for following the strategy. If you pursued this strategy, you would sell your stocks in May and re-invest the proceeds in a money market or similar cash equivalent. In order for this strategy to work, the return on your money market account from May – October would have to exceed the return you could have earned had you remained invested in stocks over this same six-month time period.
The table below shows the sub-period returns for the S&P 500 and the Russell 3000 Indexes. The S&P 500 Index is a market cap weighted index that represents the 500 largest US companies. The Russell 3000 Index is even broader and includes the 3000 largest US companies. Both indexes are significantly more representative of the US stock market than the DJIA, which only has 30 US companies.
As you can see, both of these broader indexes have experienced consistently positive returns from May through October. For comparison purposes, the 3-month T-Bill yielded an average of approximately 2% during the periods analyzed below. This suggests most stock investors would be better off not following the SMGA strategy, particularly after transaction costs and taxes are considered. In addition, the frequency of positive returns for the May-October period is not significantly different than the November-April period for these indexes.
Six-Month Index Returns
Although calendar-based anomalies can be nice headlines for the financial press, investors are better off focusing on maintaining a broadly diversified portfolio that matches their risk tolerance. Any anomaly that does exist may ultimately disappear and the actual results of the strategies may be much worse than advertised.
1Sidney B. Wachtel, 1942 ‘Certain Observations on Seasonal Movements in Stock Prices’ , Journal of Business, University of Chicago Press
2Source: BOS Research using data from DFA Returns 2.0