The last ten trading days of the year have contributed about 1% to S&P 500 performance on the average. A backtest also confirms this.
The chart below shows S&P 500 returns for the last 10 trading days of the year.
Since 1941, the index has gained 78% of the time in the last 10 trading days of the year. The average return is 1.08%. This is good performance.
There is an outlier in 1991 of more than 9%. If we replace this outlier with the average value, the new average is 0.98% or a difference of 10 basis points only.
Below is a backtest of buying and holding the S&P 500 except for after December 15 to December 31 of each year. Note that the purpose of this backtest is to determine the impact on annualized return of the second half of December; it’s not a strategy since there are tax and other considerations.
Skipping the second half of December reduces buy and hold annualized return from 8.3% to 6.4% and this is before commissions. The impact on performance of the second half of December is significant. The annualized return from holding the index in the second half of December is about 1.6%.
However, the same has not been true in the case of NASDAQ-100, as shown below, when using QQQ ETF.
The red line (buy and hold) and the black line (buy and hold ex-last two weeks of December) have only about 20 basis points difference in favor of former and in terms of annualized return, 7.3% versus 7.1%.
Therefore, for tech stocks and based on the limited sample from QQQ for 22 years, the last two weeks of December haven’t had any material impact on returns.
Also note these statistics are random variables and will change in the future. In my opinion, for very large samples, in the case of S&P 500, there shouldn’t be a calendar anomaly present.
Charting and backtesting program: Amibroker. Data provider: Norgate Data
10% off for blog readers and Twitter followers with coupon NOW10