End of year trading

I listened to Yahoo Finance this morning, where Interactive Brokers' chief strategist mentioned that investors typically sell losers and hold on to winners at the end of the year. This makes sense, considering tax planning.

Based on this, I expected the losers to perform worse toward the end of the year. However, I did a quick backtest to analyze which stocks perform best and worst during this period. To my surprise, it’s the stocks that have lost the most that perform best in the last few days of the year, while the stocks with the best performance tend to do the worst during this time.

If you buy the 10 worst performers on the first trading day after December 25th and hold them until the first trading day of the new year, the results are as follows:

  • S&P 500 stocks: 2.06% return
  • Russell 2000 stocks: 6.54% return
  • Dow Jones stocks: 0.76% return

If you buy the 10 best performers during the same period, the results are:

  • S&P 500 stocks: -0.28% return
  • Russell 2000 stocks: -1.31% return
  • Dow Jones stocks: 0.02% return

The results appear consistent across different universes—buying the losers at the end of the year seems to be a better strategy.

This backtest was conducted using data since 2003 with a 245-day Rate of Change (ROC).

What is less surpricing is that most of the return comes over New Years day. This is the result of you buy the worst 20 stocks of the year in R3000 and hold them over New Years.

These are the candidates for this new years.

Ticker Name
GOEV Canoo, Inc.
CDT Conduit Pharmaceuticals, Inc.
MOND Mondee Holdings, Inc.
SAVEQ Spirit Airlines, Inc.
BNAI Brand Engagement Network, Inc.
CABA Cabaletta Bio, Inc.
SAVA Cassava Sciences, Inc.
IVVD Invivyd, Inc.
VRCA Verrica Pharmaceuticals, Inc.
LAZR Luminar Technologies, Inc.
ALXO ALX Oncology Holdings, Inc.
STTK Shattuck Labs, Inc.
EDIT Editas Medicine, Inc.
CBUS Cibus, Inc.
AVTE Aerovate Therapeutics, Inc.
HLVX HilleVax, Inc.
CHGG Chegg, Inc.
WOLF Wolfspeed, Inc.
INO Inovio Pharmaceuticals, Inc.
AGEN Agenus, Inc.
3 Likes

How does that compare to other weeks of the year?

It actually looks like a good trade in general and the trend is quite clear.

If I test the same trade for all months ranked by ROC(245) on SP500, buy first trading day after the 25th, sell first trading day next month, yearly CAGR is as follows -

Top250 - 3.75%
Top100 - 3.53%
Top50 - 3.21%
Top25 - 3.04%
Top10 - 2.47%
Bottom250 - 5.31%
Bottom100 - 6.33%
Bottom50 - 7.86%
Bottom25 - 8.82%
Bottom10 - 11.75%

The trade seems to work well all on all months exept the summer months May to Aug.

If you hold until the 4th trading day of the month you can add quite a bit of CAGR.

Top250 - 6.04%
Top100 - 5.00%
Top50 - 4.43%
Top25 - 4.13%
Top10 - 3.15%
Bottom250 - 9.24%
Bottom100 - 10.56%
Bottom50 - 12.65%
Bottom25 - 13.25%
Bottom10 - 16.14%

I wonder how much of this is simple regression to the mean. I would surmise that if you expanded the study to any week of the year and used a much shorter ROC (20 days, for instance) you'd get even better results.