No paywall for this report!
Last week, I came across a viral tweet in which the author claimed that by starting with just $100 in 2012 and then rotating it into the previous year's best-performing S&P 500 stock, your portfolio would now be worth $176K.
For context, the same $100 left in the S&P 500 or Berkshire Hathaway would be worth about $800.
So is that all there is to investing? Can you really compound at 66% a year just by betting on last year’s winner? As the saying goes, there are lies, damned lies, and statistics.
A quick note about AI backtests
Generally, I am skeptical of these AI backtests because they almost always make simple errors. For example, here’s what you will get if you ask Gemini what the best-performing S&P 500 stock in the year 2000 was.
But a simple fact check shows that none of the top 3 stocks mentioned in this list were in the S&P 500 index in 2000. Laboratory Corp got added in Oct’04, DGX in Dec’02, and VRTX in Sep’13.
Here’s Gemini’s response when you call that out:
Pre-AI, you didn’t have to be so skeptical, since any analyst who puts their name on the research will triple-check their work or lose all credibility instantly.
So, is the above backtest correct?