The question
Buy the dip has the simplest rule in the lab. Each night it ranks the companies that have fallen furthest below their highest price of the past year, as long as they are still above their 200-day average. That average is the stock's mean price over roughly the last ten months, and staying above it is a common sign that the longer trend is still up.
Over the ten years it made a year, against for the S&P 500. So why is it still an experiment, and not a strategy we ask people to follow?
How we tested it
Like every strategy, it was tested three times: over the whole ten years, and over each five-year half from a fresh start.
For each test we recorded the worst drop. That is the biggest fall from a high point to the next low. It is measured with each day's new $100 taken out, so fresh money cannot hide a fall.
The launch bar sets a limit on that drop, because a return only counts if a person could have held on long enough to get it.
The result
What it means
. Many people would sell somewhere on the way down, and then the ten-year return would never be theirs.
So Buy the dip stays in the lab. It is still ranked every night and its picks are still published, so its record keeps building in public. It has not passed the bar.
What it does not prove
- It does not prove that buying dips is a bad idea. This is one rule, tested on one decade.
- It does not explain the gap between the tests. The figures do not say why a fresh five-year start fell so much further than the ten-year run. One likely part of it: a new portfolio holds only a few stocks, so one bad pick moves it more. We have not tested that.
- The limit is a judgment. There is no proof that 45% is the right line. It is our view of what most people can sit through.
- Nobody invested this money. It is a test on past prices, and live results are likely to be lower.