Relative Strength Rotation
Rank competing assets by price leadership and rotate only when the winner changes
- Difficulty
- Advanced
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 99%
Dorsey compares relative strength to an arm-wrestling contest. Divide one asset's price by another and plot the result: when the line rises, the numerator is winning; when it falls, the denominator is winning. Repeat the comparison across the eligible universe, rank the competitors, and own the leaders until a confirmed change identifies a stronger alternative. The process can compare stocks, sectors, weighting schemes, countries, asset classes, and cash. It does not attempt to call the exact end of a cycle. Dorsey instead waits for the former loser to win the contest, even when the prior regime lasted years. Portfolios such as the People's Portfolio and First Trust Five apply the mechanism to a small set of choices. The cost is periodic underperformance when little separates the best and worst assets and the system is searching for traction.
Origin
Dorsey Wright first performed relative-strength comparisons by hand and later automated millions of nightly comparisons across global markets.
Core principles
- 01Relative strength compares one asset directly with another.
- 02Price leadership can persist for years rather than days.
- 03The method waits for evidence of a new winner instead of forecasting a turn.
- 04Cash can compete in the ranking as a risk-management asset.
- 05A narrow performance spread can temporarily weaken the method without invalidating it.
How to run it
- 1
Choose the competitors
Define the stocks, sectors, asset classes, or portfolio structures that can compete for capital. Add cash when the process is meant to reduce market exposure.
Pro tip Compare alternatives that solve the same allocation decision.
Watch out A weak or irrelevant universe can only produce the best member of a poor set.
- 2
Calculate relative strength
Divide the price of one candidate by another or by a benchmark and plot the ratio. A rising series means the first candidate is outperforming.
Pro tip Automate broad comparison matrices when the universe is large.
Watch out Relative strength measures leadership, not absolute cheapness.
- 3
Rank the field
Use the comparison signals to order the eligible assets from strongest to weakest. Select the top-ranked number required by the portfolio rule.
Pro tip Keep the number of holdings and replacement threshold fixed in advance.
Watch out Changing the ranking rule to preserve a favored holding reintroduces discretion.
- 4
Wait for confirmation
Hold the current leaders until the relative-strength evidence shows that another candidate has won. Do not buy a laggard solely because it appears to have more upside.
Pro tip Set alerts so a long-developing leadership change cannot escape notice.
Watch out Trying to predict the crossover abandons the evidence-based mechanism.
- 5
Rotate and monitor
Replace a weakening holding with the strongest qualifying alternative and continue the comparisons. Judge temporary weakness in light of the dispersion between candidates.
Pro tip Expect fewer changes when durable trends persist for years.
Watch out Low dispersion can produce whipsaw while the strategy searches for traction.
In the wild
Dorsey compared the capitalization-weighted S&P portfolio, the equal-weighted S&P portfolio, and cash. Whichever alternative won the relative-strength contest received the portfolio allocation, with signals changing only over long cycles rather than on a daily schedule.
→ The model selected between competing equity structures and cash without requiring a forecast of the cycle's turning point.
Dorsey Wright compared twenty First Trust sector funds, selected the five strongest, and allowed them to remain until relative strength crossed the replacement threshold. The portfolio could go years without a change while leadership remained durable.
→ The portfolio concentrated in the leading sectors and excluded falling sectors such as oil during the period Dorsey discussed.
Common mistakes
Buying last place too early
A low rank may offer future upside, but the method waits until the asset actually begins winning the comparison.
Rotating on a calendar
Dorsey's examples change when price-based relative strength crosses, not simply because a review date arrives.
Expecting constant outperformance
Narrow dispersion can leave the model temporarily out of favor while it searches for a durable leader.
Is it for you?
Best for
It is best for portfolios with comparable assets, reliable price data, and permission to rotate as leadership changes.
Not ideal for
It is not ideal for investors who must predict bottoms, trade constantly, or hold every eligible asset regardless of rank.
From the transcript
“the way we do a relative strength charters, divide one thing by another.”
“instead of trying to call the point that that you're at the end of the cycle, let the work do it. Let the relative strain…”
“you have to wait until the actual change your relative strength charts begin to change and begin to win the arm wrestling contest”
From the episode
Interview With Tom Dorsey: Masters in Business (Audio)
Tom Dorsey