Consensus Bell-Curve Pruning
Map professional forecasts and remove the already-priced middle
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 97%
Fisher treats professional forecasts as information about what is already priced rather than as a reliable forecast of what comes next. His team collects professional investor sentiment for sectors, countries, and the overall market, then arrays the views into bell curves. The center represents the collective expectation. Because that outcome is already embedded in market prices, Fisher says the middle of the distribution does not occur twelve months out. This does not identify the winning alternative: the actual result may fall on either side, and further work is required to decide between them. The value is in pruning the possibility space. Like removing several hours from a clock face, excluding the crowded center stacks the odds toward less anticipated outcomes while preserving uncertainty about which tail will prevail.
Origin
Fisher described the method as something his firm always does: collect professional sentiment, array it into bell curves, and treat the middle as already priced.
Core principles
- 01Collective professional expectations are already reflected in prices
- 02Consensus is more useful for exclusion than prediction
- 03Removing likely outcomes improves the remaining odds
- 04The method does not reveal which non-consensus side will occur
How to run it
- 1
Collect professional expectations
Gather what professional investors and forecasters believe about the market, countries, and sectors. The goal is to measure consensus, not endorse it.
Pro tip Keep separate distributions for distinct questions.
Watch out A few prominent forecasts do not necessarily represent the group.
- 2
Build the distribution
Array the forecasts into a bell curve or comparable expectation distribution. Locate the middle where professional opinion is most concentrated.
Pro tip Use a consistent twelve-month horizon when applying Fisher's stated version.
Watch out Mixing horizons can create a false center.
- 3
Prune the middle
Treat the crowded center as already priced and remove it from the leading outcome set. Use consensus to identify what is least likely to surprise the market.
Pro tip Frame this as an exclusion, not a precise counter-forecast.
Watch out The center being priced does not prove one specific tail will win.
- 4
Analyze both tails
Investigate the plausible outcomes on each side of consensus. Decide which side has stronger drivers while accepting that the judgment can be wrong.
Pro tip Look for evidence the market has not already discussed heavily.
Watch out Do not turn a probabilistic narrowing tool into certainty.
In the wild
Fisher compares the method to taking about three hours off a clock. Knowing what professional forecasters collectively believe does not tell him what will happen, but it removes a portion of the outcome space because that belief is already priced.
→ The research effort concentrates on less crowded outcomes on either side of consensus.
Common mistakes
Following the middle
Treating consensus as the most likely forecast ignores Fisher's premise that the collective expectation has already been priced.
Assuming the opposite automatically
The method removes the center but does not determine whether the left or right side will occur. Both tails still require analysis.
Is it for you?
Best for
It is best for market strategists with enough professional-forecast data to estimate the shape of expectations.
Not ideal for
It is not ideal when the forecast sample is too small or unrepresentative to characterize consensus.
From the transcript
“if you can figure out what they believe will occur, you know exactly what won't happen.”
“arraying that into bell curves, and knowing that twelve months out the middle of the bell curve doesn't happen.”
“Could be the left, could be the right. We got to try to figure that out.”
From the episode
Interview With Ken Fisher: Masters in Business (Audio)