Multi-Horizon Trend Check
Judge noisy economic releases through rolling trends, not one month's swing
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 4
- Confidence
- 96%
Rosenberg's trend check separates the underlying signal from the drama of a single release. First identify how volatile the series is: he distinguishes the more volatile household employment survey from the generally steadier payroll survey. Then place the latest number beside the previous report rather than reading it alone. If one month triggered recession fears and the next triggered euphoria, resist both narratives and look for the middle ground. Inspect three-month, six-month, and twelve-month trends to see whether the direction survives across horizons. The resulting conclusion should describe the condition of the economy, not celebrate or condemn the newest print. In the episode, this process turns September's weak number and October's strong number into a restrained assessment that the US economy was in decent shape.
Origin
Extracted from Masters in Business
Core principles
- 01A single release can produce excessive angst or euphoria
- 02The truth often lies between adjacent extreme readings
- 03Longer windows reveal the underlying direction
- 04Different data series have different levels of volatility
How to run it
- 1
Classify the series
Determine whether the data source is characteristically volatile or stable. Give less weight to isolated readings from noisier series.
Pro tip Distinguish related surveys instead of treating all headline employment data alike.
- 2
Pair adjacent releases
Compare the newest report with the previous one. Note when the emotional narratives attached to them point in opposite directions.
Watch out Do not let one weak or strong month define the cycle.
- 3
Inspect three horizons
Review the three-month, six-month, and twelve-month trends. Look for a direction that remains visible as the window expands.
Pro tip Use the longer windows to dampen monthly jumps.
- 4
State the middle-ground signal
Summarize what the trend says about the economy without importing the angst or euphoria of either release. Keep the conclusion proportional to the evidence.
Pro tip Use restrained labels such as decent shape when that is all the trend supports.
Watch out Do not turn noise reduction into false precision.
In the wild
The weak September report encouraged recession talk, while the much stronger October report encouraged euphoria. Rosenberg put both in context and concluded that the earlier angst and later enthusiasm were overdone, with the underlying economy doing okay or somewhat better.
→ The paired and trended reading avoided two opposite overreactions.
Common mistakes
Extrapolating one month
A noisy release can reverse quickly and should not carry the full interpretation of the economic cycle.
Mixing unlike surveys
The household and payroll surveys have different volatility, so their individual readings deserve different treatment.
Is it for you?
Best for
It is best for recurring economic series such as payroll and household employment data.
Not ideal for
It is not ideal when a genuine structural break makes older observations irrelevant.
From the transcript
“the truth is usual somewhere in the middle.”
“focus on the three months, six month, even the twelve month.”
“the data telling you the economy is in decent shape. Full stop.”
From the episode
Interview With David Rosenberg: Masters in Business (Audio)
David Rosenberg