Cherry-picking: base period 2.0
About this app
Rebuild 2.0 (preview): the same series, every start–end pair as a triangle of returns, distribution of all windows, rebasing to 100 and axis tricks.
All time windows at once
A return depends on where the time window starts and ends. To compare windows of different length you compute the average annual growth, the CAGR: (P_end/P_start)^(1/J) − 1, with J in years. A return triangle shows all possible windows at once: each cell stands for one start–end pair, its colour for the CAGR. Near the diagonal sit the short windows with the most extreme values, because annual rates extrapolated from a few months swing widely. Showing only one window hides how many others there are.
What the app shows
You choose between a simulated stock index over ten years and the revenue of two sales regions over 15 months. On the left you see the series with the chosen window and a dashed path of steady growth; on the right, the return triangle. The right-hand column holds all windows ending today, the ones broker apps offer as 1 day to Max. A histogram shows the distribution of all windows of one year or longer and the share in the plus. In the region comparison the triangle shows which region grows faster in each window.
Spotting axis tricks
Set the y-axis to “from 0”, “fitted” or “log”. A truncated axis stretches every movement, and the app reports Tufte’s lie factor: the visible relative change in height divided by the real one. On a log axis equal height means equal percentage change, and steady growth becomes a straight line. “Window only” hides the rest of the series, as a slide would. Also note the difference between percentage points and percent, which the app shows in the region comparison.
Frequently asked questions
- What is CAGR?
- CAGR, the compound annual growth rate, is the average yearly growth over a period, computed as (end value/start value)^(1/years) − 1. It tells you the constant rate that would have taken you from the start to the end value. For periods shorter than a year it is an extrapolation and correspondingly uncertain.
- What is a return triangle?
- A return triangle is a table or chart in which each cell shows the return for a particular start and end year. Rows stand for the start, columns for the end. It shows at a glance how strongly the result depends on the chosen period.
- What is Tufte’s lie factor?
- Edward Tufte defined the lie factor as the size of an effect shown in a graphic divided by the size of the effect in the data. A value of 1 means an honest display. A y-axis that starts far above zero produces values well above 1.
- What does this version of the app offer?
- It shows every start–end pair of a series in a return triangle, the distribution of all time windows and the CAGR of the chosen window. On top come three axis types with a lie factor, indexing to 100 and a comparison of two sales regions. The “Cherry-pick” button selects a flattering window.
Subject: Data & decisions
More from Data & decisions
- Anchoring effect
- Benford's law
- Berkson's paradox
- Cherry-picking: base period
- The three-cowboy duel (truel)
- The three-cowboy duel 2.0
- The three means
- The three means 2.0
A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps