Skip to content

DCT Lexicon · Risk and uncertainty

Percentiles (P10, P50, P90) and bands

P10 · P50 · P90

P10, P50 and P90 are the values below which 10, 50 and 90 percent of the simulated outcomes fall; a band between two of them shows the spread a plan carries, not a forecast of one number.

Drafted for the author's approval from Vol. I, Sections 8.8–8.9, and the DCT Helix white paper (Plans as distributions)

In depth

In the method

A single-number business case treats the most likely outcome as the only one. DCT Helix prices every step as a distribution instead. The Certify move produces stochastic pro forma statements: the income statement, balance sheet and cash flow, simulated thousands of times under declared assumptions, with the P10, P50 and P90 of value at stake.

DCT Helix white paper (October 2026), Plans as distributions, verbatim

Formal statement

Pq=inf⁡{ v:F^N(v)≥q },F^N(v)=1N∑i=1N1{Vi≤v}.P_q = \inf\{\, v : \hat F_N(v) \ge q \,\}, \qquad \hat F_N(v) = \frac{1}{N}\sum_{i=1}^{N} \mathbf{1}\{V_i \le v\}.

Drafted for the author's approval: the empirical quantile of $N$ simulated outcomes

How to read it

Plan on distributions where curvature lives; when a single path must be quoted, label it as the mean or as a percentile, never as “the case”.

Vol. I, Section 8.8, verbatim

Common misreading

A common mistake is to treat the fan as decoration around the forecast; the fan is the forecast.

Vol. I, LOS 8.2, verbatim

How AXIOM uses it

On the Financial impact tab each bar spans the 10th to the 90th percentile of the simulated futures and the dot is the median. Where a declaration is written as a value plus or minus a range, AXIOM reads the range as a 95% band, with a standard deviation of half the half-width, everywhere in the product.

Drafted for the author's approval, from the Financial impact tab and D-065, D-091

In the book: Vol. I, Ch. 8: Stochastic Enterprise Dynamics

Related: Monte Carlo simulation · Value at stake · Stochastic business case · Expected value

V1.Thm.8.5 · HX.P.10 · HX.E.3.Control.1