Skip to content

Inspect residual sensitivity in a total-value multiple

For paid-in 100 and distributions 40, a one-unit residual change moves TVPI by 0.01×.

On this page

Specify the valuation scenarios

The original case fixes paid-in 100 and distributions 40. It considers residual estimates 50, 80, 110 and 140 solely as illustrative inputs. They are not probabilities, valuations of a named fund or a forecast path.

Calculate the conditional ratios

TVPI = (40+V)/100, so the four inputs produce 0.9×, 1.2×, 1.5× and 1.8×. DPI stays 0.4×. The derivative with respect to V is 1/100 on this fixed base.

Original residual scenarios
Residual inputRVPITVPI
500.5×0.9×
800.8×1.2×
1101.1×1.5×
1401.4×1.8×
0.91.351.85095140Residual value inputTVPI (×)0.91.351.85095140Residual value inputTVPI (×)
TVPI under stated residual inputs

Paid-in 100 and distributions 40 are fixed. The horizontal axis is a hypothetical residual input, not calendar time; no probability is assigned. This is an illustrative example.

View the chart values
TVPI under stated residual inputs: underlying values
SeriesResidual value inputTVPI (×)
TVPI500.9
TVPI801.2
TVPI1101.5
TVPI1401.8

Read the slope as a unit effect

The linear relationship shows how this ratio depends on its input. It does not identify which residual estimate is justified. An uncertainty range needs evidence beyond choosing convenient chart endpoints.

Do not ignore changes in other components

A new contribution changes paid-in, and a distribution can change both distributed and residual value. Those cases require a new complete record rather than treating every ratio movement as a valuation-only change.

Further references