How much return are you getting for the risk you are taking

Why they are one page

The Composite Efficiency Index reads whether a fund's configuration holds together. These two read something different: the shape of the outcome distribution that configuration implies. One describes the return side of that shape, the other the dispersion side, and reading either alone gives half a picture.

Return per Unit of Variance

It is displayed as a multiple, so 1.25x reads directly: this configuration is expected to produce more return per unit of dispersion than the median fund at the same lifecycle stage. Below 1.00x means less. The comparison is always within lifecycle stage, because a Conviction-stage fund and a Continuity-stage fund are not on the same frontier and comparing them would be comparing different questions.

The phrase to sit with is per unit of dispersion. This is not a prediction of return. It is a reading of return relative to the spread a configuration implies, which is why a concentrated fund and a broad fund can land at similar values for entirely different reasons: one earns it through expected magnitude, the other through a tighter spread.

The Consistency Score

A 0 to 100 reading where higher means a tighter expected distribution. It is driven by the configuration choices that influence dispersion: portfolio size, sector breadth, geographic scope, target stage, and reserve allocation. More positions and more breadth pull the distribution tighter; concentration and earlier stages widen it.

A high Consistency Score is not automatically desirable, which is the most common misreading. Venture returns come from the right tail, and a configuration that narrows the distribution narrows it at both ends. A fund optimized for consistency alone has optimized away the outcome it exists to produce.

Reading them together

The pair is more informative than either number, and the four combinations mean four different things.

High on both is a configuration expected to produce good returns without extreme dispersion, which usually means breadth with real selection discipline. High return per unit of variance with a low Consistency Score is the classic concentrated venture shape: the expected return is there and it depends on a small number of outcomes. That is not a warning, it is a description of most successful early venture funds.

Low return per unit of variance with a high Consistency Score is the combination worth pausing on. It describes a fund that has taken the diversification without the upside: spread widely enough to damp the tail, without the ownership or concentration that would make a tail event matter. Low on both usually points at something specific in the configuration rather than a strategy at all.

What these are not

Neither is a forecast. Return per Unit of Variance is not a prediction of realized IRR, TVPI, or any other return metric, and the Consistency Score is not a probability of anything. Both are readings of what a declared configuration implies, computed from the configuration alone, and both would return the same values on the day the fund closed as on the day it made its first investment.

They also say nothing about the companies. A fund with an excellent configuration and poor selection will do badly, and the model has no visibility into which companies a firm can actually reach.

And the internals stay internal. What drives these two is proprietary to Colibrí Strategies and is not published, which is a deliberate position rather than an oversight: the model describes what it evaluates and what a General Partner sees, and stops there.