The Unicorn Ratio

How many startups should you fund per batch?

A portfolio-construction calculator for YC investors. Uses the Kelly Criterion and Binomial Distribution to estimate how many companies to back and how to size each check.

Assumptions

1. Binary outcomes — each investment either becomes a success (unicorn) or does not.

2. Historically, each YC batch has an average 6.5% unicorn rate .

3. Average valuations at YC demo day are $40,000,000 post-money. This implies a 25.0× multiple needed to achieve a $1B+ outcome.

Inputs

1. No. of startups in the batch

2. Funds available to deploy

3. How much of the batch can you confidently rule out? (i.e. your picking ability)

4. Select a model

Kelly Criterion

Binomial Distribution

Objective: Optimize portfolio size and check amounts using the Kelly Criterion.

Analysis using Kelly Criterion

Invest in approximately 11 companies (out of 200 in the batch), with an average check size of ~$91,000.

Investment Scenarios

Filtering level	No. of investments	Check size
0%	38	$26,000
25%	21	$48,000
50%	11	$91,000
75%	4	$250,000
90%	2	$500,000

Investment Allocation

Write ~11 checks of $91,000

Total deployment: $1,000,000

Explanation

The Unicorn Ratio is a framework for deciding how many companies to invest in for a given YC batch. There are two ways to think about it: the Kelly Criterion and the Binomial Distribution.

Kelly Criterion

f = p - \frac{1-p}{b}

Where:

$f$ = the fraction of your fund to put into each investment
$p$ = the chance any single investment becomes a unicorn (rises as your filtering improves)
$b$ = profit per dollar invested if you succeed (a 50× return → b = 49)
$\frac{1}{f}$ = how many companies to invest in

Understanding the Kelly Criterion

The Kelly Criterion is the optimal fraction of your capital to put into each opportunity, balancing return against risk.
Applied to startup investing, it tells you both how many companies to back (portfolio size) and how much to put into each (check size).
The formula accounts for your skill at picking winners (the $p$ term) and net profit per dollar (the $b$ term).

Binomial Distribution

P(\text{at least one unicorn}) = 1 - (1-p)^n

Where:

$P$ = your chance of finding at least one unicorn
$n$ = how many companies you invest in
$p$ = chance any single investment becomes a unicorn

Understanding the probability curve

The "knee" is where adding more investments starts giving you significantly less benefit.
Before this point, each new investment meaningfully improves your odds.
After it, you need many more investments to achieve similar gains.
Mathematically, the knee is identified by the second derivative of the curve — the rate at which returns diminish.

Limitations

These calculations assume you have an accurate read on how good you are at picking (i.e. how much of the batch you can rule out).
Historical unicorn rate may not predict future outcomes.
You may want to size checks differently across deals — the price and dynamics of each round are unique.
Watch your conversion rate (# investments / # founder meetings). Reputation with founders matters.
VC returns follow a power law — non-unicorn outcomes (acquisitions, smaller exits) can still meaningfully impact portfolio returns.

Disclaimer:

The Unicorn Ratio is a framework that provides probabilistic estimates for portfolio construction. It should not be considered or used as investment advice.

Oversubscribed · Final close

OC Fund I

Join 150+ YC alumni, top VCs, frontier labs, and institutional LPs

Apply to Invest

Accredited investors only. Min. investment requirements apply. Advertising permitted under Rule 506(c).