Thread by @ry_paddy

My napkin calculations estimate @ycombinator's DPI (distributed to paid in capital ratio) for their 2005 - 2013 vintages at ~320x.

That’s a 31,900% fund-level return. In cash terms, they turned ~$11.2m of invested capital into ~$3.6B.

Here's a rundown of the numbers:

Over the years the Y Combinator deal has evolved, but they’ve always tended to take a 6 - 7% stake.

In their very first batch, they invested ~$20,000 per company. It stayed this way until 2014, when it was increased to $120k.

So until 2014, their entry price was very low. It has varied since then, and is now $125k for 7%, plus an additional $375k in the next round, on “most favoured nation” (MFN) terms (i.e. at the best valuation possible).

From 2011 - 2014, various VC’s (Yuri Milner, Khosla) automatically invested in every company in YC batches, but YC still got their 6-7% for ~$20k during that time.

We’ll ignore the question of pro-rata / follow-on for the purposes of simplicity in this analysis. YC do follow their investments now, but they didn’t used to.

By the time a company gets to exit, you can expect a dilution of about ⅔ on average (according to reputable sources, including @fredwilson)

avc.com/2021/11/seed-r…

Crunchbase ran the numbers on some of YC’s big exits in 2020.

They conservatively estimated the cash generated by exits in 2020 alone as between $1B and $2B, mainly from two companies: AirBnB and Doordash.

The return on those two investments was ~3.5 *million* percent

If we add in the previous M&A activity with disclosed valuations (only a small percentage of total exits, which were mostly for undisclosed sums) and take the middling case, YCombinator delivered about $1.7B in cash by 2020 across all exits.

Now let’s add their TEN public offerings (via IPO & SPAC) in '21. In brackets, you can see the avg mkt cap around the time the company went public (or shortly after).

If they sold recently after all of these companies went public, then their ~2.2% stake would be worth ~$2.4B.

That’s ~$4.1B IN CASH that YC has generated in the last two years, and we aren’t counting any of the numerous private acquisitions they had in 2021

ycombinator.com/blog/2021-yc-y…

90% of this value (~3.6B) was generated on investments made prior to 2013.

Between 2005 and 2013 YC only deployed ~$11.2m, so they generated very roughly a 320x cash on cash return on capital deployed.

That’s DPI, not TVPI.

That’s ~242x - 320x better than the average DPI delivered by a VC in the USA over that time period (1 - 1.32x), according to data from Cambridge Associates.

cambridgeassociates.com/wp-content/upl…

They achieved all of this in a 15 year total holding period for the 2005 - 2013 portfolio, with three quarters of the capital deployed from 2010 onwards.

So their IRR for that time period was absolutely through the roof (at least 90% or more IRR for 10 consecutive years).

At @JoinOdin, we are partnering with Fundament, an incubator and angel syndicate led by @Mentioum, cofounder of Headstart (YC 17), to offer ppl the opportunity to invest in 10 selected companies from the Y Combinator Summer 2022 (S22) batch.

Sign up:
tinyurl.com/ms7nv3kf

You can also subscribe to our newsletter to get more great content on early stage investing

odin.substack.com