The following is my answer to a Quora question: “How do venture capital firms make money if 90% of investments fail?”
The
question contains its own answer.
Venture capital does not need 90 per cent of investments to
succeed. It needs one to succeed
spectacularly enough to make the other nineteen irrelevant. This is not a flaw in the model. It is the model.
The
Mathematics
The
standard venture capital fund structure assumes a portfolio of 20
investments. The fund manager — the
general partner — raises capital from institutional investors and
high-net-worth individuals — the limited partners — and deploys it across the
portfolio over a three- to five-year investment period. The power law of venture returns dictates the
outcome. Assume a US$100 million fund
deployed across 20 companies at US$5 million each. Of 20 investments, 10 fail and return
zero. 6 return between 0.5x and 2x —
they do not fail outright, but they do not generate meaningful returns. 3 return between 2x and 5x — modest
successes. One returns 100x.
The
one company that returns 100x on a US$5 million investment returns US$500
million. The three modest successes
return between US$10 million and US$25 million each, contributing US$50
million. The six mediocre performers
return US$30 million. The
ten failures return zero. Total fund
return: US$580 million on a US$100 million fund — a 5.8x return before
fees. The general partner takes a 20 per
cent carried interest — the “carry” — on profits above the hurdle rate,
typically 8 per cent. The limited
partners receive the remainder. This is
why the 100x outcome is not a bonus. It
is the entire thesis. Without it, the
fund returns 0.8x — a loss after fees and the cost of capital.
The
Return Expectations by Market
The
expected return profile varies significantly by stage and sector.
Seed-stage
technology startups — the earliest bets, typically US$500,000
to US$2 million — carry the highest failure rate and the highest potential
multiple. The venture firm expects to
lose its investment in 60 to 70 per cent of cases. It expects 30x to 100x returns from the
survivors. Y Combinator’s early
investments in Airbnb (US$20,000 for 6 per cent, now worth billions) and
Dropbox (similar structure) represent the archetype.
Series
A technology — US$5 million to US$20 million — carries
a lower failure rate but a compressed multiple.
The expectation is 10x to 30x on the winners, with a 40 to 50 per cent
failure rate. The fund needs fewer
enormous outliers to generate returns.
Growth-stage
venture — US$20 million to US$100 million — operates on a 5x
to 15x return expectation with a lower failure rate still. At this stage, the company has product-market
fit. The question is execution, not
survival. The failure rate drops to 20
to 30 per cent.
SaaS
specifically — Software-as-a-Service companies are the
preferred venture category because of their recurring revenue model, high gross
margins (typically 70 to 80 per cent), and the compounding nature of their
customer acquisition. A SaaS company
growing at 100 per cent annually doubles its revenue each year. At year five, it is 32 times its original
size. The multiple expansion from
entering at a 5x revenue multiple and exiting at a 10x revenue multiple on a
company that grew 32x produces a return of 64x.
Deep
tech and biotech — longer development cycles, binary
outcomes (FDA approval or not; the technology works or it does not), and
capital-intensive development produce a specific return profile: higher failure
rate, longer time to exit, but potentially unlimited upside if the technology
succeeds. Returns of 50x to 300x are
documented for successful biotech exits.
The failure rate exceeds 80 per cent.
Real-World
Cases
Sequoia
Capital invested US$60 million in WhatsApp across multiple rounds. Facebook acquired WhatsApp for US$19 billion
in 2014. Sequoia’s return: US$3 billion
— a 50x return on its investment. That
single investment covered the losses of an entire fund.
Benchmark
Capital invested US$5 million in Uber in 2011 for 20 per cent of the
company. Uber’s IPO in 2019 valued
Benchmark’s stake at US$7 billion — a 1,400x return. One investment. One fund.
One outcome that defines the firm’s entire history.
Andreessen
Horowitz invested US$250,000 in Instagram in 2010. Facebook acquired Instagram for US$1 billion
in 2012. Andreessen Horowitz’s return: US$78
million on a US$250,000 investment — a 312x return in 18 months.
Accel
Partners invested US$12.7 million in Facebook’s Series A in 2005 at a valuation
of US$100 million. Facebook’s IPO in
2012 valued Accel’s stake at US$9 billion.
That investment funded Accel’s operations for a decade.
The
common thread is not diversification. It
is concentration in the one or two companies within a portfolio that the fund
manager correctly identified as having the potential for category-defining
scale.
How
a Venture Fund is Constituted
The
general partner — the venture firm — establishes a limited partnership
structure. The general partner
contributes 1 to 2 per cent of the fund’s capital — the GP commitment — which
aligns their interests with the limited partners. The limited partners contribute the remaining
98 to 99 per cent. The fund has a
defined life — typically 10 years, with two one-year extension options. Years 1 to 5 are the investment period. Years 5 to 10 are the harvesting period — the
general partner works the portfolio toward exits.
The
general partner charges a management fee — typically 2 per cent of committed
capital annually during the investment period, declining to 1.5 per cent
thereafter. On a US$100 million fund
over 10 years, this generates US$17 million in management fees regardless of
performance. This is how venture firms
pay their staff, rent their offices, and fund their operations. The carry is how they get wealthy. The carry — 20 per cent of profits above the
hurdle rate — is the primary incentive mechanism. On a US$100 million fund generating US$580
million in returns, the profit above the 8 per cent hurdle rate is US$400
million. The general partner’s 20 per
cent carry is US$80 million, split among the partnership’s principals.
How
the Fund is Disbursed
Capital
calls rather than upfront disbursement characterise the LP experience. The limited partners do not transfer all
committed capital on day one. The
general partner issues capital calls — requests for a portion of committed
capital — as investments are made. A LP
who commits US$10 million to a fund may receive capital calls of US$1 to US$2
million per year over five years. This
structure allows limited partners to manage their liquidity. It also means that venture funds operate with
uncalled capital — committed but not yet deployed — that represents a
contingent liability for the limited partner rather than an immediate cash
outlay.
How
the Fund Recovers Its Investment
The
exit mechanisms are three: initial public offering, merger and acquisition, and
secondary sale.
The
IPO — listing the portfolio company on a public exchange — provides the most
liquid exit but also the longest timeline.
The venture fund receives shares in the public company, subject to a
lock-up period typically of 180 days, after which it can sell into the public
market. Uber, Airbnb, DoorDash, and
Snowflake are examples of venture-backed companies that provided IPO exits.
The
M&A exit — the portfolio company is acquired by a larger company — provides
the most certain exit but at a negotiated valuation. Instagram, WhatsApp, YouTube, and LinkedIn
are examples of venture-backed companies acquired before IPO. M&A exits tend to be faster, more
certain, and at lower multiples than IPO exits — but in a market where IPO
windows close, they represent the pragmatic alternative.
The
secondary sale — the venture fund sells its stake to another investor,
typically a growth equity firm or a sovereign wealth fund, before a public exit
— has become an increasingly important mechanism as companies stay private
longer. The average time from founding
to IPO for a US technology company increased from 4 years in 1999 to 12 years
by 2020. Secondary sales allow
early-stage investors to realise returns without waiting for a public exit that
may be a decade away.
The
Assessment
Venture
capital is the only asset class where the professional’s primary job is to be
wrong 90 per cent of the time and still generate superior returns. The model works because the mathematics of
power law distributions — where one outcome dominates the entire distribution —
is structurally different from the bell curve distributions that govern most
asset classes. In a bond portfolio, no
single bond can return 100x. In a public
equity portfolio, no single stock can return 1,400x in 8 years. In a venture portfolio, both are documented
outcomes.
The
failure rate is not a problem to be managed.
It is a feature of the model — because the companies willing to take the
risks that produce 100x returns are the ones that might fail. A company unwilling to take those risks is
unlikely to return more than 3 to 5x — insufficient to justify the illiquidity,
the management cost, and the risk premium that venture capital commands. The 90 per cent failure rate and the 100x
return are the same phenomenon, viewed from different vantage points. The venture capitalist who refuses to invest
in companies that might fail will never invest in companies that might return
100x. This is why venture capital is not
an investment model that most investors should attempt to replicate — and why
the ones who do it well are considerably wealthier than the ones who merely
attempt it.
Terence
Nunis | Executive Chairman, Equinox Zenith | Author, The 1% Playbook: The
Billionaire Cheat Code

No comments:
Post a Comment
Thank you for taking the time to share our thoughts. Once approved, your comments will be poster.