In the last post I argued that a pre-revenue company is priced the way a developer prices a vacant block of land: start at the end, subtract what it costs to get there, discount for time and risk, and what’s left is the value today. That dissolves the “how can it be worth anything” objection. But it leaves the harder question open. Discount for risk: fine. At what rate? And once you have the rate, what return are you playing for?

The honest answer has two numbers in it that don’t match. The maths says the one company in your portfolio that works has to return about 30x. No sane investor underwrites a single deal to 30x, you underwrite to 10x. The gap between those two numbers is the whole game, and it’s worth understanding exactly where each one comes from.

The method nobody teaches you

The standard finance toolkit (discounted cash flow, a cost of capital from CAPM, and comparable-company multiples) was built for businesses with an operating history. It falls apart at seed for a dull, decisive reason: the inputs don’t exist. No cash flows to discount, no beta to calculate, no comparables that are guaranteed to survive.

So venture uses the Venture Capital Method, which does what the developer did: start at the exit and discount back. It isn’t widely known outside the industry, so here’s the one table the whole thing rests on. Each stage gets its own discount rate, and the rate falls as the company matures, because each later investor prices a narrower set of risks than the one before:

Stage	Discount rate band	The risk being priced
Pre-seed / idea	70–100%	Team, technical feasibility, whether it works at all
Seed	50–70%	A first product, earliest signs of a market
Series A	40–60%	Product-market fit, repeatable revenue
Series B	30–50%	Scaling, unit economics
Late stage	20–35%	Path to exit, most risk already gone

The bands come from the venture finance literature: Bhagat formalised the method, building on earlier work by Keeley, Punjabi, and Turki. Two things to notice before we use it. The rates are enormous at the top and ordinary at the bottom. And every cell is a band, 20 or 30 points wide, not a number. Both facts matter more than the method itself.

The 100% is a survival rate — here’s the derivation

Take the top of that first band, 100% at pre-seed, and build it from the ground up, because asserting “it encodes the failure rate” isn’t good enough.

A well-run angel portfolio returns something like 20–30% a year over time. Call it 25%. That’s the headline, and it hides a brutal distribution underneath.

• Take ten pre-seed investments.
• Eight or nine return nothing. Call it nine to zero.
• So the entire return of the book has to come from the one survivor.

What does that survivor have to do? Ten equal tickets, nine dead, and you need the whole book to compound at 25% a year for five years.

• 1.25⁵ is roughly a tripling — call it 3x on the whole $10 deployed.
• So the one survivor has to turn its $1 into about $30. The other nine contribute nothing.

The survivor needs roughly 30x. Not because 30x is a sensible expectation for any company, but because it’s carrying the funerals of the nine that died beside it.

Now turn that 30x-over-five-years into an annual rate, because that’s what a discount rate is. Here’s the back-of-envelope I keep in my head: 100% a year, compounded for five years, is 2⁵ = 32 — call it 30x. A doubling every year, five times. So the rate you apply to a single pre-seed company, if it’s to stand for the whole survival-weighted book, is around 100%. That’s the top of the band, derived rather than asserted.

The 100% was never a return forecast. It’s the annualised figure the lone winner must hit to justify a book where most entries go to zero. Read it as a survival rate and it stops being deranged. This is the same power-law engine I worked through in Roll the Die, one winner from ten carrying the whole portfolio, only here wearing the disguise of a discount rate.

Two medians, and why this matters

Here is where careless thinking causes real errors, so be precise about which median you mean.

Across the whole portfolio, the median outcome is a write-off. Most companies return nothing; the middle of the distribution is zero. That’s the survival rate above, stated as a median.

But across the companies that actually reach a successful exit, the survivors, the median is a long way from zero. Most good angel exits aren’t billion-dollar IPOs; they’re trade sales in the tens of millions. Seraf, working from historical norms, puts the most likely successful outcome at an acquisition in the $20-40M range. And by the time you get there, the later rounds have done their work on your holding: seed, Series A, and Series B each sell a slice to new money, and a pre-seed investor typically ends up with something like a third of the equity they started with. Run that through, and an exit toward the middle of that band returns you roughly 2 to 4x your money.

So both medians are true at once, and they describe the same distribution at different points. Across all your bets: a write-off. Across the ones that exit: a few times your money. Neither median is where the returns live. The returns live in a tail that medians, by construction, are built to ignore.

30x on the shares is 10x in your pocket

Now let’s look at the two numbers from the top, and the sum that ties them together.

The survival maths said the one winner has to do about 30x. But that’s 30x on the company’s shares, and you don’t keep all your shares. By the time the company exits, seed, Series A, and Series B have each sold a slice to new money, and a pre-seed investor typically holds something like a third of the position they started with. So a company whose shares do 30x puts roughly 10x in your pocket. That’s the whole envelope: thirty times, diluted to a third, is ten times to you.

This is why 10x is the number you underwrite to. It isn’t a softer target than the 30x. It’s the same outcome, measured after the cap table takes its cut. 30x is what the shares have to do; 10x is what that leaves you holding. The survival rate and the underwriting standard are one result seen gross and net.

And notice what 10x still demands. Enter at a $3.5M pre-money and a 10x on your money needs an exit somewhere around $90 million. Not the $20–40M median survivor, but a top-decile one. Even your underwriting target is a tail event. The median company that exits returns you a few times your money; 10x sits well above it, and the genuine once-a-fund monster sits further out still — a company whose shares run so far past 30x that even a third of it is a fund-making result. You can’t underwrite to that one. You can only hold enough positions to give the power law a chance to hand it to you.

Why entry price is the only lever that matters

If you can’t underwrite to the monster, and even 10x needs a top-decile exit, the one thing genuinely in your control is the price you pay to get in. Watch what it does:

• Enter at $3.5M pre-money → 10x needs an exit around $90M
• Enter at $2M pre-money → 10x needs an exit around $55M
• Enter at $1M pre-money → 10x needs an exit around $33M

Lower entry prices move 10x onto smaller, more probable exits. The cheaper you enter, the less you need the planets to align. Entry price isn’t one input among many; in power-law investing it’s close to the whole game, because it’s the only term in the return equation you set rather than hope for.

So why does the market clear at $3–5 million?

This is where the rigorous foundation distils into the rule of thumb I trade on, much like a trader who understands Black-Scholes thoroughly and then trades off a couple of heuristics rather than the full model on every ticket.

The entry-price lever above says cheaper is always better for the investor, which would push valuations toward zero. So what holds them up? The existence of large exits. The whole edifice (the 100% discount rate, the willingness to pay millions for a company with no revenue) only stands because the right tail is real. If the best plausible outcome were a $5M trade sale, nobody could justify a $4M pre-money on a pre-revenue company; the maths wouldn’t close. It’s precisely because exits run into the tens and hundreds of millions that paying $3–5M for the ticket makes sense. The fatter and more frequent the large exits, the more an early entry is worth. This is because the 10x in your pocket becomes reachable on a realistic, if top-decile, result. Large exits are the upward force on early-stage valuations.

Pushing the other way is funding risk. Price a round too high, say $8–10M at pre-seed, and you’ve borrowed against the company’s own future. The next round now has to clear above $8–10M to give your earlier investors any uplift at all, and pricing that high that early makes the step-up hard to deliver, baking down-round risk into the cap table before the company has earned anything. A down round at Series A doesn’t just dent the numbers. It kills momentum, signals trouble, and can be fatal. Funding risk is the downward force.

This isn’t theoretical. The Angel Capital Association flagged exactly this in 2025, warning that early-stage valuations rising against flat Series A pricing means more dilution to early investors and greater risk of down rounds, with potentially fatal consequences, and that discipline is needed to keep early valuations at a healthy discount to the exit.

So the $3–5M band is held between the two forces. The size of the exit distribution pulls it up and funding discipline caps it. Real, large exits make an early entry worth paying for and the need to preserve a credible up-round stops you paying too much. And here’s the part that matters for what comes next: it’s the size of the exits in your market that sets where the upward force lands. A market with bigger and more frequent large exits will price early entry higher — the same company, worth more, purely because the tail it’s being measured against is fatter. Which is the whole subject of the next post.

Why you don’t pass on valuation alone

This leads to the practical consequence, and the way I read it as an investor.

If 10x already needs a top-decile exit and the entry band is set by funding dynamics rather than a precise valuation, then turning down a deal purely on valuation, within the normal range, is usually a category error. An angel who walks from a $5M pre-seed because their model says $3.5M is making a $1.5M precision call on a process whose every input is a wide band and whose target return is a tail event. Against all that, $1.5M is noise.

Valuation matters at the extremes. A founder asking $20M pre-seed for an idea-stage company with nothing to mitigate it is genuinely mispriced. They will be refused for being outside any defensible band. But inside the normal range, the variables that decide whether you make money are the team, the market, the competition, and the founder’s ability to execute — the separate question of how I select deals at all. The valuation number is less important than the story it encodes. It’s the price of the ticket, not the quality of the destination.

The field version of all of this is the doubling rule: the single heuristic that carries the whole model without re-running it. At seed, I want the valuation to roughly double at the next round inside twelve months. That’s the top of the seed band restated in plain English, it’s the uplift the funding-risk argument exists to protect, and it’s the one number I can check on a single deal without rebuilding the portfolio maths behind it. Understand the foundation once; trade off the rule of thumb after that.

The maths, in the end, is there to tell you what the number is made of. And most of what it’s made of is judgment.

The next post takes the same arithmetic somewhere less comfortable: valuation and pricing are not the same thing, and the same company prices differently depending on where in the world it’s raising. That gap has built returns for a generation of Australian angels, and it’s closing.

Richard Moore — MooCoo Ventures

Richard Moore is co-founder of MooCoo Ventures, an angel syndicate that co-invests alongside Brisbane Angels, one of Australia’s most active angel groups. He has made over eighty personal angel investments since 2013.

moocoo.vc