The Pro-Rata Verdict: Exercise, Waive, or Sell the Follow-On

A portfolio company just opened its next round, and the offer notice is sitting in your inbox with a price, a date, and a 15-to-20-day clock already running. This guide is for angels, platform and talent partners at funds, and early-stage investors who hold an existing position and have to decide, before the window closes, whether to put more money in, let the pro-rata lapse, or sell into the round. It scores that one call from signals a small investor without board information can gather on the open web, and it ends with a verdict: exercise, waive, or sell.

Most public writing on pro-rata either defines the legal right or tells you to back your winners. Neither helps at the moment of decision. This is a decision procedure for grading a specific notice under a fixed deadline, not advice on whether to open a new position.

What the pro-rata verdict actually decides

The verdict is a three-way choice on an existing position's next-round notice: exercise the right and buy more, waive it and let it lapse, or sell existing shares into the round through a separate secondary process. These are three different instruments, and the most common error is treating them as one.

Pro-rata rights let you buy new shares in the next funding round. They are not the right of first refusal (ROFR), which lets others buy your existing shares when you want to sell. You cannot sell your pro-rata; you can only exercise it or let it lapse. If you want liquidity instead of more exposure, that is a secondary sale, and it runs on its own, longer clock. Keep the three options distinct from the first minute, because conflating them is how investors waive a right expecting to sell later, then discover the ROFR cascade has eaten the window.

The decision rests on four dimensions, scored in order: ownership math and return, public momentum, round quality, and your own reserve constraints. The first is a hard gate. The next two grade the company. The fourth grades you - whether you can afford the check without starving your ability to follow on elsewhere.

Why "back your winners" is the wrong default

The slogan is contradicted by its own data. In AngelList's 100,000 simulations, never following on beat always following on 54% of the time, while always following on won the other 46%. Because seed returns follow a power law, spreading capital across more positions raises your odds of catching the one outlier that carries the fund.

The two strategies fail in opposite directions. Always following on produced a higher mean TVPI, around $0.90 per dollar against never's $0.60, because it loads capital into the tail winners. But never following on produced a higher median, because it did not concentrate reserves into names that later disappointed. The exercise call, then, has to clear a real bar: it must beat simply indexing your reserves across more names.

Two numbers set the guardrails. Investors with three or fewer startups had a negative median portfolio value, so a small book that concentrates reserves is fragile by construction. And most funds reserve 40 to 60% of committed capital for follow-ons, with top-quartile managers clustering at 40 to 50%. Per company, reserve ratios run 1:1 up to 3:1, but most cluster right around 1:1. If you are holding more than one initial check in reserve for a single name before the round even opens, you are already overweight on conviction.

The four dimensions, and what each one proves

Score each dimension independently, then combine. A dimension is only useful if you also know what it looks like when it lies, so each one below carries its false positive.

Ownership math and return. This proves whether the incremental dollar can earn out. Compute the allocation first: pro-rata allocation equals your ownership percentage times new round size. Then test the entry price against a base-case exit. It lies when an up round reads as safety. Down rounds are now rare - under 14% of Q4 fundings, the lowest in three years - and median valuations are up at every stage, with a staggering 667% year-over-year rise at Series E+. So most notices arrive at richer prices, and the upside-to-exit test fails more often even for healthy companies. Scarce down rounds raise the exercise bar, they do not lower it.

Public momentum. This proves the company is expanding, not stalling. Read hiring velocity, senior go-to-market hires, new customer logos, and shipping cadence. It lies as momentum theater: headcount growth that is backfilling attrition rather than expansion. Check hires net of departures. In January 2026, VC-backed companies on Carta made 26,030 new hires against 20,378 departures, a 1.3x ratio - use that as your baseline and ask whether the new hires are senior and commercial, not just replacements.

Round quality. This proves the market's own verdict on the company. Read who is leading, up versus flat or down, the price against your entry, and priced versus bridge. It lies through signaling: if the original lead passes, that is data. Signaling risk refers to the message sent to new investors when an existing major investor chooses not to participate - so do not read your own exercise as independent. Ask who else is in before you decide.

Reserve constraints. This proves whether you can afford it without starving the rest of the book. Measure the allocation against remaining reserves and the 1:1-to-3:1 per-company ratio. It lies when conviction overrides breadth and you concentrate a small portfolio into one name.

The windows you are racing

Two clocks govern this decision, and they run at very different speeds. Knowing both lengths before you start is the difference between a considered verdict and a rushed one.

The exercise window is short. The NVCA model gives Major Investors 20 days after the offer notice to make the first election, then adds a 10-day window for fully participating investors to request unsubscribed securities. Real filings vary from 15 to 20 days. Eligibility is gated by Major Investor status, usually defined as holding 1 to 2% of fully-diluted equity, often with a minimum dollar purchase. Confirm you qualify before you plan around the right.

The sell window is long. Selling existing shares runs the ROFR gauntlet: the NVCA model agreement requires the seller to give notice at least 45 days before closing, with a 15-day exercise window for the company and roughly 10 more days for investors. That cascade outlasts the round you are selling into. So if the verdict is sell, you cannot start when the exercise window opens and expect to be done inside it. Start the ROFR process early or do not count on selling into this round at all.

The risk screen that overrides a good story

Run the risk screen independently of the momentum score, because strong growth routinely masks concentration risk that caps the exit. The two most common killers are customer concentration and insider signaling.

Customer concentration has published thresholds that tighten by stage. A single customer above 10 to 15% of revenue is a material risk flag, and above 20% many institutional investors will walk regardless of growth rate. The penalty is priced in: high single-customer concentration drives 20 to 30% valuation discounts from investors and acquirers. An up round does not erase this - check it independently.

Revenue quality sits alongside concentration: recurring beats one-time, and contracted beats month-to-month. A company clearing its concentration thresholds on month-to-month revenue is weaker than the headline suggests.

Insider signaling is the second screen. If the original lead passes on its own pro-rata, that message reaches every new investor, and it should reach you. Map who else is participating before you treat your own exercise as a standalone bet.

Reading momentum without board access

You do not need an information-rights seat to score momentum; the hiring graph is public and it moves before revenue does. Who a company adds, net of who it loses, is a leading indicator you can gather on the open web, and specialized hiring in particular clusters in named companies.

Refolk's index makes this concrete. Reading where senior engineers concentrate tells you both who is building momentum and where a scaling company can realistically hire. In Refolk's index of professional profiles, the United States holds about 1,085 Rust-skilled software engineers at the senior, staff, and SWE level against roughly 195 in Germany - a depth gap of about 5.6x, with the US pool concentrated among employers like Google, Oxide Computer, Shopify, Uniswap, and Helius.

Use this as a template, not a specific answer. For your own portfolio company, you want the same read: who joined in the last six months, from where, and at what seniority, measured against who left. That is the difference between expansion and backfill.

Refolk turns the hiring graph into a count you can act on inside the exercise window: recent senior hires, their previous employers, and departures, without a board seat. When the clock is 15 days, that is the difference between a momentum score and a guess.

The procedure

Work the notice in a fixed order. The first two steps are mechanical and fast; the middle three are the scoring; the last two are the screen and the verdict. Practitioners split on sequencing the scores: conviction-led funds like CRV lead with momentum, while quant-led angels lead with price. The order below leads with price because it is a hard gate - if the math fails, the softer scores do not get a vote.

Once the scores are in, the verdict falls out of two axes: whether the price clears the return test, and whether the company's momentum and round quality read as strong. The matrix below is the shape of the call.

The template below is a rubric you can paste into your notes and fill per notice. Score each line, then read the dominant pattern against the matrix.

Company: ___________   Notice date: ___________   Window closes: ___________
Major Investor status confirmed (1-2% FD): Y / N
Allocation (ownership % x round size): $___________
Reserve impact (per-company ratio toward 1:1-3:1): ___________
1. Return test - upside to base-case exit at this price: PASS / FAIL
2. Momentum - hires net of departures, senior/commercial: UP / FLAT / DOWN
3. Round quality - lead, up vs down, price vs entry: STRONG / MIXED / WEAK
4. Risk screen - single customer %, top-5 %, insiders passing: CLEAR / FLAGS
Verdict: EXERCISE / WAIVE / SELL
If SELL: ROFR notice started on ___________ (45+10+15 days ahead of close)

How this goes wrong

The failure modes below are where careful investors still land on the wrong verdict. Each one has a false positive - a reading that looks right and is not.

• Allocation math on the wrong base. Using as-converted when the agreement measures fully-diluted (or the reverse) changes your number. The false positive is believing you exercised fully when you under-subscribed. Check the IRA's calculation base: which securities count in the numerator and denominator, and is ownership measured as-converted or fully-diluted.
• Treating an up round as safety. A company can raise an up round and still be a dangerous bet if a single customer is above 20% of ARR. Growth does not erase concentration. Run the risk screen independently of the round grade.
• Confusing pro-rata with ROFR. Trying to sell your pro-rata is a category error. The false positive is waiving while expecting to sell later, then finding the ROFR cascade eats your window. Verify which agreement governs each right.
• Ignoring insider signaling. If the original lead passes, that is information, not noise. Do not score your own exercise as independent; map who else is in first.
• Underpricing the secondary discount. The headline secondary price is not the realized price. Build in 5 to 15% ROFR friction and 30 to 60 extra days before you shop the shares. A $4 negotiated price can close at $3.80 or lower.
• Momentum theater. Headcount growth can be backfilling attrition rather than expansion. Check hires net of departures against the 1.3x baseline, and confirm the new hires are senior and commercial, not replacements.
• Chasing mean TVPI. Always following on wins the top tail but loses more often; a small portfolio that concentrates reserves can post a negative median. Size reserves to portfolio breadth, not conviction alone.

Before you file the election

Run this check before you deliver a written election. The window is unforgiving, and a verdict that skips any line below is a guess wearing a scorecard.

Keeping the verdict current

The thresholds in this guide are durable, but the market context that sets the exercise bar is not. Re-check two things before each decision rather than trusting the last number you saw.

First, the valuation and down-round environment. The 667% Series E+ jump and the sub-14% down-round rate reflect one quarter's state of private markets; both move, and they move the upside-to-exit test. Re-read the current quarter's data before you set your base case, because a looser market lowers the bar and a tighter one raises it.

Second, your own reserve position. The 40-to-60% reserve range and the 1:1 per-company default assume a certain portfolio breadth. As your book grows or contracts, the point at which a single follow-on overweights you shifts. Recompute remaining reserves against live portfolio count each time, not from the plan you wrote at fund formation. The verdict is only as good as the two inputs that change underneath it: what the market will pay, and what you can afford to commit.