The IPO Lockup Expiry Sourcing Playbook: Filing to Dated Queue

When a company goes public, its employees do not get liquid on IPO day. Their shares stay locked for a disclosed window, and the moment that window opens is the moment the golden handcuffs loosen. This guide is for in-house recruiters, sourcers, and founders who want to time outreach to that window, and it delivers a repeatable method for turning the publicly filed lockup calendar into a dated, ranked queue of reachable, lockup-covered employees, with each person's outreach scheduled to the specific expiry date read from the company's own filings.

Most passive-sourcing advice treats timing generically, telling you to reach people "in the window" without saying how to find the window. Lockup content aimed at investors tells you how to trade the stock, not whom to recruit. This playbook sits in the gap: it names the exact filing fields that yield the expiry date, it handles staggered and early-release structures that break naive date arithmetic, and it schedules first contact to a known calendar date rather than waiting for a job title to change on a profile.

Why lockup expiry is a sourcing trigger, not a trading one

An IPO lockup is a contractual promise by pre-IPO shareholders not to sell for a set period after listing, and its expiry is the first date those people can convert paper equity into cash. That conversion is the recruiting trigger. A senior engineer who has been unable to sell a single share suddenly can, and the single largest financial reason to stay put, unvested or unsellable equity, weakens.

The investor reads this event for price. The recruiter should read it for people. Studies consistently find that selling pressure clusters at expiry, but the price effect is modest. US work in the Field and Hanka lineage finds abnormal returns of -1% to -3% around expiry. A VC-focused analysis found -1.61% over five days, with venture-backed firms losing 3% to 4%, particularly high-tech companies. A Nordic sample of 141 IPOs from 2009 to 2017 showed a cumulative abnormal return of -1.1%. The signal you want is not in any of those price numbers.

The signal is volume. A German Neuer Markt study found a roughly 25% jump in trading volume around lock-up expiration. That spike is insiders converting equity to cash, and it is the behavioral proof that the golden handcuffs just came off. Watch volume and Form 4 insider-sale disclosures, not the ticker.

Which filing fields give you the expiry date

The authoritative source is the IPO prospectus, and within it two named sections carry everything you need. Shares Eligible for Future Sale tells you how many shares unlock and when. Underwriting tells you who is bound and what permits an early release. The exact expiry language lives in the lock-up letter, filed as an exhibit to the underwriting agreement.

You will encounter the same provisions in three documents, and they do not always agree, so pick an authority and stick to it.

Sources disagree on which document to trust: some practitioner write-ups point to the S-1, while others insist on the 424B4 filed at pricing. Use the 424B4 as authoritative and the S-1/A for history. Pull every one of these through EDGAR full-text search for free; there is no need for a paid tracker, and investor trackers drift because they estimate earnings-linked dates.

Most US lockups run between 90 and 180 days, with 180 by far the most common. But the duration is the least interesting field. The fields that decide your schedule are the staggered release dates and the price triggers, and those are where naive date math fails.

Staggered releases and price triggers that break the headline date

A single expiry date is often the wrong date to schedule against, because roughly one third of one studied sample had more than one lockup period and about 10% had two or more. On a staggered deal, shares free in stages, by date, by price threshold, or by an earnings-linked cliff. If you schedule outreach to the headline 180-day expiry, you miss the first cliff that may have freed a large block months earlier.

Price-trigger early-release clauses are explicit in recent filings, and they share a recognizable shape: a percentage above the IPO price, sustained across a count of trading days, after an earnings release.

SpaceX is the cleanest worked example of an accelerated staggered structure, and it shows why counting to day 180 would mislead you badly. Its first cliff unlocked 20%, roughly 911.5 million shares, two trading days after Q2 earnings. On top of that sat time-releases of up to 7% each at days 70, 90, 105, 120, and 135, and a performance tranche of 10% that released only if the stock held at or above $175.50, 30% above the $135 IPO price, on 5 of 10 days. Meanwhile the CEO's own shares were locked 366 days with no early release.

The lesson is operational: enumerate every tranche date and every price threshold, and set your first-contact date to the earliest real release, not the last one. A company whose "expiry" is in December may have had sellable, newly liquid employees since August.

Who is actually locked, and how tenure proxies for it

The lockup binds only pre-IPO holders, so the usable filter is join-date, not job title. Lockup agreements typically cover anyone who held equity before the company went public: founders and executives, employees with vested stock options or RSUs, early investors, and other pre-IPO shareholders. A software engineer who joined after listing is generally not locked and holds little, which makes post-IPO hires the classic false positive on these lists.

No public dataset maps individual holdings, so you proxy with tenure. Standard grants vest over four years with a one-year cliff, so pre-IPO joiners with two to four or more years of tenure hold the most sellable equity and feel the most pull when it becomes liquid. Former employees are frequently still covered too; the standard advice to ex-employees is to keep contact information current with the company precisely because that is how they learn whether they are in the lockup.

This is where the sourcing problem becomes a people-search problem: you need everyone at a named company who joined before a date and is senior enough to hold meaningful equity. Refolk lets you express exactly that constraint in plain English, including the pre-IPO join-date cut and the current-employer filter, so you get a ranked pool back instead of scrolling filters.

Scale matters when you are building a universe across many recently-public companies. In Refolk's index, a search for US-based software engineers returns 352,498 people, and the top current employers in that sample include Figma, which went public in July 2025, alongside Google. That is the raw pool you narrow by company and join-date.

One honest limit on that pool: tighter cuts do not always populate. A query for UK senior software engineers, and one for US engineering directors in the Computer Software industry with seniority filters applied, returned no populated aggregate in Refolk's index at query time. Treat narrow country-and-seniority splits as not established via that cut, and widen the query before you conclude a pool is empty.

The procedure, start to finish

Run these seven steps in order. The roles and time budgets are the ones this method assumes; adjust them to your team, but do not skip the filing read in step two, which is the step that makes everything downstream dated rather than guessed.

The output of this procedure is a spreadsheet where every row is a person, every person carries a first-contact date, and every date cites a specific filing field. That is the artifact no generic sourcing method produces.

A locked engineer is a prospect with a known unlock date. Schedule to the date, not to a hunch.

The greenshoe, the waiver, and the blackout

Three mechanics around the expiry regularly trip up people who have not read the filings carefully. Each one either does not move the date, or moves it in a way you can only see in a disclosure.

The greenshoe, or over-allotment option, gives underwriters the right to sell roughly 15% more shares than planned and to decide over the following 30 days whether to buy those extra shares from the company at the offer price. It changes total float. It does not change the lockup date and it is not new insider supply. If you see shares appear, confirm whether they came from the greenshoe or from a lock-up release before you re-date anything.

The underwriter waiver is the mechanism that genuinely moves the sellable date earlier. Underwriters can release shareholders from the lockup early. This is uncommon, but when it happens it is disclosed: FINRA Rule 5131 requires the book-runner to announce a release or waiver for officers and directors through a major news service at least two business days before it takes effect. That two-day notice is your early-warning system, and it is why step six exists. dLocal's F-1/A, for one, defines an "Early Lock-Up Expiration Date" against a "Final Lock-Up Expiration Date", the earlier of a second-earnings day or 180 days, announced at least two trading days ahead.

The blackout period is the quietest trap. Even after expiry, companies impose recurring quarterly blackouts, typically 30 to 60 days before earnings, that suppress the moment liquidity actually matters. If you schedule outreach into a blackout, you have hit the calendar date but missed the behavioral window. Map earnings dates for each company and avoid scheduling into a blackout.

One more confirmation source closes the loop. Rule 144 caps how much an affiliate can sell, at the greater of 1% of shares outstanding or the four-week average weekly volume. In one filing, Forbright's, that 1% figure worked out to about 289,393 shares. You do not need to compute this, but it explains why insider selling appears as a steady stream of Form 4 filings rather than a single dump, which is exactly the signal you monitor.

How this goes wrong

This method fails in predictable ways, and every failure has a check. Treat the list below as the part of the playbook you reread before trusting a queue.

Two of these deserve extra weight. Date arithmetic is the cardinal error: counting 180 days from the IPO misses staggered first cliffs that unlock months earlier, and produces a company you believe is locked while its insiders are already selling. SpaceX had sellable tranches from roughly day 70, long before its headline December expiry. The only fix is to read the lock-up exhibit.

Over-trusting investor trackers is the subtler error. Third-party trackers estimate earnings-linked dates, and they drift because earnings dates are themselves estimates until confirmed. Reconcile every date against the primary filing, every time. A tracker is a starting hypothesis, not a source.

And remember the price trap, listed as its own failure mode because it is so intuitive: assuming a price drop means people are selling. The abnormal returns are small, -1% to -3%, so price barely moves. The recruiter's signal is the roughly 25% volume spike and the Form 4 filings, not the ticker.

Across these four studies, the expected selling-pressure signal brackets at roughly -1% to -4% on price and +25% on volume. Build your monitoring around volume and insider filings, which carry the signal, and ignore the price, which does not.

Keeping the queue current

A lockup queue is a living document, not a one-time export, because dates move and new IPOs arrive weekly. Treat the maintenance loop as part of the method. Before you call any cohort ready, run the checklist below.

To keep the universe fresh, re-run step one weekly against new listings, and re-run step six daily as a lightweight alert. When a waiver or early-release announcement lands, re-date the affected rows immediately, since FINRA's two-business-day notice gives you a short head start that evaporates fast. For the people side, refresh tenure cuts as companies hire and attrit; a profile that was a post-IPO hire last quarter may now clear your pre-IPO filter if you are targeting a later-listed company, and vice versa.

The named primary sources you return to are small and free: EDGAR full-text search for every 424B4, S-1, and 8-K; Form 4 for confirming who is actually selling after expiry; and FINRA Rule 5131 for the waiver-disclosure rule. Everything in this playbook is reconstructable from those three, which is the point. You are not buying a calendar. You are reading the one every public company is already required to file, and turning it into a dated recruiting queue before anyone else bothers to.