IPO Lockup Analysis: What Spreadsheets Miss

TL;DR

IPO lockup expiry is one of the most consequential and least automated workflows in fund management
In March 2026, three lockups expired within eight days. Insider behavior (not share count or price trend) predicted the direction in all three cases.
78% of lockup dates in academic databases are off by at least one day (Brav & Gompers, Journal of Finance). One day of unexpected supply can move a stock 5-15%.
Figma’s lockup was not one event but four, spanning 13 months. A single “lockup: January 27” spreadsheet entry misses three of the four.
Portfolio-level lockup aggregation (not company-by-company tracking) is where most teams fall short

The March 2026 Cluster

Post-IPO lockup expiry is one of the most consequential and least automated workflows in fund management. In March 2026, three lockups expired within eight days at companies held by the same major fund. One stock surged 21%. One dropped 15%. The variable that predicted the direction in all three cases was the same, and it was not the one most teams track.

March 9: Klarna. 335 million shares released, representing 90% of total outstanding shares. The stock was already down 67% from its IPO price. But instead of crashing further, the chairman and chief product officer bought shares around the expiry. The stock held near its lows. Insider buying at lockup expiry is a rare and meaningful signal.

March 10: Figure Technology. 147 million Class A and 38 million Class B shares released. Insiders did not sell. The stock surged 21% on the lockup day itself, closing at $40 against a $25 IPO price. The market interpreted the lack of insider selling as a strong conviction signal.

March 17: Netskope. 390 million shares released. The CFO sold 65% of his personal holdings. The stock dropped 15%, despite an earnings beat earlier that quarter. When a CFO sells two-thirds of their position the moment they legally can, it overwhelms any positive fundamental data.

Three lockups. Three completely different outcomes. The variable that predicted the direction in all three cases was not the number of shares released, not the stock’s price trend, and not the earnings results. It was insider behavior around the expiry date.

The Figma Problem: When 180 Days Is Not 180 Days

The standard lockup narrative assumes a single date: 180 days after IPO, shares unlock, done. Figma’s lockup structure illustrates why that assumption breaks analysis.

Figma went public on July 31, 2025 at $33 per share. Its lockup was not one event. It was four:

Stage 1 (Day 36, September 5, 2025): Performance-based early release. 25% of locked shares became eligible for sale if certain conditions were met. This is not standard. Most analysts tracking a simple “180-day lockup” would miss this entirely.

Stage 2 (Day 107, November 14, 2025): Tied to the Q3 earnings release. Executives used this window to sell at $43 to $48 per share. The CEO had filed a 10b5-1 trading plan just four days after the IPO, signaling intent from the very start.

Stage 3 (Day 180, January 27, 2026): The standard 180-day expiry. By this point, the stock was already in decline, down from an all-time high of $142.92 on day two to below its IPO price.

Stage 4 (August 31, 2026): An extended lockup covering 54.1% of all outstanding shares. This tranche releases in stages, with 27.5% becoming available after Q1 2026 earnings results. More than half of Figma’s total shares remain locked until late summer 2026, a full 13 months after the IPO.

A team tracking Figma with a single “lockup: January 27” entry in a spreadsheet would miss three of the four events. They would not anticipate the September early release. They would not connect the November selling to the earnings-linked window. And they would not know that the largest tranche of shares is still restricted until August.

The Data Accuracy Problem

Academic research on IPO lockup dates reveals a finding that should concern every operations team relying on third-party databases: 78% of lockup dates in the SDC New Issues Database contain discrepancies of at least one day, according to research by Brav and Gompers published in the Journal of Finance. Subsequent studies by Field and Hanka corroborated the pattern, finding systematic errors in how databases parse lockup provisions from registration statements.

One day might sound trivial. It is not. Lockup expiry creates concentrated selling pressure on a specific date. If your model says shares unlock on Tuesday but they actually unlock on Monday, your hedging, your position sizing, and your risk management are all calibrated to the wrong day. In illiquid post-IPO names, one day of unexpected supply can move a stock 5 to 15%.

The error rate stems from how databases parse lockup language in S-1 filings. Some filings specify “180 calendar days.” Others say “180 days following the date of this prospectus,” which may or may not include the pricing date itself. Staggered lockups with performance conditions, earnings ties, and extended agreements compound the ambiguity. Automated parsers get it wrong. Human readers get it wrong less often, but they do not scale across dozens of positions.

The only reliable source is the primary document: the actual S-1 text. And parsing a 200 to 400 page registration statement to find, interpret, and correctly calculate every lockup date is precisely the kind of labor-intensive, error-prone work that consumes analyst hours for minimal glory.

What Insider Filings Tell You That Dates Do Not

The lockup calendar tells you when shares can be sold. Form 4 insider filings tell you who is actually selling and how much. The combination is where the real analytical value lives.

Consider the three March 2026 lockups again:

At Klarna, insiders bought. That is the strongest possible signal at a lockup expiry: the people who could finally sell chose to buy more instead. Any model tracking only the lockup date and share count would have predicted downward pressure. The insider filing data reversed that prediction.

At Figure Technology, insiders held. No significant Form 4 selling activity around the lockup date. The market read this as confidence, and the stock responded accordingly. Again, the lockup calendar alone would have flagged risk. The insider data neutralized it.

At Netskope, the CFO sold 65% of his personal stake. When a C-suite executive liquidates the majority of their holdings at the first legal opportunity, it signals something that no earnings beat can offset. The lockup date created the condition. The insider filing revealed the intent.

Tracking lockup dates without Form 4 correlation is like tracking earnings dates without reading the results. You know something is happening, but you do not know what it means.

The Portfolio-Level Challenge

Individual lockup tracking is manageable. Portfolio-level lockup management is where teams struggle.

A fund holding five recently public companies in March 2026 faced three lockup expiries in eight days with a combined float increase of 870 million shares. That is a portfolio risk event, not five individual company events. The correlation matters: if all three stocks decline simultaneously on lockup-related selling, the fund’s IPO book takes a concentrated hit in a single week.

Now extend this to a fund also tracking private portfolio companies preparing to go public. Cerebras filed its S-1 in February 2026 for an April IPO. Databricks signaled H2 2026. One company shelved its IPO entirely due to geopolitical tensions. Each of these transitions from private to public introduces a new lockup calendar, a new set of insider filing obligations, and a new monitoring cadence.

The operational reality: a fund managing the full lifecycle from pre-IPO due diligence through S-1 filing, pricing, first-day dynamics, quarterly earnings, and lockup expiry needs a unified system that connects all of these data points. Spreadsheets handle individual events. They do not handle the interaction between events, the insider filing overlay, or the portfolio-level risk aggregation.

What Better Tooling Looks Like

The gap between current practice and what is possible comes down to three capabilities:

S-1 parsing that captures complexity. Read the full registration statement. Extract every lockup provision, including staggered releases, performance conditions, earnings ties, and extended agreements. Calculate the exact dates from the filing language, not from a third-party database that gets it wrong most of the time.

Insider filing correlation. Monitor Form 4 filings in real time and overlay them on the lockup calendar. Flag when insiders are buying at expiry (bullish), holding (neutral), or liquidating (bearish). Automate the pattern recognition that the March 2026 cluster demonstrated.

Portfolio-level risk aggregation. Show the fund manager that three lockups expire in eight days with a combined 870 million shares released. Calculate the portfolio impact if all three sell off simultaneously. This is not analysis any individual company spreadsheet produces. It requires the system to see across positions.

The Bottom Line

Lockup analysis is a workflow that rewards precision and punishes approximation. The difference between “shares unlock in late January” and “25% of shares unlock September 5, executives sell in the November earnings window, the standard tranche expires January 27, and 54% of outstanding shares remain locked until August” is the difference between being surprised and being prepared.

The funds that got March 2026 right were the ones tracking insider filings alongside dates, watching CFO selling plans alongside share counts, and aggregating lockup risk across the portfolio rather than company by company. That is not a data problem. It is an infrastructure problem.

Frequently Asked Questions

What is an IPO lockup period?

After an IPO, insiders and pre-IPO investors agree not to sell their shares for a set period, typically 180 days. When that period ends, previously restricted shares become tradeable, potentially increasing supply and putting downward pressure on the stock price. However, the actual structure is often more complex than a single date.

Share count tells you the maximum possible supply increase. Insider behavior tells you the actual supply increase. In the March 2026 cluster, Klarna released 335 million shares (90% of outstanding) but insiders bought instead of selling. The stock held. Netskope released 390 million shares and the CFO sold 65% of his holdings. The stock dropped 15%. The intent behind the shares matters more than the number of shares.

How reliable are lockup dates from financial databases?

Research published in the Journal of Finance found that 78% of lockup dates in the SDC New Issues Database contain discrepancies of at least one day. Errors stem from ambiguous S-1 language (“180 days following the date of this prospectus” vs “180 calendar days from pricing”). The only reliable source is parsing the actual registration statement.

Can a lockup have multiple stages?

Yes. Figma’s 2025 IPO had four distinct lockup stages spanning 13 months: a performance-based early release at day 36, an earnings-linked window at day 107, the standard 180-day expiry at day 180, and an extended lockup covering 54.1% of shares until August 2026. Complex lockup structures are increasingly common, especially for high-profile tech IPOs.

What tools exist for automated lockup tracking?

EdgarTools (open source, MIT license) provides real-time SEC filing monitoring including Form 4 insider transactions. Commercial providers like MarketBeat and TikR offer lockup expiration calendars. For the S-1 parsing component, most funds still rely on manual extraction or custom NLP systems, as no standard automated solution accurately handles multi-stage lockup structures.

Sources: SEC EDGAR Form 4 filings (KLAR, FIGR, NTSK, FIG, WLTH), Figma S-1/A registration statement (lockup provisions), Guru Focus lockup analysis (Klarna 335M shares), MarketBeat lockup expirations tracker, TikR analysis (Figure Technology +21% lockup surge), Seeking Alpha institutional 13F data, Brav & Gompers (Journal of Finance) and Field & Hanka on IPO lockup date accuracy

Last updated: April 14, 2026

By BetterAI | We build custom AI research infrastructure for European investment firms. See how it works