Cellebrite Cut Its Outlook and Its Chief Executive While a Japanese Parent Holds 42 Percent

Stake held by SUNCORPORATION, the Japanese parent

42.5%

Form 20-F, Feb 23, 2026

Cash left for shareholders after capex and all stock pay

$1.7M

First half 2026, our calculation

Share of revenue from recurring software

91.1%

Second quarter 2026, our calculation

1. Our call: Hold off, because the forensics toll is real but cash and leadership both slipped in 2026

Our call: Hold off. Cellebrite sells software that police, prosecutors, federal agencies and corporate investigators use to unlock, extract and analyze data from phones and other devices, then manage it as evidence ^[1].

It has about 7,000 customers, including about 5,300 federal, state and local agencies, and public sector customers gave more than 90% of 2025 revenue ^[1].

Software is almost all of it. In the second quarter of 2026, subscription services were $97.7M and term licenses $21.8M of $131.1M in revenue, so recurring software was 91.1%, our calculation ^[2].

Revenue grew 16% and annual recurring revenue, or ARR, reached $507.8M, up 21% ^[2].

We measure owner cash, the cash left for shareholders: operating cash flow, less capital expenditures, less capitalized software, less all stock pay. It was $115.4M in 2025, 24.3% of revenue, our calculation ^[9].

In the first half of 2026 it fell to $1.7M, from $29.9M a year earlier, our calculation ^[2].

The question this piece answers is whether a sticky government software franchise is worth owning while its parent, a Japanese public company, holds about 42% and the chief executive changed again in August. Our answer is not yet.

2. A subscription toll on digital evidence

2b. Cellebrite holds the evidence chain, but phone makers hold the door

A gate is valuable when traffic must pass through it and the owner can see and price what passes. For a police digital lab, the steps from seizing a phone to presenting its data in court often run through Cellebrite tools, and switching means retraining examiners and revalidating methods ^[1].

3. Growth was fast until this year

Operating income rose from $1.0M in 2022 to $66.5M in 2025, a 14.0% margin, our calculation ^[9]. Headcount grew from 1,005 at the end of 2022 to 1,285 at the end of 2025 ^[1].

4. The outlook cut came with the chief executive change

In February the company guided 2026 ARR of $567M to $573M and revenue of $565M to $571M ^[10]. In August it cut ARR to $550M to $560M and revenue to $555M to $561M, while raising its adjusted earnings target ^[2].

Second quarter ARR of $507.8M fell short of its own $510M to $513M guide from May ^[2].

The new chief executive, Shiven Ramji, blamed longer sales cycles and less expansion than expected from customers moving to Inseyets, the newer bundled platform ^[2]. He said resetting expectations now was the responsible approach ^[2].

The 21% ARR growth also includes Corellium, a virtualization software company bought on Dec 1, 2025 for an enterprise value of $170M, $150M of it in cash ^[11]. Corellium contributed $16.1M of ARR at year end, and excluding it ARR had grown 17% in 2025 ^[10].

Up to $30M more in cash is owed to its sellers if targets are met ^[11].

5. Costs grew faster than revenue

In the second quarter, revenue rose 15.8%, which the company rounds to 16%, while total operating expenses rose 21.9%, our calculation ^[2]. General and administrative spending rose 38.8% to $19.3M, and cost of subscription services rose 87.5% to $16.0M, our calculation ^[2].

The company says the subscription cost rise came from hosting, support, outside vendors and amortization of acquired technology, and subscription gross margin fell from 91% to 87% in the first half ^[3]. Operating income fell to $6.9M from $14.4M, and net income to $6.4M from $19.5M ^[2].

Stock pay rose 73.1% to $15.2M in the quarter, our calculation ^[2]. In September the board cut 3.5M shares from the equity plan reserve, and the company put potential dilution from all plans at about 9.2% ^[14].

6. Owner cash nearly vanished in the first half

Owner cash fell 94.2% year on year in the first half, our calculation ^[2]. Operating cash flow fell because receivables rose $38.1M in the second quarter, and deferred revenue fell $12.9M in the half after rising $3.3M a year earlier ^[2].

Some of the receivables gap may reverse if agencies pay on time in the second half, which is the bull case, but stock pay of $29.6M will not reverse ^[2]. Cellebrite also spent $15.6M buying intangible assets and $15.3M on an acquisition in the half, which we leave out of owner cash but note as real cash out ^[2].

The company still says it expects stronger free cash flow in the second half ^[2]. Its own free cash flow measure does not subtract stock pay, which is why it reported $144.2M of trailing twelve month free cash flow while our measure for the first half is close to zero ^[2].

7. The balance sheet is clean

Cash, deposits and marketable securities were $545.7M at Jun 30, 2026, up from $535.0M at year end ^[3]. We found no bank loans or notes on the balance sheet; total liabilities of $456.2M are mostly deferred revenue of $313.3M, cash customers have prepaid ^[16].

Book value was $538.1M ^[16].

The company pays no dividend and has not announced a buyback ^[1]. With no debt, the risk here is in the operations and the governance, not the balance sheet.

8. The main risks are access, reputation and a concentrated customer base

The first risk is access. If phone makers close the paths Cellebrite uses, its core tools lose value; the company names this risk directly ^[1].

The second is reputation: in late 2024 Amnesty International alleged that Serbian police misused its tools on journalists and activists, and the company no longer sells to China, Hong Kong, Russia, Belarus and certain other countries ^[1].

The third is the budget cycle. With more than 90% of revenue from public sector customers, a government spending freeze or slower agency procurement shows up quickly in ARR ^[1].

The fourth is control, covered in section 11.

9. A forensics leader building a broader investigation platform

Cellebrite, based in Petah Tikva, Israel, listed in the United States in 2021 through a merger with TWC Tech Holdings II, a special purpose company sponsored by True Wind Capital ^[1]. It now sells a platform that runs from collecting device data to analysis and evidence management, and it added Corellium to test and secure mobile apps on virtual devices ^[1][11].

Net retention of 117% in the second quarter, up two points from the first, shows existing customers still spend more each year ^[2]. The question for 2027 is whether the new products, Genesis, Guardian Investigate, drone forensics and advanced unlocks, add enough to restore 20% ARR growth ^[2].

10. Superintelligence tools are the bull case and the threat

Genesis, launched on Jun 10, 2026, is a superintelligence investigation assistant that sorts device data for examiners, and the company says it earned early revenue in the second quarter ^[2]. If it shortens case work, agencies may pay more per seat; the risk is that general superintelligence tools from larger vendors do the analysis once the data is extracted.

Our test is simple. If net retention holds at or above 117% and ARR growth returns toward 20% in 2027, the new products are working ^[2].

If ARR growth falls below 15%, the analysis layer is not lifting the business and the call moves toward Avoid.

11. Vote control, chief executive awards and the ties around the board

Vote control. Cellebrite has one class of ordinary shares, and every share carries the same vote ^[1].

But SUNCORPORATION, a Tokyo-listed company, held 106,065,324 shares, 42.5% of the ordinary shares at Feb 23, 2026 (42.7% on the smaller share count in its Dec 31, 2025 Schedule 13G/A), and the 20-F says it has significant influence over the election of directors, changes to the articles and any change of control ^[6][1]. The company calls it a controlling shareholder ^[1].

The parent has its own owner. In August 2024 True Wind Capital, the private equity firm led by Cellebrite chairman Adam H.

Clammer, bought 19% of SUNCORPORATION through a tender offer in Japan ^[12]. True Wind also holds 13,861,007 Cellebrite shares directly through TWC Tech Holdings II, 5.56%, and Clammer may be deemed to own 14,671,802 ^[1].

Adding True Wind's direct stake to 19% of the parent's 42.5% (both figures from the 20-F share count) gives an economic interest of about 13.6%, our calculation, if the 19% stake is unchanged ^[1][12]. True Wind's other managing member, James H.

12. The questions we would ask management, and the calls we are tracking

Question one: what owner cash, after stock pay, does the board expect in 2027? ^[2]

Question two: how much of second quarter ARR growth came from Corellium versus the core business? ^[2][10]

Question three: why are the chief executive performance units tied to stock returns rather than ARR or cash? ^[4]

Question four: does SUNCORPORATION or True Wind plan to change its stake or seek a sale? ^[1][12]

Question five: what share of new ARR came from Genesis and Guardian in the second half? That answer sets our conviction.

^[2]

Prediction one: Cellebrite posts 2026 ARR within its $550M to $560M range (65%). ^[2]

Prediction two: owner cash, as we define it, is below $90M for 2026 (65%). ^[2]

Bottom line: a strong niche with weak cash this year and a parent in control

Cellebrite earns 91.1% of revenue from recurring software and holds a real gate in digital evidence ^[2]. But owner cash fell to $1.7M in the first half, the 2026 outlook was cut and a new chief executive took over the same day, while a Japanese parent part-owned by the chairman's firm holds about 42% ^[2][6][12].

We would move to Buy if owner cash recovers above $100M a year with ARR growth near 20%, and to Avoid if ARR growth drops below 15%. Our call: Hold off, conviction 3 of 5.