Figma, Inc.
FIG · New York Stock Exchange · Technology, application software · Reporting currency USD · As of 9 August 2026
A category-defining product with a narrow moat, priced as though the moat were wide. The business passes; the price and the pay structure do not.
Moves to BUY at or below USD 12.50, or without any price fall if stock based pay drops below 20% of revenue on a trailing four quarter basis, which lifts the range itself. Objective is capital growth over a three to five year holding period. Allocation band 0 to 2% of a growth allocation, set by volatility and the bear case rather than by liquidity.
| Metric | Value | Metric | Value |
|---|---|---|---|
| Market capitalisation | USD 12.41bn | EV / sales, trailing | 8.4x |
| Diluted shares | 542.6m | EV / gross profit, trailing | 10.7x |
| Net cash, 30 Jun 2026 | USD 1,599m | Gross margin, trailing | 79.1% |
| Revenue growth, Q2 2026 | 48% | Net dollar retention | 136% |
| Stock pay, H1 2026 | 45% of revenue | Free cash flow margin | 20.5% |
| Daily traded value | USD 535m | Float sold short | 26.9% |
Source: Figma Q2 2026 results release and Form 10-Q for the quarter ended 30 June 2026 (Tier 1); market data from S&P Global Market Intelligence via stockanalysis.com, 7 to 9 August 2026 (Tier 4). All figures in US dollars, Figma's reporting currency, so no exchange rate is applied.
Fair value from a discounted cash flow in which stock based pay is charged as a cost rather than added back. The anchor is the base case of USD 19.19, not the midpoint of the range. Calculated by the author.
Three reasons to look
- Growth is still accelerating at scale. Revenue grew 48% in the June quarter, every quarter since listing has beaten the company's own range, and full year guidance has been raised twice to USD 1.463bn to 1.467bn.
- Customers spend more each year without being sold to. Net dollar retention of 136% means the existing base alone grows the business by more than a third before a single new logo is added.
- The balance sheet removes financing risk. USD 1.60bn of net cash and no borrowings mean Figma never has to issue equity into a weak price, which is why this is a WATCH rather than an AVOID.
Three major risks
- AI agents may remove the design file entirely. A manager who describes a screen and receives working code never opens a design tool. That is the bear case, and it values the shares at USD 6.53.
- Two fifths of revenue is paid to staff in new shares. Stock pay ran at 40% of revenue in the June quarter and the count has risen about 9.7% a year since listing. The whole valuation turns on that falling.
- The margin is going the wrong way. Gross margin fell from 89% to 84% in twelve months and free cash flow margin from 27% in March to 14% in June, as AI inference costs land.
Business overview
Figma sells a browser-based tool where teams design software interfaces together. The file lives on a server rather than a laptop, so a designer, an engineer and a product manager can open the same screen at the same time. Revenue comes from seats sold by subscription, increasingly with usage-based AI credits on top. Customers do not buy a drawing tool so much as a shared place where a company's design system lives: the component library, the colour tokens and the conventions every file points at.
The company listed on the New York Stock Exchange on 31 July 2025 after the European Commission signalled it would prohibit Adobe's USD 20.0bn takeover and the UK Competition and Markets Authority found against it, so the deal was abandoned in December 2023. Guided 2026 revenue is USD 1.465bn at the midpoint.
Figma is moving from seats to seats plus credits. That is a defensive move as much as a commercial one: if AI reduces the number of people who open a design tool, revenue tied to headcount falls, and revenue tied to usage does not. Over 80% of customers spending above USD 10,000 used AI credits weekly by June 2026.
| Line | What it is |
|---|---|
| Seats | Per-user subscriptions across Figma Design, FigJam, Slides, Sites and Buzz |
| AI credits | Usage-based charges for Figma Make, Code Layers and the design agent |
| Enterprise plans | Annual contracts with the design system, admin controls and single sign-on |
| Deferred revenue | USD 626.8m collected and not yet delivered, at 30 June 2026 |
Source: Figma Q2 2026 results release and Form 10-Q, 5 August 2026. Tier 1.
Industry, moat and disruption
Figma put its own addressable market at USD 33bn in July 2025. Morgan Stanley used about USD 26bn in March 2026 and said Figma had reached 4% of it. Both come from interested parties. Against guided revenue of USD 1.465bn, the claimed market is 18 to 23 times the company's own sales and two to three times an independent measure of the category, which third-party research put at USD 10.5bn in 2022. Figma's figure counts the workforce doing design work, not money spent on tools. That is headroom, not a revenue pool, and neither number enters the model here.
| Competitor | Status | Scale | Where it competes |
|---|---|---|---|
| Adobe | Listed, ADBE | USD 25.2bn revenue | Creative tools broadly. Its own interface tool, Adobe XD, lost the category to Figma |
| Canva | Private | USD 42bn valuation | Marketing and social design, moving up into teams. Different job, overlapping buyer |
| AI coding agents | Mixed | Claude Code above USD 2.5bn run rate | The prompt-to-interface path that skips the design file entirely |
Sources: Adobe revenue from S&P Global Market Intelligence via stockanalysis.com (Tier 4); Canva valuation from an employee share sale reported by Bloomberg, August 2025 (Tier 3); Claude Code run rate reported February 2026 (Tier 3). Sketch and Framer remain in the market at small scale.
Share, barriers and pricing power
Within interface design, expert surveys put Figma's share at roughly 80% to 90%. What holds a team is not the drawing tool but the design system, and rebuilding one costs months. On pricing power the evidence is mixed. Figma raised the price of Figma Design in March 2025 and kept its retention rate, but it has not passed the AI compute cost through: gross margin fell five points in twelve months, and a company with full pricing power does not absorb an input cost that size.
The mechanism is switching costs from the design system plus a network effect inside each team. Not a licence, not scale, not brand. It is narrow rather than wide because what is defended is the design file, and the plausible disruption removes the file entirely. Anthropic's chief product officer resigned from Figma's board on 14 April 2026, the same day reporting emerged that its next model would carry design tooling, and Figma spent July 2026 buying the team behind a vibe-coding startup.
Regulation is light-touch and the one event that mattered has already happened. A rescue bid from a large incumbent is unlikely to clear a competition review, so no takeover floor sits under this price.
Financial performance
| Line | 2024 | 2025 | H1 2026 | 2026E | 2027E | 2028E |
|---|---|---|---|---|---|---|
| Revenue growth | 48% | 41% | 47% | 39% | 28% | 23% |
| GAAP gross margin | 90% | 88% | 82% | 83% | 83% | 84% |
| GAAP operating margin | -122% | -122% | -36% | -27% | -15% | -5% |
| Adjusted operating margin, company basis | 17% | 12% | 13% | 9% | 11% | 13% |
| GAAP net income | -732 | -1,300 | -254.6 | -410 | -300 | -140 |
| Free cash flow | 204 | 243 | 142 | 220 | 338 | 484 |
| Stock pay | 948 | 1,364 | 317 | 615 | 638 | 623 |
| Stock pay, share of revenue | 127% | 129% | 45% | 42% | 34% | 27% |
| Economic free cash flow, after stock pay | -744 | -1,121 | -175 | -395 | -300 | -139 |
Sources: Figma results releases Q2 2025 to Q2 2026 and the Q2 2026 Form 10-Q (Tier 1); Form S-1 for 2024 (Tier 1). Full year stock pay for 2024 and 2025 is this desk's derivation from the gap between reported and adjusted operating profit, including USD 889.3m on the tender offer and USD 975.7m on the listing. Columns marked E are this desk's estimates.
The 2024 and 2025 bars are capped at 60% because the true values, 127% and 129%, run off the scale. Forecast bars are this desk's assumption and carry the whole valuation. Reported figures from Figma results releases, Tier 1.
Figma reported USD 141.8m of free cash flow in the first half of 2026, but paid USD 161.6m to tax authorities to settle employee share awards, a payment recorded in financing rather than operating. Net of it the company consumed USD 19.8m of cash, and USD 174.8m after the full stock charge. Return on invested capital was -65% on trailing figures against a 10.0% cost of capital.
The growth is earned rather than bought. Figma's four acquisitions were team hires with immaterial terms, and cash paid for business combinations was USD 21.0m across 2025. Capital investment is about 1% of revenue, so the usual question of whether growth destroys value does not apply here. The question is who receives the growth: the owner, or the employee holding new shares.
Balance sheet, dilution and the downturn test
| Item | USD m | Comment |
|---|---|---|
| Cash and cash equivalents | 445.8 | Immediately available |
| Marketable securities | 1,221.3 | Short dated, available for sale |
| Digital assets | 25.9 | Bought for USD 30.0m in 2025, carried lower after USD 5.0m of remeasurement losses in the half |
| Deferred revenue | 626.8 | Cash already collected for services not yet delivered |
| Borrowings | Nil | Lease liabilities of USD 67.8m are the only debt-like item |
| Total equity | 1,477.9 | Against an accumulated deficit of USD 1,697.9m |
Source: Figma Q2 2026 results release, condensed consolidated balance sheet, 5 August 2026. Tier 1.
The downturn test
Assume growth stops dead: revenue flat at the guided USD 1,465m, gross margin down to 80% as AI compute costs rise, and operating costs held at the annualised first half 2026 level of USD 1,658m. That gives a GAAP operating loss near USD 486m, but roughly USD 615m of that cost is paid in shares. On a cash basis the company stays modestly positive and the net cash is untouched. Figma cannot be forced into a financing at a bad price, which is why the rating is WATCH rather than AVOID.
Sources: Figma Form 10-Q cover pages for the quarters ended 30 September 2025 and 31 March 2026 (Tier 1 to 2); current count from S&P Global Market Intelligence via stockanalysis.com, 9 August 2026 (Tier 4). The July 2025 step from 217m to 493m is the listing conversion of preferred stock, not dilution of a public holder.
The count went from 493m to 533m in ten months, 8.2% in the period and about 9.7% a year, all of it employee share issuance. A holder buying today owns a shrinking share of the company every quarter unless that rate falls. No dividend has been declared and no buyback announced, so the company is a net issuer.
Cash conversion is deteriorating as the AI costs land. The free cash flow margin was 23% for full year 2025, 27% in the March 2026 quarter and 14% in the June quarter. Management attributes the swing to inference costs and to Config, the annual conference. Both recur.
Ownership, control and governance
Source: Figma Form 10-Q cover page for the quarter ended 31 March 2026, as at 11 May 2026 (Tier 2). Class A 445.7m shares, Class B 82.7m shares. Voting percentages calculated by this desk. A Class C with no votes is authorised but none is outstanding.
A block holding 15.7% of the economics controls 73.6% of the votes, and the chief executive directs most of it. At the July 2025 listing the prospectus stated that Dylan Field would hold or control approximately 74% of total voting power, including shares voted under an irrevocable proxy granted by co-founder Evan Wallace and the Wu-Wallace Family Trust over 26.7m Class B shares. Outside holders cannot change the board, force a sale, or block one. The structure is common among recent listings and still a real cost.
Insider dealing and the incentive structure
Field has sold under a Rule 10b5-1 plan adopted 4 August 2025: 312,500 shares in November 2025 near USD 34 to 36 and 486,930 in February 2026 near USD 30 to 32. The chief technology officer sold 261,000 shares for USD 6.6m and the chief financial officer filed to sell 30,000 in July 2026. One aggregator counts insiders as net sellers of about USD 116m over the year. Scheduled selling under a pre-set plan is the least troubling kind.
The compensation fact that matters is in none of those filings. Stock pay ran at 40% of revenue in the June quarter. The incentive that creates is to grow revenue, because the cost of the people doing it is charged to shareholders rather than to the income statement anyone quotes.
Board, audit and a note on the leadership
All directors were re-elected and the auditor ratified on 2 June 2026, and nothing in the filings reviewed carries a qualification, restatement or late filing. Two things sit outside that record. Mike Krieger, Anthropic's chief product officer, resigned from the board on 14 April 2026. And with the June results the company said its chief technology officer would move to a chief architect role with a successor search open, its chief product officer would leave after seven years, and its chief marketing officer at the end of August: three senior roles in one announcement, in the quarter margin guidance fell.
Valuation
Model. A discounted cash flow in which stock pay is a cost rather than an add-back. An earnings multiple has nothing to attach to, since the company loses money under normal accounting rules, and a revenue multiple makes the answer a guess about sentiment. Figma hands two fifths of its revenue to staff in new shares, so the only question worth modelling is how much of each revenue dollar reaches the owner.
| Input | Used | Derivation |
|---|---|---|
| Risk free rate, equity risk premium | 4.60%, 4.23% | US 10 year Treasury at 7 August 2026, and Damodaran's implied S&P 500 premium |
| Beta | 1.27 | No five year beta exists: the shares have traded twelve months. Bottom-up peer average of Adobe 1.40, Atlassian 1.04, Datadog 1.51 and monday.com 1.12. All four carry negligible net borrowings, and Figma holds net cash, so relevering leaves 1.27 |
| Cost of capital | 10.0% | 4.60% plus 1.27 times 4.23%. No borrowings, so the weighted cost equals the cost of equity |
| Revenue 2026E, growth to 2035E | USD 1,465m, 28% to 6% | 2026 is the guidance midpoint. Growth then runs 28, 23, 19, 16, 13, 11, 9, 7.5 then 6%, the 2027 rate matching the exit rate implied by fourth quarter guidance |
| Free cash flow margin | 18% to 32% | Rising toward a mature software level below Adobe's 40.8%. Cash tax sits inside this path, normalising toward 20% by 2031 |
| Stock pay path | 34% to 9% | Falling as pre-listing awards vest out. The model's most important assumption |
| Terminal growth and margin | 3.0%, 23% | Growth slightly above long-run nominal output, justified by the recurring base. The margin is 32% free cash flow less 9% stock pay |
| Diluted shares and net cash | 542.6m, USD 1,599m | Q2 2026 adjusted diluted weighted average, held flat because charging stock pay as a cost and diluting the count too would count it twice. Net cash is securities and digital assets less lease liabilities at 30 June 2026 |
Sources: Trading Economics (Tier 4); Aswath Damodaran, Data Update 2 for 2026 (Tier 3); peer betas from S&P Global Market Intelligence via stockanalysis.com (Tier 4); guidance and balance sheet from the Figma Q2 2026 release (Tier 1). Growth, margin, stock pay and terminal assumptions are this desk's.
| Component | USD m |
|---|---|
| Present value of economic free cash flow, H2 2026E to 2035E | 1,697 |
| Present value of the terminal value | 7,119 |
| Enterprise value | 8,816 |
| Add net cash at 30 June 2026 | 1,599 |
| Equity value | 10,415 |
| Divided by 542.6m diluted shares | USD 19.19 |
This desk's model. No minority interests to deduct, and net cash is the reported 30 June 2026 position with no post-balance-sheet adjustment, so the market profile above and this bridge sit on the same basis. 81% of the value sits in the terminal period.
Source: S&P Global Market Intelligence via stockanalysis.com, 7 August 2026 close. Tier 4. Earnings multiples are omitted because three of the five carry no meaningful trailing earnings.
| Company | Market cap | EV / sales | Gross margin | FCF margin | 3yr growth | 12m price |
|---|---|---|---|---|---|---|
| Figma | USD 12.4bn | 8.4x | 79.1% | 20.5% | 28.6% | -74.2% |
| Adobe | USD 105.4bn | 4.2x | 89.4% | 40.8% | 9.8% | -23.3% |
| Atlassian | USD 37.8bn | 5.8x | 85.6% | 20.1% | 14.7% | -19.6% |
| Datadog | USD 84.0bn | 20.2x | 79.5% | 29.8% | 25.1% | +70.8% |
| monday.com | USD 4.0bn | 2.3x | 89.1% | 23.6% | 17.2% | -64.0% |
Source: as above, in US dollars. EV to gross profit and the derived comparisons are this desk's calculation.
Both sides of the peer comparison
The row that argues against this note. Datadog trades at 20.2 times sales with slower forecast growth and a similar gross margin, so Figma at 8.4 times looks cheap. Three things separate them: Datadog converts 29.8% of revenue into free cash flow against Figma's 20.5% and falling, its share count rose 1.7% against Figma's near 10%, and it has risen 71% while Figma fell 74%.
The row that argues for it. monday.com is the closest analogue, a seat-priced collaboration tool selling to the same buyer and now moving to seats plus credits as Figma has. It trades at 2.3 times sales, which is what the market pays once it concludes AI shrinks the seat count.
Hold the growth path and the 10.0% cost of capital fixed and solve for the margin that justifies today's USD 23.29. The answer is 27.5% against the 23% assumed here: the price needs Figma to keep 27.5 cents of every revenue dollar after paying staff, permanently. No peer manages that above 20% growth. A second cross-check on multiples gives USD 16.45 at the 5.0x peer median and USD 24.55 at 8.0x, on the same guided revenue.
Analyst coverage, and where this note differs
Seven houses are named with actions since 7 July 2026: four Buy, three Hold, on a low of USD 28.00, a mean of USD 32.00 and a high of USD 37.00. Citi raised to 37 on 7 August, Morgan Stanley cut to 33 on 6 August, RBC raised to 28 on 6 August and Bank of America reinstated at 30 on 7 July. Counts disagree by provider: S&P Global has 14 at a mean of 30.50, Benzinga 15 at a stale 43.58, Public.com 8. This anchor sits about 40% below the mean of the four confirmed targets, and all of the gap is one assumption. The street values Figma on adjusted operating income and forward sales, which add stock pay back. This note charges it.
| Fair value | Per share | Assumption at that end |
|---|---|---|
| Low | USD 15.52 | Terminal margin 19%: stock pay settles near 13%, not 9% |
| Midpoint, the anchor | USD 19.19 | Terminal margin 23%, the base case |
| High | USD 22.87 | Terminal margin 27%: stock pay falls to 5%, conversion reaches Adobe's level |
The base case discounted cash flow of USD 19.19 is the anchor, the only candidate with a printed assumption behind every input. The probability weighted USD 19.52 and the bear-to-bull midpoint of USD 24.75 are not used. Margin of safety 35%, wide because of a twelve month trading history, no five year beta, 81% of value in the terminal period, and substitution rather than cycle risk. Preferred entry USD 12.50.
Catalysts, scenarios and sensitivity
| Catalyst | Window | Direction | What it is worth |
|---|---|---|---|
| Extended lock-up fully expired | 7 August 2026, done | Negative | 77.7m Class A shares became sellable, 39% of the float. No value effect, but the supply is live |
| Equity grants and roll-off of pre-listing awards | Q4 2026, Q1 2027 | Either way | The recurring cost the thesis turns on. Each 5 points of stock pay is worth about USD 4.60 |
| Q3 2026 results, then Q4 with FY2027 guidance | Q4 2026, Q1 2027 | Either way | First test of the guided slowdown to 36%, then the year's key data point. This note assumes 28%; below 25% moves the anchor to the low end |
| Leadership: CMO departs, CTO successor named | End August 2026, then unannounced | Either way | Third senior change in one quarter. AI engineering reports to the chief executive until filled |
Sources: lock-up mechanics from the Q2 2025 release (Tier 1); leadership and Config from reporting on the Q2 2026 call (Tier 3); guidance from the Q2 2026 release (Tier 1). Magnitudes are this desk's. No scheduled licence auction, tariff review or statutory obligation falls in the window, and S&P 500 inclusion needs profitability Figma lacks.
Vertical marker = current price USD 23.29
Bear: AI agents take the design step, growth 22% fading to 3%, conversion stalls at 26%, stock pay reaches only 10%, cost of capital 11.5%. Base: growth 28% fading to 6%, conversion reaches 32%, stock pay falls to 9%, cost of capital 10.0%. Bull: Figma owns the AI design layer, growth 34% fading to 8%, conversion reaches 36%, stock pay falls to 7%, cost of capital 9.0%. Returns annualised over four years from the USD 23.29 close. This desk's model.
| Cost of capital | m 15% | m 19% | m 23% | m 27% | m 31% |
|---|---|---|---|---|---|
| 8.50%, beta about 0.92 | 15.82 | 20.60 | 25.38 | 30.16 | 34.94 |
| 9.25%, beta about 1.10 | 13.58 | 17.74 | 21.90 | 26.06 | 30.22 |
| 10.00%, beta 1.27, base | 11.85 | 15.52 | 19.19 | 22.87 | 26.54 |
| 11.50%, beta about 1.63 | 9.35 | 12.31 | 15.28 | 18.24 | 21.20 |
The grid runs the cost of capital against the terminal economic margin, what Figma keeps of each revenue dollar after paying staff, rather than terminal growth. Computed by this desk.
Reaching today's USD 23.29 needs either a terminal economic margin of 27% or better at the base cost of capital, or a cost of capital near 8.5% on the assumed 23% margin. The first requires stock pay near 5% of revenue, the second requires a beta of about 0.92 on a stock that fell 74% in twelve months and moved 15% in one session.
Risks and invalidation
| Risk | How it damages the thesis | Fair value impact | Leading indicator | Severity |
|---|---|---|---|---|
| AI agents replace the design file | Prompts produce working software, so seats stop growing | USD 6.53, -72% | Retention, count of USD 10k customers | High |
| Stock pay does not fall | Owners keep less of every revenue dollar, permanently | USD 15.52, -19% | Stock pay over revenue | High |
| Growth slows faster than guided | The multiple resets a second time | USD 3 to 6 a share | The 36% guided for Q3 | High |
| AI compute eats the gross margin | A point of gross margin lost is a point of cash margin | USD 0.92 a point | Gross margin, 84% and falling | Medium high |
| Cost of capital rises | 81% of the value sits in the terminal period and reprices hard | USD 15.28 on 150bp | The US 10 year yield | Medium high |
| Share supply, control and churn | Lock-up expiry plus 10% a year issuance; a 15.7% stake holds 73.6% of votes | A discount, not a level | Form 4 filings, short interest, the CTO search | Medium |
Fair value impacts are this desk's calculation from the sensitivity grid. Underlying facts are sourced in the sections above.
The bear case, argued properly
One: the design file stops being where software starts. AI coding agents attack Figma's position directly, and a manager who describes a screen and receives working code never opens a design tool. Figma's answer, an agent on its own canvas, concedes the ground has moved. Two: the stock pay never comes down, because Figma bids for engineers against model labs paying in their own appreciating equity, and the cheapest way to keep them is more shares. Three: the margin damage is structural rather than cyclical. Gross margin fell from 89% to 84% in twelve months and free cash flow margin from 27% in March to 14% in June, and guidance of 9% against 12% delivered supports that reading.
The bull case, argued properly
One: Figma becomes where AI software gets built, not just designed. Code Layers, Figma Make and the agent try to own the whole path from idea to shipped product, and if that works the seat count stops being the ceiling because credits scale with usage rather than headcount. Over 80% of customers above USD 10,000 used AI credits weekly by June. That path gives USD 42.96. Two: growth keeps beating the company's own guidance, and three consecutive quarters of acceleration at this scale is rare. Three: USD 1.60bn of net cash and no debt mean Figma never has to sell equity into a weak price.
This note is wrong on the upside if stock pay falls below 20% of revenue on a trailing four quarter basis, which alone lifts the anchor above today's price, or if retention returns above 140% having slipped from 139% in March to 136% in June. It is confirmed on the downside if retention falls below 120% or customer growth drops under 20%.
Portfolio fit, entry and the decision
Volatility sets the size, not liquidity
Liquidity is not the constraint here, unusual for a name this volatile: Figma trades about USD 535m a day, so any realistic position can be built or exited inside a session. The binding constraints are volatility and the bear case. The shares fell 74% in twelve months, moved 15% in one session after the June results, and 27% of the float is sold short. That puts the instrument in a 0 to 2% band of a growth allocation. This is a high beta US software holding correlated with the Nasdaq and most closely with the AI-disruption theme now moving monday.com and Atlassian on the same days, so anyone holding a broad technology fund already owns some of this exposure.
The bear case takes the shares from USD 23.29 to about USD 6.53, a fall of 72%. Anyone uncomfortable with that should not own this at any price. The preferred purchase price of USD 12.50 is the anchor less the 35% margin of safety, which is 46% below today and below the 52 week low, so this is a name to track rather than buy. A second route to BUY does not need the price to fall at all: if stock pay drops below 20% of revenue, the range itself moves up. Buy in tranches either way.
Monitoring plan
- Each quarter: stock pay as a share of revenue, gross margin, and free cash flow after share settlement tax.
- Each quarter: growth against the guided 36% for Q3 and implied 28% for Q4, retention, the customer counts and the share count.
- As they occur: the CTO appointment, further departures, and competing design tooling from a model vendor.
| Test | Pass | Evidence |
|---|---|---|
| The business is understandable | Yes | Seats plus AI credits, one product surface |
| A moat is identified with a named mechanism | Yes | Narrow: design system switching costs |
| The balance sheet survives a downturn | Yes | USD 1.60bn net cash, no borrowings |
| Management is fully accountable to owners | No | 15.7% of the equity holds 73.6% of the votes |
| Earnings quality is clean | No | Cash flow negative after share settlement tax |
| Price below fair value less the margin of safety | No | Trades 86% above the USD 12.50 entry |
| Risks are tolerable at the current price | No | The bear case implies a 72% fall |
| The position is sizeable and exitable | Yes | USD 535m traded daily |
Four passes and four fails, concentrated on price and governance rather than on the business.
The product won its category and the balance sheet removes the risk of ruin. What is not resolved is who gets paid: two fifths of every revenue dollar currently goes to staff in new shares, and the price already assumes that stops. Buy it at USD 12.50, or buy it when stock pay falls below 20% of revenue. Not before.
Appendix
| Rating | Definition |
|---|---|
| Buy | Price at or below the fair value midpoint less the margin of safety, with business, balance sheet, governance and valuation all passing. |
| Watch | Quality passes, price does not, or a catalyst must be confirmed first. A WATCH names its trigger: a price, a threshold, or an event that would move it to BUY. |
| Avoid | Fails on business quality, balance sheet, governance or valuation, with no realistic path to the required return. |
The equivalent house convention is BUY, HOLD, SELL. WATCH replaces HOLD because these notes inform a decision to buy rather than advise an existing position.
| Source | Tier | As of |
|---|---|---|
| Figma Q2 2026 results release and Form 10-Q, plus the Q2 2025 to Q1 2026 releases for the quarterly series, guidance and the lock-up schedule | 1 | Sep 2025 to Aug 2026 |
| Figma Form S-1, S-1/A and 424B4, for 2024 results, the tender offer charge, the Adobe merger and the share classes | 1 | Jul 2025 |
| Figma Form 10-Q cover pages for share counts by class, Schedule 13G filings for ownership and the irrevocable proxy, and Form 4 filings on officer sales | 2 | Oct 2025 to Jul 2026 |
| TheFly analyst actions for the seven houses named; TechCrunch on the Bud acquisition and the Anthropic board resignation; reporting on the Q2 2026 call | 3 | Apr to Aug 2026 |
| Aswath Damodaran, Data Update 2 for 2026; the Morgan Stanley market estimate; an AlphaSense expert survey on category share; Bloomberg on the Canva share sale and the Claude Code run rate | 3 | Aug 2025 to Mar 2026 |
| S&P Global Market Intelligence via stockanalysis.com, for prices, multiples, float, short interest, betas and peer statistics | 4 | 7 to 9 Aug 2026 |
| Trading Economics for the 10 year Treasury yield; Benzinga and Public.com for conflicting analyst counts; Simply Wall St for insider selling; Contrary Research for the 2022 market size | 4 | 2025 to Aug 2026 |
All figures are in US dollars, Figma's reporting currency, so no exchange rate was applied. Prices are the New York close of 7 August 2026 and balance sheet figures are as at 30 June 2026 with no post-balance-sheet adjustment, so every section sits on the same basis.
Limits of desk research
Four things sit outside what public filings answer, and that is where the remaining edge is. The vesting schedule of pre-listing equity awards, which decides how fast stock pay falls and is the most important input here. The unit cost of AI inference per credit and whether it is falling. Customer-level evidence on whether AI credits expand accounts or substitute for seats, which needs channel checks rather than a filing. And how quickly the AI coding agents now in the market actually displace design work in ordinary product teams, as opposed to in demonstrations. Beyond that this is desk research, untested against management.
Basis of analysis: a discounted cash flow in which stock based pay is charged as a cost rather than added back, at a 10.0% cost of capital built on a bottom-up peer beta of 1.27, with revenue growth fading from 28% to 6% by 2035, a terminal economic margin of 23% and 542.6m diluted shares held flat against USD 1,599m of net cash at 30 June 2026. All valuation arithmetic, scenario probabilities and the sensitivity grid are this desk's work. Figures marked E are estimates.
The views expressed in this report accurately reflect the analyst's personal views about the subject instrument. No part of the analyst's compensation was, is, or will be directly or indirectly related to the specific recommendation or views expressed. The analyst holds no position in Figma, Inc.
This document is independent research prepared for informational purposes. It is not investment advice, not a recommendation to buy or sell any security, and does not constitute an offer or solicitation. Figures are drawn from public sources believed reliable as of the date shown and may change without notice. Anyone acting on this material does so at their own risk and should seek their own professional advice.