FactSet Research Systems Inc.(FDS) · Professional Information Services

FactSet Research Systems Deep Value Investment Research

Other languages
Quick ReadPlain-language overview · read this first

FactSet sells financial data and investment research workflow software to institutional clients, covering the full investment chain from research, portfolio analytics, trading, risk, to reporting. Its client base is mainly buy-side, and annual subscriptions drive high retention. At its core, this is a data-plus-tooling business embedded in investment workflows. Rating: Cautious Buy. This is a good business, but it looks more like a quality company available at a discount than a deep-value name cheap enough to make winning almost automatic.

The tension lies in the mismatch between valuation and business quality. At the current price of 245 dollars and about 15.8 times PE, it is far cheaper than peers such as MSCI, Moody's, and S&P Global. ASV retention has long stayed above 95%, free cash flow has consistently come close to or even exceeded net income, growth consumes almost no capital, and returns on capital are not built by piling on leverage. Still, the market has practical reasons for assigning a discount: it is not the strongest player in the industry. Its brand falls short of Bloomberg's, its moat is less institutionalized than those of the index and ratings leaders, growth is usually only in the mid-single digits, and it is now caught in a dual transition around AI and a new CEO.

The most important long-term risk is that AI flattens the scarcity value of basic data and standard analytics tools, narrowing the workflow moat. The new CEO's compensation package is expensive, and insider ownership is not high, both of which also deserve monitoring. The analyst puts the ideal buying range at 200—230 dollars, believes a small staged position can be initiated at the current level, but would keep waiting if the goal is an extremely thick margin of safety.

Lead

FactSet is a high-retention, asset-light financial data and workflow software company with strong cash generation. At about $245, or roughly 15.8x PE, it trades at a clear discount to peers such as MSCI, Moody's, and S&P Global, although AI substitution risk and execution under the new CEO keep the margin of safety from being thick. Research rating Cautious Buy: a quality compounder at a reasonable discount, not a deeply mispriced asset.

Full report

Prices in the article are as of publication; see the valuation band above for the live price.

Conclusion First

Investment rating: Cautious Buy. From the perspective of a long-term business owner, FactSet is a readable, cash-generative, sticky-customer, asset-light data and workflow software business with a moderately strong moat. For many years, the company has sustained high retention, steady revenue growth, high operating margins, and stable free cash flow, showing that its products are embedded in clients' critical investment research, portfolio analytics, trading, and risk-management workflows. The issue is that this is not a worry-free great company: AI is increasing the substitutability of basic data and analytics tools, FactSet still competes with larger or more specialized rivals such as Bloomberg, LSEG, S&P Global Market Intelligence, MSCI, and Morningstar, and the incentive package after the new CEO took office in 2025 is generous while management ownership is not especially high. After weighing business quality, valuation, and risk, my conclusion is: this is a good company, but at the current price it looks more like a quality company at a certain discount than an extremely cheap deep-value opportunity.

As of May 30, 2026 (Asia time zone; corresponding to the U.S. trading day of May 29), FDS traded at about $245.47, with a market capitalization of roughly $9.098 billion and a platform-shown trailing PE of about 15.8x. Based on the latest 10-Q disclosure of $268 million in cash, $17 million in investments, and $1.369 billion in long-term debt, this report estimates enterprise value at about $10.2 billion. On a rolling free-cash-flow basis, the current equity free cash flow yield is already meaningfully higher than the historical level usually seen for high-quality information-services companies, and valuation appeal has improved materially from the highs of 2025.

Does the current price offer a margin of safety? The answer is yes, but not a thick one. If you are willing to underwrite a long-term path of mid-single-digit to mid-high-single-digit organic growth, operating margins around 30%, and strong buybacks, the current price has some appeal. If your standard is that even a growth slowdown must be protected by the balance sheet or an extremely low valuation, FDS is not cheap enough, because its valuation is supported by continuing earning power, not asset liquidation value.

By investor type, it is better suited to long-term value investors / quality-growth investors / investors who prefer high-cash-flow software and data-infrastructure assets. It is less suitable for deep-value investors who require a very thick margin of safety, and it is also not ideal for short-term traders focused only on near-term catalysts.

The three biggest uncertainties are: first, the speed at which AI affects financial-information intermediaries and workflow software; second, execution and capital-allocation quality under the new CEO; and third, whether FactSet can sustain high retention and mid-to-high margins while facing stronger rivals from multiple sides.

Business Understanding and Industry Structure

II. Business understanding: how does this company make money?

Fact. FactSet's revenue mainly comes from subscription-based financial data, analytics tools, and enterprise workflow solutions, covering the investment lifecycle from research, portfolio construction and analytics, trade execution, performance measurement, risk management, and reporting. The company also provides data feeds, cloud-based digital solutions, APIs, and identification and data-infrastructure businesses such as CUSIP Global Services. The company explicitly discloses that revenue is mainly derived from client access to its multi-asset platform during the contract term, and most contracts are treated in accounting as a single performance obligation to continuously provide platform access, with revenue recognized over time.

Fact. FactSet's client base is mainly buy-side institutions. As of August 31, 2025, the company had 8,996 clients and 237,324 users; buy-side clients accounted for about 82% of Organic ASV, while sell-side clients accounted for about 18%. By February 28, 2026, the client count had risen to 9,101, users to 241,352, and the ASV-based annual retention rate remained above 95%, while the client-count retention rate was about 91%. This shows the company does not make money from one-off projects, but from highly recurring and sticky annual subscription revenue.

Fact. In pricing, the company's core metric is ASV, the forward-looking value of subscription revenue contracted and being delivered to clients over the next 12 months. ASV at the end of fiscal 2025 was $2.4056 billion, and ASV at the end of fiscal Q2 2026 was $2.4502 billion. The company treats ASV as a key indicator of recurring revenue and cash-flow capacity.

Fact. The cost structure mainly consists of employee compensation, third-party data costs, computer and cloud-related costs, and amortization of intangible assets. In fiscal 2025, cost of services was 47.3% of revenue, SG&A was 20.5%, and the operating margin was 32.2%. This is a typical high-gross-margin, high-fixed-cost, low-marginal-cost business that depends on continuous investment in talent and infrastructure.

Fact. Client concentration is not high: the company discloses that no single client accounted for more than 3.5% of total revenue in fiscal 2023 to 2025. The supply side is not free of dependence: the company says two third-party data suppliers each accounted for more than 10% of total data costs in 2025. In addition, among multiple cloud-service providers, one supplier provided most cloud-computing support in fiscal 2025. This means buyer demand is diversified, while upstream data and cloud infrastructure still carry some dependency.

Inference. Overall, this business is simple, transparent, and understandable. Clients are not paying for one isolated database. They are paying for an integrated set of "data + tools + workflows + services" deeply embedded in investment research and investment workflows. If the stock market were closed for five years, as long as global capital markets, asset management, and investment-banking activity still existed, I would still be willing to own this business. The caveat is that it is not a consumer staple. It is tied to financial-industry budgets and technology-substitution risk.

Business understandability score: 4.5/5.

III. Industry and competitive structure

Fact. FactSet names its largest competitors as Bloomberg, S&P Global Market Intelligence, and LSEG Data & Analytics. Other competitors include BlackRock Aladdin, MSCI, Morningstar, and others. The company also stresses that its environment is highly competitive, with both old and new rivals accelerating innovation through alternative data, analytics tools, fintech platforms, and AI models.

Fact. Public disclosures from comparable companies show that long-term demand in this industry remains solid. MSCI disclosed in Q4 2025 that recurring subscription Run Rate organic growth was 7.7%, with a retention rate of 93.4%. LSEG's preliminary 2025 results disclosed 5.0% growth in Data & Analytics and 5.9% ASV growth in subscription businesses. S&P Global's 2025 annual report showed total revenue growth of 8%, with all five divisions growing. Morningstar's full-year 2025 revenue grew 7.5%, with organic growth of 8.0%. Together, these data points show that financial data, indices, research, risk, and workflow software remain a mature but steadily growing industry in asset management, wealth management, investment banking, research, and compliance, not a declining industry.

Inference. This is more like a mature growth industry than a highly cyclical industry. Demand does not come from end-consumer sentiment, but from the data, models, identifiers, research, and workflow infrastructure that financial institutions need to operate. Still, it is not non-cyclical. FactSet explicitly notes that asset-management clients' profitability is tied to AUM, and that financial-market declines, M&A consolidation, budget pressure, and industry layoffs can slow new subscriptions and user expansion.

Inference. Long-term industry demand is stable, but technology-disruption risk is higher than in the past. The "raw data layer" of basic market data and public information is most vulnerable to erosion by AI, alternative data, and cheaper interfaces. The more stable part is the analytics, portfolio, trading, reporting, and enterprise-integration layer deeply embedded in client workflows. If FactSet can keep moving from the "terminal screen" toward "embedded workflow + API + enterprise solutions + AI context integration," the industry remains attractive. If it stays at the level of basic search and data display, the moat will narrow.

View. My characterization of the industry is: a good company in a good industry, but not the strongest company in that industry. Bloomberg has a stronger terminal advantage, LSEG and S&P Global have greater scale, data assets, and brands, and MSCI is stronger in the index and asset-management ecosystem. Yet FactSet still has considerable depth in buy-side research, portfolio analytics, and multi-asset workflows, and its current valuation is clearly lower than several of the "strongest" players.

Industry attractiveness score: 4/5.

Moat and Management

IV. Moat analysis

The table below reflects judgment based on company disclosures, industry structure, and the competitive landscape. It is not the company's own description. The evidence base mainly comes from client retention, the revenue model, product embedding, the competitive set, and risk disclosures.

Moat factor Judgment Basis
Brand advantage Moderately strong Reliable brand among professional buy-side/sell-side users, but below Bloomberg's outright dominance
Cost advantage Medium Scale and platform reuse create operating leverage, but third-party data and talent costs are not low
Scale advantage Moderately strong 9,000+ clients, more than $2.4 billion in ASV, global platform and COEs create economies of scale
Network effects Medium to slightly weak Not a typical two-sided platform; more of a data/workflow ecosystem effect than a strong network effect
Switching costs Strong High retention, workflow embedding, API and enterprise integration, multi-module use across research/portfolio/trading
Channel advantage Medium Strong direct sales and deep service capability for institutional clients, but not an exclusive channel
Patent/license/regulatory barriers Medium CGS, compliance, data contracts, and industry know-how create barriers, but are not impossible to replicate
Data advantage Strong Long-built connected financial databases, aggregation of third-party content, identifiers, and workflow integration
Culture and operating capability Moderately strong Strong client service, talent development, and long record of revenue growth
Capital-allocation capability Medium Long-term dividend and buyback discipline is good, but acquisition and buyback prices have not always been excellent

Fact. FactSet's moat is not based on "the lowest price," but on deeply connected data, analytics, and workflow embedding. The company discloses that its solutions cover the full investment lifecycle and are delivered through desktop, data feeds, cloud-based digital solutions, and APIs. Most contracts are serial services providing continuous platform access, and ASV retention was above 95% in both 2025 and the first half of 2026. This structure typically maps to high switching costs.

Fact. The company also has some pricing power. The 2024 10-K explicitly states that part of Organic ASV growth came from higher demand and price increases, especially for workstations and CGS subscriptions. In 2025 to 2026, the company continued to maintain high retention and roughly 5% to 7% ASV growth, showing that among core clients, FactSet still has some modest pricing power, rather than being a pure price taker.

Inference. My view of the moat is: overall stable, with the underlying data layer slightly narrowing while the workflow layer remains relatively solid. AI will lower the threshold for "querying, summarizing, and preliminary analysis," but the closer the task gets to real order flow, portfolio risk, compliance, reporting, enterprise data governance, investment-research collaboration, and deep API embedding, the higher the replacement cost becomes. Replicating FactSet requires more than spending money to buy data. It requires years of sustained investment in product, workflow, implementation, service, and client trust. For a new entrant, replication would take years and large amounts of capital. For existing giants such as Bloomberg / LSEG / S&P, it is continuous competition from an advantaged base, not a start from zero.

Inference. In an inflationary environment, FactSet looks more like software/data infrastructure that can raise prices in small steps than like a monopoly resource that can raise prices at will. In a downturn, it would probably remain profitable because revenue is recurring, clients are diversified, cash flow is strong, and leverage is not high. But it will not ignore the cycle like a utility. A prolonged slump in global financial markets, project delays, and client consolidation would still slow ASV and user growth. I believe the company's historically high margins mainly reflect structural advantages, not a one-cycle tailwind.

Moat strength score: 4/5.

V. Management and capital allocation

Fact. The governance framework is broadly sound. The 2025 Proxy shows that the company has an annual say-on-pay mechanism; shareholder support for the 2024 executive-compensation plan was 94.6%; directors and executives have stock-ownership requirements, with the CEO required to hold shares equal to 6x annual salary and the CFO 3x; the company prohibits executives and directors from hedging and pledging shares, has a clawback policy, no tax gross-ups, uses double-trigger change-in-control vesting, and receives advice from independent compensation consultant Meridian.

Fact. Management's economic alignment with shareholders is not especially strong. As of October 1, 2025, all directors and executives together held only about 439,600 shares, or 1.2% of total shares outstanding. New CEO Sanoke Viswanathan disclosed beneficial ownership of 0 at the time. Although management is "in compliance or on a path to compliance" under the rules, this is still far from founder-like long-term ownership.

Fact. The CEO transition in 2025 is a major governance item to monitor. Viswanathan formally became CEO on September 8, 2025. His contract includes a $1 million annual salary and a 200% target bonus, plus a sizable "make-whole / buyout" package: a $3 million one-time cash payment, $10 million one-time cash payment subject to clawback conditions, $26 million in RSUs, $10 million in three-year performance RSUs, and $22 million of performance options. Even if part of this compensated him for rights forfeited when leaving JPMorgan, the package still shows that the board paid a high price to recruit an external CEO.

Judgment. This leads me to assess management as follows: honesty and the governance framework are broadly acceptable, but shareholder alignment and capital-allocation taste are above average rather than excellent. The positives are a relatively complete governance system, candid risk discussion, and ongoing buybacks and dividends. The negatives are an expensive new-CEO package, low insider ownership, and the fact that in fiscal Q4 2025 the company repurchased shares at an average price of $410.51, which was not cheap in hindsight. On the other hand, in the first two quarters of fiscal 2026, the company continued repurchases at average prices of $292.61 and $250.14, respectively, showing that management actually increased buyback activity after the share price fell. That deserves credit.

Fact. In cash use, the company invests, acquires, pays dividends, and repurchases shares. In fiscal 2025, the company returned $460.4 million to shareholders. In the first half of fiscal 2026, it repurchased $302.9 million of stock, buying back 1.1299 million shares in six months and materially reducing shares outstanding. In fiscal 2025, the company also strengthened OMS, IR, AI, and private-market capabilities through acquisitions and product expansion. But because the transactions are recent and intangible assets and amortization have risen clearly, it is too early to conclude that "the acquisitions have created value."

Management and capital-allocation score: 3.5/5.

Financial Quality and Owner Earnings

VI. Financial quality analysis

The table below is based on the 2025 10-K, the Q2 2026 10-Q, and 2025/2026 quarterly press releases. TTM figures are rolling estimates based on disclosed data and are derived values.

Metric FY2024 FY2025 TTM to 2026-02-28 Note
Revenue ($mn) 2,203.1 2,321.7 2,401.1 First-half 2026 year-on-year growth about 7%
Revenue YoY - 5.4% About 3.4% versus FY2025 TTM basis affected by rolling period
Operating profit ($mn) 701.3 748.3 748.5 First-half 2026 margin declined slightly
Operating margin 31.8% 32.2% 31.2% Affected by compensation, technology, and amortization
Net income ($mn) 537.1 597.0 587.8 TTM basically flat
Net margin 24.4% 25.7% 24.5% Still high quality
CFO operating cash flow ($mn) About 700.4 726.3 798.9 2024 backed out from FCF + Capex
FCF free cash flow ($mn) 614.7 617.5 682.9 Strong cash generation
FCF / net income 114% 103% 116% Cash earnings better than accounting earnings
Capital expenditure ($mn) 85.7 108.8 116.0 Mainly related to internal software development
Diluted EPS ($) 13.91 15.55 About 15.54 TTM did not keep growing
Latest shares outstanding (mn) - 37.65 (period end) 36.64 (period end) Significant buybacks in first half of 2026
Latest liabilities/assets - - 49.6% Based on the 2026-02-28 balance sheet
Latest net debt/adjusted EBITDA - - About 1.1x This report's estimate; leverage is not high

Fact. Revenue grew 5.4% in fiscal 2025, and revenue in fiscal Q1 and Q2 2026 grew 6.9% and 7.1%, respectively, showing acceleration at the start of the new fiscal year. But margins did not expand in parallel. GAAP operating margin in fiscal Q2 2026 fell to 30.3%, adjusted operating margin fell to 35.0%, and management explicitly attributed the pressure to compensation, technology-related expenses, and intangible-asset amortization. In other words, demand is improving, but cost investment is also rising.

Fact. Cash-flow quality is very good. Fiscal 2025 net income was $597 million, and free cash flow was $618 million. In the first half of fiscal 2026, net income was $286 million and free cash flow was $276 million, while rolling TTM free cash flow is estimated in this report at about $683 million. This shows that most of FactSet's profits really convert into cash, rather than being built up through large receivables or one-off accounting treatments.

Fact. Growth does not require very heavy capital investment. Fiscal 2025 capital expenditure was $108.8 million, about 4.7% of revenue. Even including the rolling TTM for the first half of 2026, capital-expenditure intensity is only around 4.8%. This is a typical asset-light, high-cash-conversion model. One point to note is that the company explicitly says the increase in 2025 capital expenditure mainly came from capitalized internal software development, while part of the increase in amortization of intangible assets within cost of services in 2025 and 2026 also came from these capitalized software assets. My judgment is: this is not an obvious accounting problem, but it requires investors to keep tracking cash flow and the capitalized-software policy.

Fact. The balance sheet is broadly solid. As of February 28, 2026, the company had $268 million in cash, $17 million in investments, and $1.369 billion in long-term debt; total liabilities were $2.091 billion, and shareholders' equity was $2.129 billion. This report estimates net debt at about $1.084 billion, and net debt/adjusted EBITDA at about 1.1x, which is not stressful.

Inference. Capital returns are excellent. On a TTM basis, estimated ROE is about 27%, and ROA about 14%. If using after-tax EBIT and net invested capital, ROIC is roughly around 19%. These numbers are not created by extreme leverage. They are built on high retention, an asset-light model, and buybacks.

Fact. Working capital consumed cash in the first half of 2026 in the short term. In the 10-Q, the company explains that first-half fiscal 2026 operating cash flow of $333 million included $285.6 million of net income and $154.2 million of non-cash expenses, but also $106.8 million of working-capital cash outflow, mainly because of the timing of client collections and annual variable compensation payments. In the same balance sheet, accounts receivable increased to $320 million from the end of 2025, and deferred revenue rose to $189 million. This looks more like a seasonal/timing issue than demand deterioration.

Judgment. I do not see clear signs of financial fraud or aggressive manipulation. Profit and cash flow match, client concentration is low, and deferred revenue is growing. The real items to watch are threefold: internal software capitalization, amortization of intangible assets from acquisitions, and continued margin erosion from AI/cloud investment. In addition, the company has almost no inventory, so for it, inventory turnover is less meaningful than receivables, deferred revenue, capitalized software, and share-count changes.

VII. Owner Earnings analysis

Definition and basis. For an asset-light data/software company such as FactSet, I prefer to define Owner Earnings as: net income + non-cash expenses - maintenance capital expenditure - working-capital investment needed to maintain the business.

Fact. The latest verifiable TTM net income is about $587.8 million. Non-cash expenses in the first half of fiscal 2026 at least included: depreciation and amortization of $87.84 million, amortization of lease right-of-use assets of $16.19 million, stock-based compensation of $39.32 million, deferred taxes of $2.38 million, and other non-cash items of $8.39 million. First-half capital expenditure plus capitalized software spending totaled $56.84 million, while working capital saw cash outflow.

Inference. Because first-half working-capital volatility has clear seasonality, I do not directly annualize "net income + all non-cash items - half-year capex" in a straight line. Instead, I start more conservatively from rolling FCF. Based on disclosed data, TTM FCF is about $683 million. To be conservative, I apply a small discount and assume there is still some maintenance working-capital need and ongoing software investment. Under that approach, conservative Owner Earnings of around $650 million is reasonable. At the current share price, that corresponds to a conservative Owner Earnings multiple of about 14x, and Owner Earnings per share of about $17 to $18.

Judgment. These numbers communicate a clear message: The company's long-term free cash flow is above or close to net income, so it is not the kind of business with attractive accounting earnings but poor cash flow. Growth also does not continuously consume large amounts of capital. From the Owner Earnings perspective, FDS's real earning power is better than the first impression given by net income alone.

Intrinsic Value and Margin of Safety

VIII. Intrinsic value estimate

Method 1: Owner Earnings DCF

This section does not use a "perfect scenario." It uses conservative Owner Earnings of $650 million as the base. This is not company guidance. It is a conservative inference made in this report based on TTM FCF, working-capital seasonality, and maintenance investment.

Scenario OE growth over next ten years Discount rate Terminal growth rate Intrinsic value per share
Conservative 3% 10% 2.5% About $250
Base 5% 9% 3% About $350
Optimistic 7% 8.5% 3.5% About $480

Interpretation. In this framework, the current market price of $245 is roughly close to the conservative scenario and well below the base scenario, but it is not low enough to be "a sure win even if the future goes wrong." Put differently, the current price looks more reasonably low than significantly undervalued.

Method 2: Relative valuation

The table below mixes current market data with public secondary valuation metrics. A reminder is necessary: different sources may calculate EV/EBITDA, EV/Sales, and P/B on either TTM or forward bases, so they are not fully comparable. They should therefore serve only as supporting evidence, not as a substitute for intrinsic-value analysis.

Company Current/approximate PE EV/EBITDA EV/Sales Comment
FactSet 15.8x About 10.4x About 4.25x Current price is clearly below most high-quality data peers
Morningstar 18.6x About 10.7x About 3.31x Valuation is not high, but earnings quality and institutional workflow depth are slightly weaker than FDS
S&P Global 26.8x About 23.9x About 8.3x Stronger scale, brand, index, and ratings barriers
Moody's 32.5x About 25.2x About 10.2x Very strong moat, also very expensive valuation
MSCI 36.1x About 44.5x About 14.9x Extremely strong position in indices/asset-management ecosystem, with a significant premium

Judgment. If FactSet is placed in the basket of "high-quality financial information and index/analytics infrastructure," its relative valuation is cheap at the current price. The issue is that the market has real reasons not to give it the same premium as MSCI, Moody's, or S&P Global: FactSet's brand and platform control are weaker than Bloomberg's, its moat is less "institutionalized" than index/rating leaders, growth is usually only mid-single-digit, and it is currently in a dual transition of AI and a new CEO.

Method 3: Asset or liquidation value

Fact. As of February 28, 2026, the company had total assets of $4.220 billion, including $1.287 billion of goodwill and $1.885 billion of intangible assets, for a combined total of about $3.172 billion; shareholders' equity was $2.129 billion. In other words, if goodwill and intangible assets are stripped out, tangible net assets are actually negative. Therefore, FDS's value cannot be protected by book assets. It must be supported by its ability to keep generating cash flow in the future.

Judgment. This means two things: First, it is not an asset-discount stock, so investors should not take comfort from "liquidation value." Second, the margin of safety must come from business quality + a reasonable purchase price, not asset revaluation.

Overall valuation conclusion:

  • Conservative intrinsic-value range: $230 to $280

  • Reasonable intrinsic-value range: $320 to $380

  • Optimistic intrinsic-value range: $430 to $500

  • Current price versus intrinsic value: close to the lower end to midpoint of the conservative range; discounted versus the base range

  • Required margin of safety: at least 20% to 25%

  • Ideal buy-price range: $200 to $230

  • Acceptable holding-price range: $230 to $320

  • Clearly overvalued range: above $380 to $400

IX. Margin of safety

Judgment. The current price is not especially cheap, but it has started to offer a margin of safety. The most fragile valuation assumption is not "whether this year's EPS misses by 5%." It is: can FactSet maintain mid-single-digit growth + high retention + operating margins around 30% over the next 5 to 10 years? If the answer is yes, today's price is not expensive. If AI rapidly turns this business from "workflow embedded" into "replaceable information intermediary," the valuation will be revised downward.

Inference. If future growth is below expectations, as long as the company can still sustain about 3% growth, margins around 30%, and ongoing buybacks, long-term returns will probably not be too bad. But if margins fall into the high-20% range, retention declines, and the market compresses the multiple to around 12x PE, then using the current $15.55 EPS as a rough reference, the share price could return to about $186, implying nearly 24% downside. In other words, valuation-multiple compression alone is enough to cause a medium-sized permanent loss of capital unless operating growth keeps up.

View. So there is indeed a shadow of "a good company, but not a great price." My conclusion is not "you must buy immediately." It is: if you can accept building a position in tranches, holding for the long term, and treating $200 to $230 as the more attractive zone, then a small initial allocation can begin now; if you act only when the margin of safety is very thick, you can keep waiting.

Risks, Comparison, and Final Conclusion

X. Risks and the bear case

The most important risk is not share-price volatility, but permanent capital loss. I summarize the core risks as follows:

Risk Why it matters Current observation
Competition risk Bloomberg, LSEG, S&P, MSCI, and Morningstar are all increasing data and AI investment The industry is crowded, and large companies have deeper capital resources
Technology-substitution risk AI may compress the value of basic information and standard analytics tools The company has explicitly listed this as a major risk in the 10-K
Regulatory/data-rights risk Data licensing, privacy, AI compliance, and financial-regulation changes can affect supply and costs Dependence on upstream content and licenses cannot be ignored
Financial-leverage risk Debt is not low, but remains within a manageable range Net debt/EBITDA is about 1.1x, not high
Management risk The new CEO has not been in office long, and culture and execution still need proof Transition-period compensation package is expensive
Valuation risk Even if the company is good, lower market valuation can hurt returns Current price is not "extremely cheap"
Cyclical risk Buy-side client budgets are tied to AUM and investment-banking conditions Stable long term, still affected by markets in the short term
Client/supplier-structure risk Clients are diversified, but data suppliers and cloud suppliers create dependencies Two data suppliers each account for 10%+ of data costs; one cloud vendor provides most cloud support
Accounting risk Capitalized software and acquisition amortization affect profit metrics Requires ongoing tracking
Business-model disruption risk If workflow embedding is replaced by AI middle layers, the moat will narrow This is the key long-term risk

The above judgments are mainly based on the company's risk disclosures regarding competition, AI, cloud, market volatility, talent, and leadership changes, as well as the latest balance-sheet and operating-cash-flow facts.

The strongest bear case: "FactSet looks like a high-quality subscription business, but it is actually in a dangerous middle zone: it does not have Bloomberg's screen monopoly and terminal necessity, it does not have MSCI's institutional barriers from the index/fund ecosystem, and it does not have Moody's ratings-regulation moat. AI is eroding the scarcity of information retrieval and standardized analytics tools, and clients may concentrate budgets with larger platforms or cheaper solutions. FactSet may ultimately lose both its growth premium and the stickiness of traditional data terminals." I think this is a bear case that deserves serious attention.

What facts would make me admit I was wrong: If several of the following occur over the next 4 to 6 quarters, I would clearly turn more conservative or even bearish: first, ASV retention falls below 94% and keeps deteriorating; second, Organic ASV growth keeps falling below 3%; third, operating margin falls below 30% with no visible recovery path; fourth, FactSet's AI/workflow product progress fails to convert into sales and clients visibly shift to larger platforms; fifth, management conducts expensive acquisitions or again repurchases large amounts of stock at clearly overvalued prices.

The biggest permanent-capital-loss scenario: It is not a recession or one quarterly miss. It is the reverse of the idea that "the workflow moat is underestimated and the speed of technology substitution is overestimated": clients gradually migrate analytics and data access to cheaper, larger, AI-native platforms; FactSet is reduced to a back-end data supplier; margins and buyback capacity both step down; and the market re-rates it from a "high-quality compounder" into an "ordinary low-growth information-services company." In that case, a 30% to 50% long-term capital loss is not impossible, especially considering it lacks strong liquidation value on the balance sheet.

XI. Comparison with other opportunities

Compared with the strongest competitors. If the comparison is purely "business quality," FactSet may not beat Bloomberg, MSCI, Moody's, or S&P Global. But if the comparison is "quality / valuation," FactSet is currently more interesting. Its current PE is about 15.8x, clearly lower than MSCI's 36.1x, Moody's 32.5x, S&P Global's 26.8x, and Morningstar's 18.6x. In other words, the market has already applied a price discount for "not being the strongest player in the industry."

Compared with broad indices. FactSet disclosed in its five-year shareholder-return chart in the 2025 10-K that if $100 had been invested in August 2020, by August 2025 FactSet became $107, while the S&P 500 became $185. This shows that over the past five years, holding FDS was worse than holding the index. This reminds us not to romanticize a "good company," and it also means current relative neglect may provide a better entry point. The reason to buy FDS should not be "it has risen a lot in the past." It should be "at a reasonable or even somewhat low price today, it still has recurring cash flow and buyback capacity."

Compared with risk-free yields or high-grade bonds. This report did not separately pull the current 10-year U.S. Treasury yield or high-grade corporate-bond yield, so this is only a directional judgment: if your main goal is certain cash income, bonds are more suitable. If your goal over more than 10 years is a combined return from cash-flow growth + share-count reduction through buybacks + valuation repair, then FDS becomes meaningful. In other words, it is not a "cash substitute." It is a candidate high-quality compounder with equity upside. Before buying, you still need to recheck the latest bond yields.

View. On whether it deserves capital, I would summarize it this way:

  • Clearly better than the index? Not necessarily.

  • Clearly better risk-reward than most peers? Possibly.

  • Can it enter a portfolio with only five assets? Borderline. For conservative long-term investors, I would not put it among the top 1 or 2 opportunities across the whole market. But if you already prefer high-cash-flow, low-capital-intensity, buyback-driven financial-infrastructure companies, FDS is qualified to enter a watchlist or even a small-position build list.

XII. Investment Checklist

The table below reflects judgments based on the full analysis. "Pass / Fail / Uncertain" are the author's judgments.

Checklist item Conclusion
Can I understand this business? Pass
Does it have stable long-term demand? Pass
Does it have a durable moat? Pass
Does it have pricing power? Pass, but not strong
Can it generate stable free cash flow? Pass
Are its returns on capital excellent? Pass
Is management trustworthy? Pass, but the new CEO needs observation
Is capital allocation rational? Pass, but not outstanding
Is the balance sheet solid? Pass
Is valuation below intrinsic value? Pass, but by a limited margin
Is the margin of safety sufficient? Uncertain, leaning pass
Would I feel comfortable holding it long term? Pass, but must accept AI risk
What key facts would make me sell? Sell if retention/growth/margins/capital allocation deteriorate
Am I buying only because the price rose or market sentiment improved? This impulse fails; the current logic should be based on cash flow and valuation

XIII. Final investment conclusion

【Final Rating】 Cautious Buy

【One-sentence investment thesis】 FactSet is a high-retention, asset-light financial data and workflow software company with strong cash flow. Its current valuation is already clearly below most high-quality peers, but the moat is not impregnable, and the margin of safety exists but is not thick.

【Core bullish reasons】 The company's revenue is mainly subscription-based, ASV retention has long stayed above 95%, and both client count and user count are still growing. Cash-flow quality is high; in recent years, free cash flow has generally been close to or above net income, and growth does not consume large amounts of capital. The current valuation is clearly discounted versus comparable high-quality assets such as MSCI, Moody's, and S&P Global. The balance sheet remains solid, net debt/EBITDA is not high, and the company continues to repurchase shares and reduce share count. The company has gradually extended from a single terminal toward APIs, enterprise solutions, AI, and multi-asset workflows; if execution is good, the moat can still hold.

【Core bearish reasons】 AI may compress the scarcity value of basic financial information and standard analytics tools, narrowing the moat. The strongest competitors are larger and stronger, and pressure from industry price competition or product substitution is real. The new CEO has not been in office long, the compensation package is expensive, and management's ownership alignment with shareholders is not high. The company is not an asset-discount investment; goodwill and intangible assets are a high share of assets, and liquidation value is weak. The current price is not expensive, but it is not cheap enough to ignore execution risk.

【Key assumptions】 FactSet can sustain long-term organic growth of about 4% to 6%. ASV retention continues to stay around 95%. The operating-margin center stays around 30% to 32%, instead of structurally falling below 30%. AI is not a one-way hit to the business, but is partly converted by FactSet into product strength and sales opportunities. Buybacks continue at relatively rational prices, instead of heavy repurchases again in an overvalued zone.

【Reasonable buy price】 $200 to $230 is the buy range I would be more satisfied with; $230 to $250 can justify starting a small position; above $320, I would be more inclined to hold than to add. This is based on a combination of conservative Owner Earnings DCF, relative-valuation discount, and the company's lack of asset backing.

【Target holding period】 More than 10 years. This type of company is better suited to realizing returns through revenue growth, profit compounding, share-count shrinkage, and valuation normalization over a long cycle, rather than betting on one or two quarters of upside surprise.

【Expected annualized return】 Conservative scenario: 6% to 8% Base scenario: 9% to 12% Optimistic scenario: 13% to 15% These are not share-price forecasts. They are long-term return assumptions based on Owner Earnings growth, buybacks, and valuation-range evolution.

【Maximum downside risk】 If AI and competition accelerate erosion of the moat, margins step down, and growth stalls, the market could compress FDS's valuation back to about 12x PE or lower. Based roughly on current earning power, the share price could return to the $150 to $190 range, down about 20% to 40% from the current price. In an extreme structural failure scenario, an even larger decline cannot be ruled out.

【Tracking indicators】 The most important indicator is Organic ASV growth. Second is ASV retention. Third is GAAP/adjusted operating margin. Fourth is the match between free cash flow and net income. Fifth is share-count changes and average buyback price. Sixth is changes in capitalized internal software and amortization. Seventh is net additions to clients and users. Eighth is signals of new bookings/cross-selling from AI product commercialization. Ninth is net debt/EBITDA. Tenth is post-acquisition integration results and returns on intangible assets.

【Signals that would trigger reassessment】 Organic ASV stays below 3% for multiple consecutive quarters. ASV retention falls below 94%. Operating margin falls below 30%, while management continues to invest heavily without a clear return path. Large expensive acquisitions or renewed large-scale buybacks in an overvalued zone. Clients visibly shift to lower-cost AI-native platforms, and FactSet's presence in key workflows declines.

【Final recommendation】 Put calmly, FDS is not the kind of stock that should be heavily weighted after one glance. But it is not a mediocre company either. It is more like a quality company with good long-term economics whose valuation has returned to a zone that is researchable and actionable. If you prefer holding companies with high cash flow, sticky customers, and low capital consumption for the long term, the current price can justify entering gradually, with a small position and a long-term mindset. If you insist on a very thick margin of safety, or if you are more pessimistic about the speed at which AI will reshape the financial-information value chain, then waiting for a lower price better fits a conservative approach.

Open questions and limitations This report prioritized the latest 10-K, 10-Q, investor-relations materials, and official press releases. Due to public collection limitations, a complete standardized 5- to 10-year full financial series was not fully expanded year by year, and some long-term trends use the last three years of official data and rolling TTM estimates. This report did not separately extract the current 10-year U.S. Treasury yield or high-grade corporate-bond yield, so the comparison with the "risk-free yield" is mainly directional. If you are preparing to actually build a position, I suggest rechecking two items before placing an order: the latest bond yields and ASV/margin trends around the next earnings report.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.

MSCISPGIMCOMORNLSEGBLK

FactSetfinancial datainvestment research workflowsubscription modelmoatvalue investingmargin of safety
Reader Q&A10

Baillie Framework · Ten Questions for Growth Investing

10

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

Baillie Framework · Ten Questions for Growth Investing — score profile: 45/100 total Ceiling 6/10 · Revenue 2x 2/10 · Next engine 4/10 · Moat 6/10 · Reinvention 5/10 · Management 4/10 · Customer need 6/10 · Unit economics 6/10 · 5x path 3/10 · Blind spot 3/10 0510 How large is its market ceiling? Is it expanding an existing pie, or creating an entirely new market? — 6/10 Ceiling 6 Can its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses? — 2/10 Revenue 2x 2 Five years from now, what will take over as the next growth engine? Does this "second curve" exist today? — 4/10 Next engine 4 What is its core competitive advantage? Will this moat widen or narrow over the next three to five years? — 6/10 Moat 6 If the core business is disrupted, does it have the DNA to reinvent itself? How does it handle mistakes and bad news? — 5/10 Reinvention 5 Does management, especially the founder, have a long-term view and deep alignment with the company? Is it willing to sacrifice current profit for results five to ten years out? — 4/10 Management 4 If it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable and not dependent on harming society or exploiting regulation? — 6/10 Customer need 6 What are the unit economics of this business, in gross margin and incremental returns? Do they improve or deteriorate as scale increases? Where does the money it earns go? — 6/10 Unit economics 6 What conditions must all be true for it to rise fivefold in ten years? Are those conditions realistic? What expectations are embedded in today's share price? — 3/10 5x path 3 Why has the market not recognized all this yet? Is it because investors do not understand it, dismiss it, or cannot look far enough ahead? What could become the "narrative inflection point"? — 3/10 Blind spot 3
  • How large is its market ceiling? Is it expanding an existing pie, or creating an entirely new market?6/10

    Conclusion: FactSet has a meaningful market ceiling, but it is mainly expanding a mature existing pie, not creating a new 0-to-1 market. That pie is the global budget for financial data, investment research workflows, portfolio analytics, trading, risk, reporting, wealth management, and enterprise data infrastructure. Demand is durable, customers have strong ability to pay, and stickiness is high, but the industry already has strong competitors such as Bloomberg, LSEG, S&P Global, MSCI, Morningstar, and BlackRock Aladdin. This is therefore not a story of explosive penetration into a new market, but a story of raising wallet share and extending into adjacent workflows within a large mature market.

    FactSet is already a sizeable company by its current scale: FY2025 revenue was $2.3217 billion, with 8,996 clients and 237,324 users, while buy-side clients accounted for about 82% of Organic ASV; by Q2 FY2026, revenue was $611 million, up 7.1% year over year, Organic ASV reached $2.4491 billion, up 6.7% year over year, and the company also raised FY2026 revenue guidance to $2.450 billion to $2.470 billion. This shows the core market is still growing, but the growth slope looks more like mid-single digits to mid-to-high single digits than the high double-digit expansion of early-stage SaaS.

    The market ceiling can be viewed in three layers. The first is core buy-side and sell-side workflows, namely research, portfolio, trading, risk, and reporting. This is the mature main battlefield where FactSet can still grow through retention, pricing, cross-selling, and net client additions. The second is enterprise data, APIs, data feeds, cloud solutions, and embedded workflows, which can push FactSet from a "terminal/desktop tool" into customers' internal systems and expand its share of existing budgets. The third is wealth management, private markets, corporate clients, and AI workflows. For example, the company disclosed that Q2 growth was driven by institutional buy-side and wealth management clients, and it launched products such as AI Doc Ingest and financial crime risk workflows. These are adjacent market expansions, but they still serve existing data, compliance, and investment process budgets at financial institutions.

    So, if we make a rough calculation based only on the current ASV of about $2.45 billion and future organic growth of 5%-7%, ASV could reach about $4.0 billion-$4.8 billion ten years from now. If execution goes well in AI, APIs, wealth, private markets, and enterprise data, with acquisitions and stronger cross-selling added on top, the revenue ceiling could be higher, and moving toward more than $5.0 billion is not unimaginable. But this requires FactSet to keep taking budget from competitors and customers' internal build-outs; the market will not naturally grow tenfold for it.

    For the Baillie Gifford framework, the key judgment is: the ceiling is large enough, but not "unlimited." FactSet's market can support long-term compounding because financial institutions cannot operate without trusted data and workflows. But its upside comes more from going deeper, becoming more automated, and embedding further inside existing financial information infrastructure than from creating an entirely new consumer-grade or platform-grade market. FDS is therefore more like a high-quality share gainer in a mature large market, not a new-market definer.

    Jun 6, 2026
  • Can its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses?2/10

    Conclusion: Based on the current evidence, the probability that FactSet's revenue will "at least double" over the next five years is not high. Starting from FY2025 revenue of $2.3217 billion, a five-year doubling would require revenue to reach about $4.64 billion, implying roughly a 14.9% CAGR. The company's latest trajectory is clearly one tier lower: FY2025 revenue grew 5.4%, organic revenue grew only 4.4%, and the company's disclosed FY2026 guidance is only $2.450-$2.470 billion of revenue and $130-$160 million of Organic ASV growth. This looks more like a mature 5%-7% compounding model than a five-year doubling model.

    In terms of growth sources, the main driver is not a single explosive new business, but renewals of the existing subscription base + modest pricing + client/user expansion + cross-selling. Q2 FY2026 revenue was $611 million, up +7.1% year over year; Organic ASV was $2.4491 billion, up +6.7% year over year; and the company also disclosed ASV retention above 95%, 9,101 clients, and 241,352 users. This shows the base business remains healthy, but it is steady expansion of a high-retention subscription business, not a sudden acceleration in demand.

    Broken down into volume, price, and new businesses: price contributes, as the company has some pricing power in subscriptions such as core workstations, data, and CGS; volume also contributes, especially from institutional buy-side clients, wealth, dealmakers users, and net client additions; new businesses are more like enhancers, including wealth, private markets, buy-side front office, APIs, data feeds, AI workflows, and small acquisition integrations. The annual report makes clear that the strategy focuses on expanding data delivery, embedding into client workflows, and advancing AI, while Q2 also noted that growth was driven by institutional buy-side and wealth management clients.

    My judgment is therefore that a five-year revenue doubling would require FactSet to lift its current growth rate from the mid-single digits to nearly 15%. Unless new curves such as AI/wealth/private markets/APIs scale materially, or the company completes larger acquisitions, this should not be the base case. A more reasonable base case is that revenue continues growing in the mid-single digits to mid-to-high single digits, while profit and per-share value are driven mainly by high retention, light capital intensity, buybacks, and valuation repair.

    Jun 6, 2026
  • Five years from now, what will take over as the next growth engine? Does this "second curve" exist today?4/10

    Conclusion: The most likely successor five years from now is not the number of traditional Workstation seats itself, but a set of new use cases around AI-embedded workflows + enterprise data/APIs + private markets/wealth/front-office. This second curve already exists today, but only as a product layer and cross-selling lead embedded in the core business, not as a separately disclosed new business segment large enough to carry growth on its own.

    The core business remains healthy, but it is already fairly mature. FactSet's 2025 annual report shows that buy-side clients accounted for about 82% of Organic ASV, with 8,996 clients, 237,324 users, and ASV retention above 95%; by Q2 FY2026, the company had $611 million of revenue, up +7.1% year over year, Organic ASV growth of +6.7%, and raised FY2026 guidance. This shows the base business can still compound, but the room for reacceleration from traditional desktop terminals alone is limited.

    The direction of the second curve has shifted from "showing users data" to "embedding FactSet's data, models, and compliance capabilities into customers' real workflows." In 2026, the company appointed a Chief AI Officer and a new CTO, then disclosed that the Document Search beta covered more than 85,000 users. Q2 operational highlights also mentioned integrating AI-driven financial crime risk management into Workstation; at the same time, AI Doc Ingest for Cobalt has already entered the workflow for extracting portfolio-company documents and structuring data for private capital clients.

    I would rank the priorities in three layers: the first is AI workflows, turning investment research, due diligence, compliance screening, and document search into traceable and auditable productivity tools; the second is enterprise data/APIs, extending FactSet from a front-end interface into the back-end data layer inside client systems; the third is private markets, wealth, and front-office, including Cobalt, growth in wealth management users, and the OMS/IBOR and front-office trading workflow capabilities brought by LiquidityBook.

    But this second curve has not yet been proven. FactSet has not broken out AI, Cobalt, APIs, or front-office workflows as separate revenue segments; Q2's "early AI contributions" are more signals of client engagement and operating efficiency than a clear scale of new revenue. The most reasonable judgment today is therefore: the second curve exists in the product and sales funnel, but not yet in the financial reporting structure. If FDS exceeds expectations five years from now, it will most likely come from these new workflows raising ARPU, expanding the client base, and deepening switching costs. If future AI products cannot translate into ASV growth, they will look more like defensive feature upgrades than a true new growth engine.

    Jun 6, 2026
  • What is its core competitive advantage? Will this moat widen or narrow over the next three to five years?6/10

    Conclusion: FactSet's core competitive advantage is not Bloomberg-style absolute monopoly from an irreplaceable terminal, nor an institutional moat like MSCI or rating agencies. It is a more practical composite moat: workflow embedding, connected financial data and identifiers, enterprise API/data-feed integration, long-term service capability, and high retention. Over the next three to five years, AI will compress the basic search and standard analytics layers, but enterprise workflows, compliance data, integration of customers' proprietary data, private-market data, and the API layer have room to widen. Overall, the moat is moderately strong, but with narrowing risk.

    The hardest evidence is stickiness. In fiscal 2025, FactSet had 8,996 clients, 237,324 professional users, an ASV retention rate above 95%, and client-count retention of 91%, while buy-side clients accounted for about 82% of Organic ASV. This shows that what it sells is not a one-off database, but something embedded in the research, portfolio, trading, risk, performance attribution, and reporting processes of asset managers, wealth firms, investment banks, PE/VC firms, corporate IR teams, and other customers. No single customer accounted for more than 3.5% of total revenue, which also reduces the bargaining or churn risk from any one large customer.

    The core of the moat is that "replacing it is a hassle." The company's products cover workstations, portfolio analytics, enterprise data solutions, data feeds, cloud solutions, and APIs, while CGS also supports security master files for the front, middle, and back offices of the investment industry. The company itself emphasizes that its applications, client support, and services are embedded in many financial professionals' workflows and handle sensitive information such as clients' proprietary portfolio data. This kind of embedding creates high switching costs: replacing FactSet is not just changing a screen, but reconnecting data definitions, permissions, models, reports, customers' internal systems, and compliance processes.

    But it is not an invincible moat. FactSet discloses major competitors including Bloomberg, S&P Market Intelligence, and LSEG Data & Analytics, and also faces competing products such as BlackRock Aladdin, MSCI, and Morningstar. Its advantage is therefore more like "strong switching costs + strong data integration + a moderately strong brand + service depth," not a harder institutional barrier such as terminal monopoly, regulatory licenses, or an index ecosystem.

    The areas likely to widen are mainly the enterprise layer and workflow layer. FactSet is extending delivery from traditional workstations to data feeds, APIs, cloud-based digital solutions, and AI workflows; Q2 FY2026 Organic ASV grew 6.7% year over year, showing that core subscriptions are still expanding even after AI disruption has been widely discussed by the market. If customers need auditable, permissioned financial data that can connect to internal systems, rather than open-web answers from general-purpose large models, FactSet's position as enterprise-grade data infrastructure may strengthen.

    The risk of narrowing is also real. AI will reduce the scarcity of basic information retrieval, summaries, initial modeling, and standard charts. The company also acknowledges that its competitiveness could be harmed if competitors or third parties integrate generative/agentic AI into products faster and more successfully, or if AI technology reduces customer demand for its products. In addition, in fiscal 2025 two third-party data suppliers each accounted for more than 10% of data costs, and one cloud provider supplied most cloud support. This shows that part of FactSet's advantage comes from licensing, integration, and service, not wholly owned exclusive data assets.

    Over the next three to five years, the moat is likely to see "the upper layer widen and the lower layer narrow," with a net effect that is relatively stable. If the new CEO can move the company into institutional AI workflows and enterprise data infrastructure, the moat will widen modestly. If customers shift basic research and data queries to cheaper AI/cloud platforms while FactSet only defends the traditional workstation, the moat will narrow. Q2 FY2026 GAAP operating margin of 30.3% and adjusted operating margin of 35.0% still show high earnings quality, but the year-over-year margin decline also reminds us that this moat requires continuous investment to maintain and should not be valued as an absolute monopoly.

    Jun 6, 2026
  • If the core business is disrupted, does it have the DNA to reinvent itself? How does it handle mistakes and bad news?5/10

    Conclusion: FactSet has a moderate capacity for self-reinvention, but it is not a platform-revolution company. It has extended from financial data terminals/workstations into workflow layers such as data feeds, cloud solutions, APIs, enterprise data, CGS, research, portfolio, trading, performance, risk, and reporting (2025 10-K). This shows that it is not merely defending a single screen that sells data, but can embed its data assets into customer processes.

    The reinvention path looks more like an "incremental mosaic." FY2025 acquisitions of Irwin, LiquidityBook, and LogoIntern strengthened IR/corporate workflow, OMS/IBOR, and investment-banking automation respectively; the company also divested RMS Partners, moving a non-core sell-side research authoring platform out of the business and shifting resources toward Dealmakers and integrated workflows (Irwin, LiquidityBook, LogoIntern, FY2025 Q4). This proves it understands that when the core data layer is squeezed, it needs to migrate into adjacent high-stickiness processes. It also shows that the main tool is product extension and small to midsize acquisitions, not a one-time reconstruction of the industry platform.

    AI is the core stress test. In its 10-K, the company directly acknowledges that if competitors or third parties use generative/agentic AI in products faster and more successfully, this could weaken its competitiveness and even reduce customer demand for its products and services (2025 10-K). The appointment of a Chief AI Officer and CTO in March 2026 shows that it has elevated AI/platform strategy to the organizational level (SEC 8-K exhibit). But commercialization still needs to be proven, so it should not be treated as an already successful second curve.

    It is generally fairly candid about bad news. The Q2 FY2026 results directly disclosed that GAAP operating margin fell from 32.5% to 30.3%, adjusted operating margin fell from 37.3% to 35.0%, with causes including compensation, technology expenses, and intangible asset amortization; the EPS decline was also attributed to higher expenses and asset impairment (Q2 FY2026). The 2025 10-K also disclosed a material weakness in IT general controls related to revenue, accounts receivable, and deferred revenue. Although the company said no material misstatement had been identified, the weakness had not been fully remediated as of FY2025 (2025 10-K). These disclosures show that the company is not only packaging a growth story.

    But "being able to disclose bad news" does not mean capital allocation is already excellent. New CEO Sanoke Viswanathan took office on 2025-09-08 and came from JPMorgan; the 2025 Proxy shows that his compensation package includes a $1 million annual salary, a 200% target bonus, $22 million of performance options, and sizable make-whole awards (2025 Proxy). In addition, the company repurchased stock in FY2025 Q4 at an average price of $410.51, so capital allocation discipline still needs to be observed (FY2025 Q4).

    Therefore, if traditional data queries and workstations are partially disrupted by AI, FactSet is unlikely to stand still. It will probably continue migrating toward APIs, enterprise data, front office, wealth, private markets, IR, OMS, and automation tools. The Q5 assessment is upper-mid: it has evolutionary capacity and is willing to disclose uncomfortable information, but the speed of reinvention, AI commercialization, and quality of capital allocation have not yet been fully proven.

    Jun 6, 2026
  • Does management, especially the founder, have a long-term view and deep alignment with the company? Is it willing to sacrifice current profit for results five to ten years out?4/10

    Conclusion: FactSet has standardized governance and some long-term orientation, but it is not the typical Baillie Gifford profile of "deep founder alignment and management highly in the same boat as shareholders." The company was founded in 1978 and is no longer founder-led today; current CEO Sanoke Viswanathan is an externally hired professional manager who became CEO on September 8, 2025, after previously serving as International Consumer & Wealth CEO at JPMorgan and sitting on the JPMorgan Operating Committee. This brings experience in financial clients, wealth management, and international business, but it also means management quality here has to be validated through systems, compensation design, and execution results, rather than being directly underwritten by founder-owner mentality.

    Alignment is the main shortcoming. The 2025 Proxy shows that the company has a fairly complete governance framework, including a CEO share ownership requirement of 6 times annual salary, 3 times for the CFO, and mechanisms such as bans on hedging/pledging and clawbacks. But as of October 1, 2025, all directors and executives together held only 439,643 shares, or about 1.2%, while Sanoke Viswanathan's beneficial ownership was 0 at that time. The new CEO's compensation package is also expensive: $1 million base salary, a 200% target cash bonus, an $11 million annual equity target from FY28 onward, and $22 million of performance options. These arrangements can link part of future returns to the share price, but they look more like "bringing in a professional manager at a high price" than "management already betting heavily on the company with large amounts of its own capital over the long term."

    Is it willing to sacrifice current profit for five to ten years from now? The answer is: there is evidence, but the degree is limited. FactSet continues to invest in AI, platform, and technology. In Q2 FY2026, the company disclosed the launch of AI Doc Ingest, the appointment of a Chief AI Officer and CTO, and stronger enterprise AI and platform strategy; in the same quarter, GAAP operating margin fell from 32.5% to 30.3%, and adjusted operating margin fell from 37.3% to 35.0%, with the main pressure coming from compensation, technology-related expenses, and intangible asset amortization. This shows that management is indeed accepting short-term margin pressure in exchange for upgrades to product and AI capabilities.

    But this is not an aggressive reinvestment company. FactSet still maintains the capital return discipline of a mature company. In Q2 FY2026, it repurchased 651,750 shares at an average price of $250.14, spending $163 million, and still had $697.1 million of repurchase authorization remaining at quarter-end. This is shareholder-friendly and also shows that the board has capital allocation discipline; but under the Baillie Gifford framework, it looks more like "a high-quality mature compounder making transformation investments" than a founder-led company willing to sacrifice current profit substantially to pursue nonlinear growth ten years out.

    So my judgment for Question 6 is: management is qualified and somewhat positive, with long-termism supported by systems and actions, but shareholder alignment is not deep enough, the founder factor is absent, and the new CEO still needs to prove himself through execution results over the next 2-3 years. For FDS, Q6 is not a fatal flaw, but it is not a strong positive contributor to a ten-year five-bagger story either.

    Jun 6, 2026
  • If it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable and not dependent on harming society or exploiting regulation?6/10

    Conclusion: Customers would miss FactSet a lot, but it would not be completely irreplaceable; its growth model is broadly sustainable and does not depend on social harm or regulatory arbitrage. If FactSet disappeared tomorrow, the biggest pain would not be the loss of one data terminal, but the disruption of customers' daily workflows in investment research, portfolio analytics, trading, risk, reporting, data feeds, APIs, enterprise data governance, and CGS/CUSIP. In its 2025 annual report, the company disclosed that in fiscal 2025 it had 8,996 clients and 237,324 users, ASV retention of >95%, client retention of 91%, buy-side clients accounting for about 82% of Organic ASV, and no single client accounting for more than 3.5% of total revenue in fiscal 2023-2025. By Q2 FY2026, clients had further increased to 9,101 and users to 241,352, while ASV retention remained >95%. These numbers show that it is not a dispensable software seat, but infrastructure embedded in the daily processes of many institutions.

    But "customers would miss it a lot" does not mean "it cannot be replaced." Bloomberg, LSEG Data & Analytics, S&P Global Market Intelligence, MSCI, Morningstar, BlackRock Aladdin, AlphaSense, and customers' internally built data platforms can all replace parts of FactSet's functionality. The real migration cost comes from portfolio and risk models, research templates, Excel/Office plugins, APIs, data mapping, permissions, historical data, reporting processes, compliance audits, and employee habits, not from any single database itself. Its stickiness is therefore very real, but it is more like a high-switching-cost workflow platform than a uniquely irreplaceable monopoly facility.

    In terms of growth quality, FactSet's model is relatively healthy: it relies mainly on subscriptions, cross-selling, data depth, workflow embedding, AI-assisted efficiency, and expansion of enterprise solutions, rather than inducement-driven consumption, regulatory avoidance, or customer-dependency traps. It serves institutional asset managers, wealth firms, asset owners, hedge funds, investment banks, corporations, and data/technology partners. Its core value is improving the efficiency with which financial professionals process data, analyze securities, manage portfolios, monitor risk, and generate reports. Identifier businesses such as CGS/CUSIP also have a financial-market infrastructure character, closer to standardization, compliance, and operational efficiency improvement.

    The main constraints are also clear: FactSet must continue to obtain third-party data and exchange content legally, comply with privacy and cross-border data rules, handle accuracy, bias, intellectual-property, and compliance risks in AI-generated content, and rely on upstream data suppliers and cloud service providers. The annual report also disclosed that two data suppliers each accounted for 10%+ of fiscal 2025 data costs, and that one cloud supplier provided most cloud computing support. These risks do not negate the sustainability of the business model, but they do limit its pricing power and expansion speed.

    So the answer to Baillie Gifford Q7 is: customers would clearly miss FactSet, especially heavy buy-side users and customers in enterprise data and middle/back-office processes; but customers are not without alternatives. Replacement would be painful, slow, and expensive, but not impossible. The growth model is broadly legal and sustainable, with social externalities ranging from neutral to positive. The core risk is not "making money by harming society," but whether AI, data rights, financial regulation, upstream supply, and client budget pressure will weaken its retention and pricing power.

    Jun 6, 2026
  • What are the unit economics of this business, in gross margin and incremental returns? Do they improve or deteriorate as scale increases? Where does the money it earns go?6/10

    Conclusion: FactSet has very good unit economics. It is a light-capital financial data/workflow software business with high cash conversion and high incremental returns; in theory, scale should make it better, but at the moment AI, cloud, compensation, data, and intangible amortization are consuming part of the operating leverage.

    In FY2025, the company generated $2.3217B of revenue, $748.3M of operating income, a 32.2% operating margin, $597.0M of net income, and $15.55 EPS. Cost of services was 47.3% of revenue, implying a rough service gross margin of about 52.7%; SG&A was 20.5%. This means that each dollar of revenue carries meaningful costs for data, engineering, client support, cloud, and amortization, but the operating profit left over is still thick. The report estimates TTM FCF at about $682.9M, FCF/net income at about 116%, and capex intensity at about 4.8%, which is also consistent with the company's FY2025 disclosures of $617.5M of free cash flow and $108.8M of capital expenditures: the profit quality is not only accounting profit, but real cash conversion.

    Incremental returns are also good. StockAnalysis' current view shows P/FCF of about 13.6, EV/EBITDA of about 11.6, and ROIC of 17.65%. For a company with high client retention, subscription-led revenue, and capital expenditures below 5% of revenue, this shows that new growth does not require constant additions of factories, inventory, or heavy assets. The real investments are in products, data, engineers, AI, cloud infrastructure, and client service. Marginal cost is higher than pure SaaS, but far lower than in traditional capital-intensive industries.

    The answer on scale effects is: better over the long term, but currently held down by reinvestment. FactSet's platform, data connections, APIs, workflows, and client service can be reused across more customers, so fixed R&D, infrastructure, and sales/service costs should theoretically be spread over a larger revenue base as scale expands. But pressure has already appeared in Q2 FY2026: the company had $611.0M of revenue, up 7.1% year over year, but GAAP operating margin fell to 30.3% and adjusted operating margin fell to 35.0%. Management pointed to compensation, technology-related expenses, professional fees, and intangible asset amortization. This does not mean the unit economics have turned bad; it means the company is trading part of its margin for AI, cloud migration, products, and acquisition integration capabilities.

    The money it earns mainly goes to four places: first, continued investment in products, AI, technology, and internal software development; second, acquisitions, including FY2025 transactions such as Irwin and LiquidityBook, with the annual report disclosing acquisition-related cash spending of about $348.3M; third, buybacks, with Q2 FY2026 again repurchasing 651,750 shares for $163.0M at an average price of $250.14, leaving $697.1M of authorization; fourth, dividends, with FY2025 dividends of $160.0M and 26 consecutive fiscal years of dividend increases. Overall, FDS is a good business where cash flow comes in first and is then allocated among products/AI, acquisitions, buybacks, and dividends. The key question is not whether it can make money, but whether those reinvestments can offset the erosion of margin from AI and rising costs.

    Jun 6, 2026
  • What conditions must all be true for it to rise fivefold in ten years? Are those conditions realistic? What expectations are embedded in today's share price?3/10

    Conclusion: A fivefold return in ten years is not completely absurd, but its realism is low to medium-low. Using the 2026-06-05 share price of about $255.62, PE-TTM of about 16.45, forward PE of about 13.85, and P/FCF of about 13.64, a fivefold return would imply a share price of about $1,278 and a market capitalization of about $47B. If the market assigns only 20x PE ten years from now, FactSet would need about $2.37B of net income; at 25x PE, it would still need about $1.89B; and at only 15x PE, it would need about $3.16B. Compared with the company's FY2025 net income of about $597M, diluted EPS of $15.55, and free cash flow of $617.5M, net income would need to grow to roughly 3.2-5.3 times its current level, with a net income CAGR of about 12%-18%, plus effective buybacks or sustained valuation.

    Several conditions would need to hold at the same time. First, Organic ASV and revenue cannot stay at mid-single-digit growth for long; they would need to lift from the current steady pace toward high single digits or even low double digits. But the company's latest view is only FY2026 revenue guidance of $2.45-$2.47B and Organic ASV additions of $130-$160M, or about 5.4%-6.7%. Second, AI must not commoditize basic financial data and standard analytics tools; it must instead deepen FactSet's embedding in workflows, APIs, enterprise data, portfolio analytics, and compliance scenarios. Third, margins must regain operating leverage rather than being persistently consumed by AI, cloud, compensation, and amortization. This has not yet been proven, because the company has already disclosed in Q2 that GAAP operating margin fell to 30.3% and adjusted operating margin fell to 35.0%. Fourth, buybacks must continue to shrink the share count effectively, preferably when the stock is undervalued rather than consuming capital at high prices. Fifth, ten years from now the market must still view it as a high-quality financial data compounder, not a low-growth information services provider, and give it a terminal valuation of at least around 20x.

    Individually, none of these conditions is an absurd assumption; FactSet has high retention, light capital intensity, strong cash flow, and real customer embedding. But having them all hold simultaneously for ten years is not easy. The facts today look more like a company moving from "high-quality slow compounder" toward "can it reaccelerate," rather than one already entering explosive growth. The core judgment in the report is similar: FDS is a good company and the current valuation is discounted, but it is not a super-growth stock where one can assume a ten-year fivefold outcome with eyes closed.

    Today's share price does not embed high expectations. Trailing PE around 16 times, forward PE around 13-14 times, and an FCF yield of about 7%, combined with a share-price decline of about 41% over the past 52 weeks, suggest the market is pricing in: FactSet can still earn money, still repurchase shares, and still has high retention, but growth is likely only 5%-7% ASV / revenue growth + high-single-digit EPS compounding, while AI and competition suppress the valuation premium. In other words, the current price embeds an expectation of "decent quality but questioned growth," not a blue-sky expectation of "fivefold in ten years." If the future proves that AI has not eroded the moat, margins recover, and EPS can compound at double digits, there is room for re-rating, but a fivefold outcome would still require a stronger set of results than the current operating trajectory.

    Jun 6, 2026
  • Why has the market not recognized all this yet? Is it because investors do not understand it, dismiss it, or cannot look far enough ahead? What could become the "narrative inflection point"?3/10

    Conclusion: The market has not missed FactSet's quality; it is simply unwilling for now to pay a "great growth stock" valuation for it. Today it looks more like a mature compounding asset categorized as "high quality but potentially diluted by AI, with margins under pressure during an investment phase and not the strongest player in the industry," rather than an overlooked high-growth dark horse.

    This is not a failure to understand; it is more dismissal + limited willingness to look far ahead. The dismissal is about relative position: FactSet has high retention and strong cash flow, but it does not have Bloomberg's terminal mindshare, nor the more institutionalized barriers of MSCI, Moody's, or S&P Global. The limited long-range view concerns the value of workflow embedding: the market can easily treat it as a "financial data provider" while underestimating the switching costs inside research, portfolio, risk, enterprise data, APIs, and client process migration. Even so, the discount is not without reason: as of 2026-06-05, FDS was about $255.62, with TTM PE of about 16.45x and forward PE of about 13.85x, but the stock had still fallen about 41.07% over the past 52 weeks, and short interest was about 15.81% of float, showing that the market is still betting that growth and the moat will be compressed.

    The real disagreement is whether AI weakens FactSet or strengthens it. The short thesis says AI will commoditize search, summaries, and basic analysis, reducing customer dependence on traditional data terminals and standard analytics tools. The bull thesis says AI needs trusted data, permissioning systems, enterprise workflows, and audit trails, so FactSet can instead embed AI into existing client processes. Q2 FY2026 has already provided some positive evidence: revenue was up +7.1% year over year, Organic ASV was up +6.7%, ASV retention was >95%, and the company raised FY2026 Organic ASV and revenue guidance. But the negative evidence is also clear: GAAP operating margin fell from 32.5% to 30.3%, adjusted operating margin fell from 37.3% to 35.0%, and margin pressure mainly came from compensation, technology expenses, and intangible asset amortization.

    A positive narrative inflection point would require several signals to appear together: Organic ASV accelerating further from the mid-single digits, with ASV retention staying above 95%; AI, enterprise data, and workflow products translating into visible orders such as multiyear client agreements like Barclays, AI Doc Ingest, and risk-control workflows, rather than staying at the product-launch level; margins stopping their decline and the market believing operating leverage can recover after the 2026 investment phase; and the company continuing to repurchase shares at low valuations. In Q2 it repurchased 651,750 shares at an average price of about $250.14, with remaining authorization of about $697.1M, which reinforces the compounding logic of "cash flow + share-count reduction".

    Conversely, the negative inflection points are also clear: ASV retention falling below 94%, Organic ASV repeatedly dropping below 3%, GAAP operating margin falling below 30% with no visible recovery path, or AI products failing to generate sales conversion while customers visibly shift toward larger platforms or lower-cost AI-native solutions. In other words, the market misunderstanding around FDS is not "nobody knows it is good." The market has not yet believed it can be retold during the AI cycle from a traditional financial data company into a platform company built around "trusted financial data + enterprise workflows + AI distribution layer."

    Jun 6, 2026
Ask about this report

Members can ask about this report; once answered it appears under "Reader Q&A" on this page. You can also highlight a passage in the text to ask about it directly.