Terry Smith

Buying a high-return company is only the start. The return must be cash, growth must earn an attractive incremental return, the price must leave something for the buyer, and inactivity must remain conditional.

Three Rules Form One System

Terry Smith founded Fundsmith in 2010 and manages the Fundsmith Equity Fund. The current fund factsheet names him as portfolio manager and identifies the vehicle as a UK open-ended investment company. Fundsmith LLP is regulated by the Financial Conduct Authority, while a depositary, registrar, auditor, authorised corporate director, and investment manager perform separate legal and operational functions.

Smith's 2025 annual letter repeats the three-part process: buy good companies, do not overpay, and do nothing. The order matters. Inactivity creates value only after the company and price pass demanding tests. It cannot repair weak cash economics or an excessive purchase price.

The Fundsmith Owner's Manual defines a good business as one that can sustain a high return on operating capital in cash. It then adds growth: the company must be able to redeploy at least some cash at similarly attractive returns. Earnings-per-share growth without the capital required to generate it does not answer either question.

Quality Has to Become Cash

Fundsmith favors repeat purchases, installed bases, client relationships, patents, brands, and distribution that competitors cannot readily reproduce. It also prefers businesses that do not require leverage to create an acceptable equity return. The theory is that durable intangible assets can resist mean reversion without attracting the same debt-funded capacity as a replicable factory or property.

The test remains physical. Customers have to continue buying; prices and volume have to cover wages, suppliers, tax, working capital, and maintenance assets. Cash collected has to reconcile with profit. Acquisitions, stock compensation, receivables, inventory, restructuring, and capital spending can produce attractive earnings while reducing the amount available per share.

Fundsmith's 2025 look-through table reported a weighted 31% return on capital employed, 62% gross margin, 28% operating margin, 94% cash conversion, and median interest cover of 29 times. The S&P 500 and FTSE 100 comparisons excluded financial businesses. These are weighted Fundsmith and Bloomberg calculations for the holdings, not consolidated or audited line items of a single company.

The letter also says weighted portfolio free cash flow grew 16% in 2025. That is supportive, but cash conversion remained below its historical level around 100%. A quality diagnosis needs both the high return and the movement of profit into spendable cash.

AI Spending Tests the Asset-Light Assumption

Smith attributes much of the remaining cash-conversion gap to increased capital expenditure at Alphabet, Microsoft, and Meta. Those companies were building data centres and buying chips for artificial intelligence. The software, user relationships, and data may be intangible, but the next competitive stage requires physical power, servers, buildings, cooling, and network capacity.

The letter calls the eventual return on that spending an open question. AI can change products and work without delivering an adequate return to every infrastructure owner. A communicated opportunity is not a cash return. Analysts need utilization, pricing, incremental revenue, depreciation, energy cost, replacement cycles, and competitive supply before classifying the expenditure as productive growth.

This also complicates Fundsmith's valuation adjustment. The Owner's Manual estimates free cash flow after tax and interest and adds back discretionary capital expenditure that is not needed to maintain the business. If spending is necessary to preserve product quality or competitive position, calling it discretionary overstates distributable cash. If it creates a high-return new capacity, treating all of it as maintenance understates value. The classification needs operating evidence.

Do Not Overpay

Fundsmith compares free cash flow per share with market value, long-term interest rates, and alternative investments. In 2025, the portfolio's weighted free-cash-flow yield rose from 3.1% to 3.7%; the letter reports year-end yields of 2.8% for the S&P 500 and 3.1% for MSCI World. A higher-quality portfolio can rationally trade at a lower yield, but the expected growth and duration must justify the difference.

A yield is not a return promise. Free cash flow may fall, rates may rise, currency can move, a patent can expire, or the terminal multiple can contract. Smith's approach therefore needs scenarios for the operating runway and a price that does not assume every favorable outcome. A familiar brand at a high multiple is not automatically safer than an unfamiliar business.

Do Nothing Does Not Mean Never Sell

Low turnover reduces brokerage commission, bid-offer spread, transaction tax, and the risk of replacing a sound holding with a worse idea. The 2025 letter reports portfolio turnover of 12.7% and voluntary dealing cost equal to 0.008% of average fund value. The T accumulation share class had a 1.04% ongoing charge and a 1.06% total cost of investment after dealing costs connected with fund flows.

The same year, the fund sold two companies, bought four, and received a spun-off holding. Three of 29 year-end companies had been held since inception, ten for more than ten years, and sixteen for more than five. Inactivity is a low-intervention policy, not an absence of monitoring or transactions. Daily subscriptions and redemptions can also force the open-ended vehicle to trade even when the manager's security view has not changed.

Novo Nordisk created the sharper revision. Smith wrote that Fundsmith had favored businesses that could perform without exceptional management, but recent experience showed the set was smaller than the firm thought. After describing lost operating position and illegal generic competition, he said selling may be more effective than engagement when poor management is encountered. The franchise, management, and feasible remedy now sit explicitly inside the hold decision.

The Fund Record Has Defined Boundaries

The 2025 annual letter reports that the T accumulation shares returned 0.8% that year versus 12.8% for the MSCI World Index in sterling with dividends reinvested. From November 1, 2010 through December 31, 2025, the fund returned 612.9% cumulatively and 13.8% annualized, versus 467.6% and 12.1% for the index.

The fund figure is net of fees and priced at noon UK time; the index is priced at the US market close. Both are in sterling, and the fund is not managed against that benchmark. An investor's tax, subscription date, withdrawals, and currency situation change the result received. The vehicle also lost 13.8% in 2022 and materially lagged the comparator in 2025. A shareholder who sold during weakness did not earn the since-inception return.

The record belongs to a defined share class, Fundsmith's investment team and process, investee-company managers, portfolio weights, service providers, and market prices. It is not Smith's personal account. The 2025 short-form report and prospectus provide the formal accounting, mandate, authority, and risk context behind the manager's letter.

Applying the Three Rules Outside Fundsmith

The usable process is specific: normalize operating capital, measure cash returns through a cycle, identify the cause of persistence, test the incremental growth runway, reconcile profit with cash, distinguish maintenance from growth investment, inspect leverage and management, estimate per-share free cash flow, compare its yield with risks and alternatives, then monitor the thesis with low unnecessary turnover.

Fundsmith's sterling base, institutional access, research resources, regulatory structure, concentrated portfolio, sector exclusions, fee schedule, and ability to absorb years of relative underperformance do not transfer automatically. The compact rules are useful because they organize hard work. Used without that work, "good company" becomes a compliment and "do nothing" becomes an excuse not to update.

Inside CompanyGraph

The population the question applies to is observable: companies whose return on equity, return on assets, and asset turnover all sit elevated against their own industry.