Tom Gayner

Insurance can supply long-duration investable funds, but claims own the first call on that money. Patient equity capital begins only after reserves, liquidity, currency, duration, and regulatory stress have been covered.

Float Is a Liability Before It Is Capital

Tom Gayner is Markel Group's Chief Executive Officer and Chief Investment Officer. The 2026 proxy says he oversees capital deployment across its insurance, industrial, financial, and consumer businesses. That is a wider role than managing a stock portfolio and a different institution from a redeemable fund.

An insurer commonly receives premium before the covered period ends or the final claim is known. The timing creates float, but the money is attached to obligations. Markel's 2025 Form 10-K defines net policyholder funds through unpaid losses, unearned premiums, insurance payables, and life benefits, reduced by items such as premium receivables, reinsurance recoverables, prepaid reinsurance, and deferred acquisition costs.

Underwriters price the risk; actuaries and claims teams estimate ultimate losses; finance records reserves; regulators and rating agencies constrain capital; investment staff need cash and bonds available when claims are paid. A reserve is a recorded estimate of a future physical cash outflow. Calling it conservative does not pay the claim if the estimate is wrong.

Liability Matching Comes First

Gayner's 2025 shareholder letter states the ordering directly: Markel first invests enough in high-quality fixed income to more than cover insurance reserve liabilities. The 10-K says the bond portfolio generally matches the duration and currency of loss reserves, with principal protection aimed at claim payment.

That division separates policyholder and shareholder capital. High-quality bonds and cash support near and medium-term obligations. Shareholder capital can then bear more public-equity volatility, subject to stress and regulatory requirements. Markel says its insurance portfolio holds a larger equity proportion than typical insurers and consequently must retain more capital in regulated subsidiaries.

Permanent capital therefore does not mean pressure-free capital. There may be no fund redemption, but catastrophes, reserve deficiencies, credit loss, market declines, rating pressure, subsidiary dividend limits, and acquisition commitments can all reduce feasible action. Patience is financed by balance-sheet strength, not by corporate form alone.

Premium arrives. Reserves and capital are established. High-quality fixed income is matched to claims. Only genuine excess and shareholder capital can move toward volatile equities, whole-business acquisitions, debt reduction, or Markel repurchases.

Underwriting Determines the Cost of Float

Markel Insurance wrote $12.5 billion of gross premium volume in 2025, including $10.6 billion of underwriting premium and $1.9 billion of fronting premium. The segment recorded a 95% combined ratio, its eighteenth underwriting profit in twenty years. A ratio below 100% usually means earned premium exceeded incurred losses and expenses before investment income.

The composition matters. Markel reported $484 million of prior-year reserve redundancies, worth 5.8 combined-ratio points, and twenty-one consecutive years of redundancy. That history supports the record of cautious reserving, but it also shows that current results include updated estimates of earlier claims. Accident-year performance, pricing, catastrophe load, fronting fees, reinsurance, and reserve development need separation.

Gayner's letter combines underwriting profit and interest income associated with float into a company-defined "total spread." It reports more than $11.5 billion of cumulative positive spread over 2005-25 and $1.3 billion in 2025. This explains the engine: when underwriting pays for its own costs and the linked assets earn interest, the holding company receives useful funding. It is not a GAAP measure of investment alpha, and costly claims can reverse the economics.

The Four Filters Apply After the Balance Sheet Test

For public equities, Gayner states four filters. The business should earn good returns on capital with minimal debt. Management needs talent and integrity. The company needs reinvestment opportunity or demonstrated discipline in acquisitions, dividends, and repurchases. Markel must pay a fair price.

Each term needs evidence. Historical return can decline as capacity or competitors arrive. Integrity is difficult to infer and cannot replace internal control. Reinvestment may widen an advantage or fund an empire. A repurchase creates per-share value only below a reasonable estimate of value. Minimal debt at an investee does not remove Markel's own insurance liabilities.

In 2025 Markel's public-equity portfolio earned a 10.5% total return. The letter reports an $8.9 billion unrealized gain and a deferred tax liability near $2 billion using a 22% rate. Deferring tax while appreciated shares remain unsold provides financing, but the liability remains recorded and a sale or law change can alter timing.

This is not Gayner's personal return. It is one component of Markel's accounts, produced by investment personnel, investee-company managers, portfolio weights, market prices, taxes, and the insurance balance sheet that holds many securities. The result is also distinct from Markel stock, which includes every operating business and liability.

Controlled Businesses Change the Available Actions

Markel now reports industrial, financial, and consumer-and-other segments rather than one undifferentiated "Ventures" engine. Its controlled companies operate across equipment, services, plants, construction products, and financial activities. The industrial businesses sent $145 million of dividends to the holding company in 2025 after capital spending and working-capital needs.

Public-stock ownership and control are not equivalent. Markel can vote or sell a quoted security, but it usually cannot set the operating budget. In a controlled subsidiary it can choose leaders, financing, incentives, and the destination of excess cash. That authority creates more ways to improve a business and more responsibility for failure.

Gayner describes subsidiary managers as long-term owners who are not loaded with debt or pressed to meet unrealistic budgets. They can reinvest locally; cash that is not needed can move to the parent for another use. The holding company compares internal projects, public equities, acquisitions, fixed income, preferred redemption, debt, and its own shares.

Decentralization Needs Visible Accountability

Markel's 2025 insurance reorganization is a useful test of the institutional philosophy. Gayner wrote that the operation had become too centralized: underwriting and claims authority moved too far from customers, while accountability for costs and results weakened. Markel appointed Simon Wilson as insurance CEO and gave distinct operations one accountable profit-and-loss owner.

The physical claim and broker decision happen close to the front line; consolidated accounts record the aggregate later. Moving authority closer can shorten feedback, but decentralization also risks inconsistent controls. Clear limits, common reserving, escalation, audit, and comparable records still have to connect local judgment to group capital.

Markel also sold the renewal rights of its reinsurance operation after concluding that adequate returns would require substantial scale and better uses of capital existed elsewhere. It retained old reserves to run off. That separates stopping new production from extinguishing the liabilities already written and shows that permanent ownership does not prohibit exit when incremental economics fail.

Company Performance Is Jointly Produced

The 2025 letter reports a closing Markel share price of $2,149.65 and a five-year compound annual growth rate of 16%. It separately states management's estimate that intrinsic value per share grew 12% in 2025, 15% annualized over five years, and 16% since the initial public offering. Market price is observable; intrinsic value is a management estimate, not an audited return.

Markel's result reflects insurance pricing, claims, reserve changes, investment income, public shares, controlled-company teams, acquisitions, financing, repurchases, taxes, and the market's valuation. Gayner directs much of the allocation but does not personally produce those operating flows. The 2025 results release appropriately reports contributions from every segment.

What Survives Outside Markel

The transferable sequence starts with obligations. Map their amount, timing, currency, and stress behavior. Hold liquid assets against them. Only then estimate risk capital. Seek high-return businesses with restrained debt, capable and honest managers, credible reinvestment or distribution discipline, and a fair price. Compare every use of cash and measure the result per share.

Insurance float, regulated subsidiaries, deferred tax financing, public-company borrowing, acquisition access, claims expertise, and authority over operating CEOs do not transfer to an individual. The four filters travel more easily than the balance sheet. Used without the liability map, the checklist misses the institutional condition that makes Gayner's patience possible.

Inside CompanyGraph

CompanyGraph tracks the quality-compounder print live: companies whose earnings arrive backed by operating cash while revenue growth and cash margins read elevated together.

Cash-Backed Growth Configuration

Three present-state observations co-occur: OCF/Net Income elevated, revenue growth composite elevated, and trailing OCF margin elevated

Cash-Backed Growth Configuration
growth consistency
ocf to net income
ratio cashflow income opcf margin
Open in Screener

A match is a recorded configuration, not a compounding forecast. Durability, reinvestment runway, and price are the judgments the screen leaves open.