Nick Sleep

Scale Economies Shared

Nick Sleep ran the Nomad Investment Partnership with Qais Zakaria from 2001 to 2014, returning roughly 20% a year net before voluntarily winding the fund up and returning capital. His central idea is "scale economies shared": most companies that grow take the savings of scale as profit, while a rare few — Costco and Amazon are his canonical examples — deliberately hand those savings back to customers as lower prices. That choice suppresses reported margins in the short run and compounds the customer proposition in the long run, creating a moat competitors cannot cross without destroying their own economics. Sleep paired this with extreme patience: Nomad held a handful of positions for many years and measured itself over decades, arguing that the main edge available to an investor is time horizon rather than information.

Methodology fidelity

Thematic proxySources reviewed 2026-08-24

The Nomad letters strongly support scale economies shared, destination businesses, owner-minded management, robustness, and an unusually long horizon. They do not publish an eight-factor screen or the app’s 7% growth, 12% return-on-capital, 8% capital-intensity, 70% cash-flow, or 5-of-8 thresholds. The margin pattern in row 1 is an observable app interpretation of shared scale, not a universal Sleep formula.

Valuation Method

⚙

No intrinsic value computed

No investor formula

Sleep's "destination value" — imagining the business at maturity and asking what it would then be worth — was described in prose in the Nomad letters and never reduced to a formula. Publishing a number for it would invent precision the source does not contain, so this framework scores criteria only.

Screening implementation (8 criteria)

#CriterionMetricThresholdData Source
1Scale Economies SharedRevenue growth vs. gross and operating margin trendRevenue rising; gross margin flat or falling; operating margin holdingIncome Statement: Revenue, Gross Profit, Operating Income (3-year trend). Falls back to an operating-margin-only reading when gross profit is untagged (43.5% of the corpus), at reduced confidence.
2Revenue Growth Runway5-year Revenue CAGRCAGR > 7%Income Statement: Revenue, 5-year history
3High Return on CapitalROIC (5-year); banks use ROEAbove 12% in a majority of yearsNOPAT / Invested Capital (banks: Net Income / Shareholders Equity)
4Low Capital IntensityCapEx / Revenue (3-year average)Below 8%Cash Flow Statement: CapEx; Income Statement: Revenue
5Deep Moat / Destination BusinessCustomer-captured economicsIs this where customers end up, and why would that change?Qualitative: Nomad letters on Costco, Amazon and the shared-scale model
6Owner-Operator ManagementManagement incentives and time horizonDoes management think and hold like an owner?Qualitative: proxy statements, shareholder letters
7No DilutionNet change in share count over one yearShare count flat or falling, net of issuancestandardized_statements.shares_retired_1y_pct — used instead of a payout measure because dividends and buybacks are untagged for roughly two thirds of the corpus
8Robustness Over TimePositive Operating Cash Flow frequencyPositive in at least 70% of available yearsCash Flow Statement: Operating Cash Flow, 10-year history

Verdict Logic

⚖

PASS when at least 5 of 8 criteria hold, applied as a proportion of the criteria actually evaluated so sector-gated companies face the same hurdle. The framework deliberately produces no intrinsic value and therefore no BUY/NEUTRAL split.

What Would Nick Sleep Buy Today?

Sources

📖

Nomad Investment Partnership Letters

primary2001–2014

Nick Sleep and Qais Zakaria

The complete partnership letters, in which the scale-economies-shared model is developed and applied to Costco, Amazon and Berkshire. The primary and near-only source for the framework.

View source →
📖

Nomad Investment Partnership Final Letter

primary2014

Nick Sleep and Qais Zakaria

The wind-up letter, which sets out why the partnership returned capital and restates the case for very long holding periods as the durable edge.

View source →