Coffee Can Portfolio

The original concept of the coffee can portfolio was articulated by Robert Kirby of Capital Group in 1984, drawing on a metaphor of the American West where settlers kept valuables in a coffee can hidden under the mattress — set it and forget it. Kirby observed that client accounts where he had bought stocks and been prevented from trading (due to client inattention) had often dramatically outperformed the actively managed portions of the same client's portfolio over many years. The insight was that inactivity compounded into extraordinary returns when the underlying businesses were genuinely excellent.

Saurabh Mukherjea adapted and formalised the coffee can framework for Indian equity markets through his writing — including the books Coffee Can Investing (co-authored with Rakshit Ranjan and Pranab Uniyal, published in 2018) and Diamonds in the Dust — and through the investment strategies of Marcellus Investment Managers. The Marcellus version of the coffee can required constituent companies to satisfy demanding quantitative filters over a rolling ten-year lookback period: revenue growth of at least 10% per year in each of the ten years, and return on capital employed (ROCE) of at least 15% in each of the ten years. Very few companies in the Indian listed universe passed both filters simultaneously across a full decade, typically producing a concentrated portfolio of fifteen to twenty names.

The rationale for extreme holding discipline was rooted in the mathematics of compounding. If a portfolio of ten companies compounded capital at 20% per year without any trading friction — no transaction costs, no capital gains tax, no behavioural errors from over-trading — the theoretical advantage over a strategy with identical stock selection but active trading was substantial over a 10-20 year horizon. India's long-term capital gains tax regime further incentivised holding periods exceeding one year to access the 10% LTCG rate (post the 2018 budget reintroduction of LTCG), though the coffee can logic extended well beyond one-year tax considerations.

The companies that typically qualified under the Indian coffee can criteria shared observable characteristics. They operated in industries with durable competitive advantages — either through brand, distribution, regulatory moats, or network effects. They were asset-light relative to their revenue generation, which produced high ROCE without requiring large capital reinvestment. Their promoters had demonstrated integrity across business cycles — no accounting irregularities, no equity dilution to benefit insiders, no related-party transactions that extracted value from public shareholders. This governance filter, while not formally codified in the quantitative screen, was emphasised strongly by Mukherjea as essential to the strategy's long-term validity.

Critiques of the coffee can approach centred on survivorship bias (the strategy only looked elegant because we knew which companies had survived and thrived over the prior decade), concentration risk (holding fifteen stocks meant a single fraud or business implosion had outsized impact), and the challenge of identifying in advance which companies would sustain their metrics for the next decade rather than just the previous one. Additionally, the valuation at which one entered a coffee can portfolio was not irrelevant — buying companies meeting the criteria at excessive PE multiples could result in long periods of subdued returns even if the underlying businesses continued to perform.

Despite these critiques, the coffee can framework served an important educational purpose in Indian retail investing by emphasising quality, holding discipline, and the corrosive effects of over-trading and transaction costs on long-run compounding — principles broadly applicable even if the specific quantitative filters were debated.