6.9 Quality frameworks: moats, Coffee Can, Lynch, Fisher, Kedia, Prasad

6.9.1Moats and the Coffee Can filter#

Moats (Buffett) A moat is a durable competitive advantage: brand trust, network effects, switching costs, cost advantages from scale or process, regulatory licences. In value-driver terms a moat is what keeps RONIC above WACC after competitors respond. Without one, competition drives RONIC toward WACC, and by (6.13) the value of growth goes to zero. Titan's Tanishq brand against unorganised jewellers is the Indian textbook example: purity trust is hard to copy.

Coffee Can (Mukherjea, Marcellus) The filter, exactly:

• Revenue growth of at least 10% in every one of the last 10 years, and
• ROCE of at least 15% in every one of the last 10 years.
• For lenders: loan-book growth ≥15% and ROE ≥15% in every year.
• A market-capitalisation floor of about ₹100 cr is reported in secondary sources.
• The portfolio is bought and left untouched for 10 years.

It is a test on every year, not on the CAGR. A company with revenue up 12% a year for nine years and 5% in one year fails, even though its 10-year CAGR (11.3%) clears 10%; with faster growth in the other years the CAGR of a failing company can exceed 12% (lab F3: a series with one 5% year and later 20% years has a 14.39% CAGR and fails). The every-year test selects for consistency: low variance of ROIC and growth, which in (6.11) means low uncertainty in the inputs of 6.7.

6.9.2Lynch, Fisher and Indian investors#

Peter Lynch's six categories (One Up On Wall Street, 1989). Slow growers (mature, low single-digit growth, held for dividends); stalwarts (10–12% growers, large and steady); fast growers (20–25%+, small aggressive firms, the source of multi-baggers and of the worst losses); cyclicals (earnings follow the economic or commodity cycle); turnarounds (recovery from distress); asset plays (assets worth more than the market price). Each category has different expected returns and different sell rules. For cyclicals the usual P/E logic inverts: P/E is lowest at peak earnings, which is often the time to sell (6.10).

Phil Fisher (Common Stocks and Uncommon Profits, 1958). "Scuttlebutt": gather information from customers, suppliers, competitors and former employees. His fifteen points cover the runway for sales growth, management's determination to develop new products, R&D effectiveness, the sales organisation, profit margins, labour relations, depth of management, cost accounting, candour with shareholders in bad times, and integrity. As an operator you can do scuttlebutt better than most: talk to distributors, check hiring and attrition, read the customer reviews.

Vijay Kedia's SMILE Small in size, Medium in experience, Large in aspiration, eXtra-large market potential. Management quality is the foundation. It is a stock-selection heuristic, not a valuation method; in value-driver terms it selects for a long runway of high $g$ with a small base.

Pulak Prasad (Nalanda Capital; What I Learned About Investing from Darwin, Columbia Business School Publishing, 2023). Three parts: (1) avoid big risks; (2) buy high quality at a fair price; (3) "don't be lazy, and be lazy": do the research, then hold for a very long time and reject most ideas. Nalanda avoids turnarounds, high leverage, poor governance and fast-changing industries. In the language of 6.7, avoiding big risks means rejecting companies whose value distribution has a fat left tail that the DCF cannot model.

Raamdeo Agrawal's QGLP: Quality, Growth, Longevity, Price. Longevity is the length of the competitive-advantage period.

6.9.3How long the advantage lasts#

A unifying model: the competitive-advantage period Suppose RONIC $= R > W$ on new investment for $T$ years, then RONIC $= W$ afterward. By Proposition 6.7, growth after year $T$ adds nothing, so value $=$ NOPAT$_1/W$ plus the PV of the economic profit created by investments made in years $1..T$. Moats, Coffee Can consistency and QGLP longevity are all claims that $T$ is long. Problem 6.6 derives the formula.