Compounding Isn’t a Finance Trick - It’s a Behavioral Edge
Everyone talks about compounding. Few have the patience to let it happen.
We don’t “do” compounding. We behave in a way that allows compounding to do us.
That’s it. That’s the whole edge.
Everyone can run a DCF, sort a screener, or memorize the latest factor alphabet soup. Very few can buy great businesses at fair prices, then sit still while the math of inevitability goes to work. Patience isn’t a mood. It’s a methodology. And around here, “do nothing” is an active verb.
But that invites a hard question…
Okay, then what is a “high-quality” asset?
Every fund deck says the same thing: “We buy high quality at a fair price.”
Sounds great. Now define “high quality” without waving your hands.
Is it revenue growth?
Fat margins?
Low P/E?
Dividend yield?
“Great management”?
Moat?
There are a thousand data points but only one first principle: a business is a machine that turns invested dollars into more dollars, at attractive rates, for a long time.
If we want one number that forces us to look at both operating excellence and capital discipline, we need a metric that connects the income statement to the balance sheet.
Enter Pulak Prasad and the importance of Return on Capital Employed (ROCE).
ROCE: the cleanest starting line
As Pulak teaches, if you want to understand the quality of a business, don’t start with stories - start with ROCE:
ROCE = Operating Income (EBIT) ÷ Capital Employed,
where capital employed is net working capital + net fixed assets (excluding excess cash).
Why EBIT? Because we want the operational engine before financing and tax architecture muddy the water.
Why “capital employed”? Because quality isn’t just what a business earns - it’s how much capital it must chain to the floor to earn it. ROCE forces that trade-off out into the open.
A few practical notes we follow:
Use history, not promises. We start with historical ROCE across a full cycle (10 years minimum) - not slide-ware. Pulak’s method is to let the track record speak before we do.
Exclude excess cash. Cash mountains flatter ROCE and tell you nothing about the core engine.
Prefer operating truth to accounting cosmetics. Margins can look pretty while inventory bloats; ROCE catches that. Growth can look heroic while capital needs quietly spike; ROCE catches that too.
Pulak’s team at Nalanda built a whole process around this, and I’ve shamelessly cloned it: filter first on consistently high ROCE, then only then do the deep work on the business, people, incentives, durability, and price.
He calls it selecting a single trait that “brings many others along for free.” High ROCE correlates with capable managers, sound capital allocation, real competitive advantages, and the oxygen to take smart risks without courting ruin.
Pulak’s three-step philosophy that I’ve adopted:
(1) Avoid big risks.
(2) Buy high quality at a fair price.
(3) Don’t be lazy - be very lazy (hold on).
On step (1): he avoids crooks, serial acquirers, debt junkies, fast-changing fads, turnarounds, and misaligned owners. It’s the Munger rulebook of eliminating stupidity first - or as Pulak shows, the evolutionary advantage of organisms that minimize fatal errors and tolerate missed opportunities. Missed out on Tesla? Fine. We play for decades, not news cycles.
On step (2): high ROCE is the front door to “quality.” It is not the whole house. But if the door is rotten, don’t bother touring the kitchen.
On step (3): once you own a compounding machine, the work becomes behavioral. Get out of the way and let time do what time does.
Why ROCE beats the usual shortcuts
Growth ≠ quality. You can grow yourself to death if each incremental dollar earns a subpar return. ROCE asks: what’s the return on the next dollar?
Margins lie. Costco runs low net margins by design and still creates enormous value because it turns capital fast and reinvests wisely. Tiffany ran higher margins for years - Costco still produced superior ROCE and value creation.
P/E is a weather vane. It tells you what others feel today, not what the business will earn on its capital tomorrow. High-ROCE businesses can look “expensive” on headline multiples and still be mispriced if their reinvestment runway is long and disciplined.
“But isn’t management quality the real edge?”
Yes - and ROCE is how it shows up in the world. You can’t sustain 20%+ ROCE across cycles with a clumsy capital allocator, absent moat, or weak culture. Sustained high ROCE is evidence that incentives, operating discipline, and advantage are aligned.
Our screen (and why it’s behavioral, not just analytical)
We begin with historical ROCE over a full cycle (10yrs).
If a company can’t clear that bar, we don’t pretend to be smarter than the numbers.
If it does, then we earn the right to study:
Durability (industry structure, customer captivity, replacement cycles)
Reinvestment runway (can they put large amounts of capital to work at similar returns?)
People & incentives (will they resist diworsification, buybacks at peaks, and M&A sugar highs?)
Price vs. value (we still need a fair price for the stream of high-ROCE compounding)
The outcome we want is simple: buy high-ROCE compounding machines at fair prices, then be “very lazy.”
Patience isn’t optional - it’s the mechanism.
As Nick Sleep put it (another hero we shamelessly learn from): you can buy “fifty-cent dollars,” but the big money comes from letting those dollars compound once they reach fair value and keep earning superior returns year after year.
The discipline before the discipline: Avoid big risks
This is where Pulak’s Darwin lens - and Munger’s avoidance of stupidity - really helps.
We skip:
Governance hair (crooks, cash leaks, “trust me” accounting)
Turnarounds (heroic narratives; dismal base rates)
High leverage (debt narrows optionality and magnifies error)
Serial M&A (most deals destroy value; the opportunity cost is invisible but massive)
Fast-changing arenas (hard to skate to where the puck will be)
Avoiding fatal errors is how you stay available for compounding. Protect the downside; let the upside take care of itself.
As Buffett framed it, rule #1 and rule #2 both rhyme with don’t die - Pulak translates that into process.
Why this is a behavioral edge (and not a spreadsheet trick)
We start with a filter that punishes wishful thinking. High, sustained ROCE across time is earned, not narrated.
We accept boredom. Most of the work is not trading; it’s not interrupting compounding. That’s emotionally hard.
We keep our circle of competence small. If ROCE is great but the engine is something we can’t truly explain, we pass.
We leave room for “no.” Saying “I don’t know” is cheaper than learning with real money.
Our house style (Stewarding Legacies)
Stewardship means we spend every dollar like it belongs to our kids - not our egos.
We filter on historical ROCE, then underwrite durability, people, and price.
We avoid big risks first. We buy boring excellence. We let time brag on our behalf.
Compounding isn’t a trick.
It’s a behavior - patience made visible.
Appendix: Quick ROCE notes (how we do it)
Definition: ROCE = EBIT / (Net Working Capital + Net Fixed Assets).
We generally exclude true excess cash and, for acquisitive companies, we evaluate the capital invested in deals separately.Window: 10 years (full cycle). We want consistency, not a one-year spike.
Thresholds: We lean in above ~20%+ average through the cycle (context matters by industry).
Adjustments:
Remove non-recurring gains/losses from EBIT.
Normalize for obvious cycle extremes.
Watch inventory turns and receivables - ROCE dies quietly there.
Interpretation:
High, sustained ROCE suggests capable management, real advantage, and a culture that understands capital.
Low or falling ROCE (amid growth) is a yellow flag - growth may be value destructive.
Pulak illustrates this perfectly: if a company is growing revenue at 10% or more but earns less than 20% ROCE, it will need external cash to fund that growth. But if ROCE exceeds 20%, the business can self-fund expansion and still accumulate cash. In other words, below 20% ROCE, growth consumes capital; above 20%, growth creates it. That’s the real dividing line between a good business and a compounding machine.
Munger footnote (because it’s fun to be reminded)
“The big money is in the waiting.”
“It’s remarkable how much long-term advantage people like us have gotten by trying to be consistently not stupid.”
“Knowing what you don’t know is more useful than being brilliant.”
Amen.
 | A guest post by
|