JB Hi-Fi vs Harvey Norman: same score, 23× the return
Two ASX retailers, nearly identical Buffett scores — and the one with the fatter margins was the worse business
Ideally, Warren Buffett likes to see a gross profit margin of 40% or more. A fat margin is one of the clearest signs of pricing power — customers paying up rather than shopping around. But it’s not the be-all and end-all. Buffett himself owns plenty of businesses with margins well below 40%, as long as the rest of the numbers are strong. Costco is the classic case: its gross margin is barely 13%, yet he and Charlie Munger have called it one of the great businesses of the age — Munger sat on its board for decades.
I learned that one the slightly hard way. When I built my portfolio in 2022, I let a fat gross margin sway me more than it should have. Two of the retailers I bought scored almost identically on my 100-point Buffett screen: JB Hi-Fi 64, Harvey Norman 60, just four points apart.
Sixty and 64 might look like modest scores, and they are — neither is a “perfect” or “exceptional” business. But both are good, solid, and above all consistent, and predictable earnings are something Buffett prizes: a business you can forecast is a business you can value. Truly exceptional businesses going cheap don’t come along often — and for Australian and New Zealand investors the pool of listed companies is far smaller than in the US, so “good and steady at a fair price” is often the best buy actually on offer.
Then the results came in. Since 2022, JB Hi-Fi — the one with far thinner margins — has returned 75.8%. Harvey Norman, with more than double the gross margin, has returned 3.3%. One made me roughly 23 times more than the other.
Two near-identical scores; a 23-fold gap in outcome. The answer is the most useful lesson I can give you about reading a screen — and about what a moat actually is.
The scores
JB Hi-Fi (JBH) — 64/100 (”Possible”) | +75.8% since 2022
Gross profit margin: 22% → 2/15
Net earnings trend: rose 5 of 7 years → 7/10
Operating expenses: 70% of gross profit → 1/5
EPS trend: rose 6 of 7 years → 10/10
Current assets vs total debt: 1.05× → 4/10
Debt safety: → 15/15
Return on equity: 32% → 15/15
CapEx: low → 10/10
Net profit margin: 4.7% → 0/10
Harvey Norman (HVN) — 60/100 (”Possible”) | +3.3% since 2022
Gross profit margin: 55% → 15/15
Net earnings trend: rose 4 of 7 years → 3/10
Operating expenses: 64% of gross profit → 1/5
EPS trend: rose 4 of 7 years → 3/10
Current assets vs total debt: 1.18× → 4/10
Debt safety: → 15/15
Return on equity: 14% → 5/15
CapEx: → 10/10
Net profit margin: 14% → 4/10
Look at the totals and they’re twins. Look at the rows and they’re opposites.
The mirror image
Harvey Norman wins one metric enormously: gross margin. At 55%, it keeps 55 cents of every sales dollar — full marks. JB Hi-Fi keeps just 22 cents — almost nothing on the screen’s scale. On the face of it, Harvey Norman looks like the one with pricing power.
But look at the row that actually predicts whether a business builds wealth — return on equity. JB Hi-Fi earns 32% on shareholders’ equity. Harvey Norman earns 14% — less than half.
That’s the whole story in two numbers. Harvey Norman has the fatter margins; JB Hi-Fi has the better business.
A fat margin isn’t a moat — it’s a symptom
This is the bit worth slowing down for, because it’s where most people (me included, once) go wrong.
A high gross margin is not a moat in itself. It’s often a symptom of one — but not always, and that distinction is everything. A genuine economic moat comes from one of about five places:
Intangible assets — a brand, patent or licence customers will pay up for
Switching costs — it’s a hassle or a risk for customers to leave
Network effects — the product gets more useful as more people use it
Cost advantages — you can produce or sell cheaper than anyone else
Efficient scale — you serve a market that’s only really big enough for one or two players
When a fat margin is the symptom of one of those — a beloved brand, say — it’s a wonderful sign. But Harvey Norman’s 55% isn’t really that. It’s largely a product of its structure: a franchise model and a big property portfolio. That property ties up an enormous amount of shareholders’ capital earning only a modest return — which is exactly why the fat margin never shows up as a fat return on equity.
Costco, from the top of this piece, is the proof of the flip side. Its moat was never in the margin — it’s a cost advantage and efficient scale: buy in colossal volume, sell on wafer-thin markups, lock customers in with membership. The thin margin isn’t a weakness; it’s the whole strategy. JB Hi-Fi is a smaller cousin of the same idea: thin margins, a low-cost, high-turnover model, a genuine cost advantage in its niche. Low margin, real moat.
Why the screen couldn’t tell them apart
The screen hands 30 of its 100 points to margin-related metrics — and margins are the one thing Harvey Norman is genuinely good at. So it scored almost level with JB Hi-Fi, even though the engine underneath was far weaker. The total couldn’t separate them.
A human reading the rows could, in about five seconds. One glance at the ROE gap — 32% versus 14% — tells you which business turns a dollar of capital into wealth and which mostly just sits on it. The screen narrows the field; it doesn’t make the judgement. That part is still yours.
The mistake I actually made: opportunity cost
Here’s the part I’d do differently. Harvey Norman wasn’t a disaster — it went up 3.3%. I didn’t lose money. But that’s the wrong way to keep score.
Charlie Munger banged on about opportunity cost his whole life: the true cost of any decision isn’t zero, it’s whatever your next best option would have done instead. Every dollar I put into Harvey Norman was a dollar I didn’t put into JB Hi-Fi — and JB Hi-Fi’s dollar worked more than twice as hard. Measured against zero, Harvey Norman was fine. Measured against the alternative sitting right beside it on my own screen, it cost me dearly.
And the clue was there the whole time, in the ROE row. I had the number; I just didn’t weight it heavily enough. The same went for the consistency I’d bought them both for: JB Hi-Fi grew earnings per share in 6 of the last 7 years; Harvey Norman managed 4, its profit lurching from $841 million in 2021 to $352 million in 2024. One was a metronome, the other a rollercoaster — and I’d told myself they were both metronomes.
What this means for you
Don’t buy the total score. Read the rows.
A screen is brilliant for narrowing hundreds of companies down to a handful worth studying. But two can reach the same number for opposite reasons, and the reasons are where the investment case lives. So:
Treat a fat gross margin as a clue, not a verdict — then ask which of the five moat sources, if any, is actually producing it
Read return on equity early, and weight it heavily — sustained high ROE without much debt is the truest single sign of a durable advantage
Judge every buy against your next best option, not against zero
JB Hi-Fi and Harvey Norman both scored “possible.” Only one had the engine of a genuinely good business underneath it — and the row that gave it away was sitting in plain sight the whole time.
Next time: how I actually put a price on a business once it’s passed the quality test — because, as Apple showed, a wonderful business at the wrong price is still a poor investment.
This is not financial advice. Returns shown are from September 2022 to June 2026 and relate to my own portfolio. Past performance does not guarantee future results. The information in this newsletter is educational and based on publicly available data. It does not take into account your personal financial situation, goals, or risk tolerance. Always do your own research or consult a licensed financial adviser before making investment decisions.