The Letter — Sample issue: two companies, the same apparent bargain, opposite books

Why this issue

Most market letters ask you to trust them. This one takes the opposite bet: to convince you by showing, never by asserting.

So we opened two real files, from the first figure to the last. Two American companies whose published accounts, read on a given date, carried the same apparent price on their assets — and said the opposite of one another underneath. Neither is a buy suggestion: both are historical, and the point is not to hand you a trade but to let you watch a method work on documents anyone can pull up.

You will see two bargains that look identical on the surface split apart on a single line of the accounts, and then what the market did with them in the years that followed. Nothing here rests on genius or on foresight. Everything rests on process: separating what can break a business from what can lift a stock, and refusing to let one answer hide the other.

Let us open the files.

Act I — The same bargain, three hundredths apart

First file. On 17 February 2023, M/I Homes — a homebuilder based in Ohio — filed its 2022 accounts. The stock closed that day at $57.05. The balance sheet put net assets at $74.28 per share: you were paying 0.77 times what the company owned.

Second file. On 28 February 2022, Teladoc Health — American telemedicine, one of the most talked-about names of the pandemic — filed its own. The stock closed at $75.91, against book value of $102.24 per share: 0.74 times net assets.

Stop there for a moment. At this stage a conventional screen would have returned them side by side, and would in fact have ranked Teladoc ahead: it was the cheaper of the two on its assets. Three hundredths separated the two files.

This is precisely the situation that loses money. An apparent bargain is where value investing starts; it is never where it concludes.

What Graham and Buffett tell us

Let us pause, because these two cases illustrate to the letter three ideas that are more than a century old and have not aged a day.

The first is the distinction between price and value. Benjamin Graham, in The Intelligent Investor, imagines "Mr. Market": a partner who every day offers to buy your stake or sell you his, at a price set by his mood. Some days he is euphoric and overprices everything; on others, despondent, he dumps. The intelligent investor does not catch the mood; he uses it. But you still have to know what you are buying on the other side — because Mr. Market sometimes dumps for very good reasons.

The second is the margin of safety, which Graham held to be the three most important words in all of investing. Buying well below value gives you a cushion: nobody knows the future, and the most careful estimate can be wrong. But the margin has to sit on something solid. A discount on assets that produce earnings is a cushion; a discount on assets that consume them is a slope.

The third is quality, which Warren Buffett added to his teacher's legacy, under Charlie Munger's influence. Buffett moved beyond original Graham-ism — the "cigar butts" you pick up for one last free puff — toward sound businesses bought at a fair price. "It is far better to buy a wonderful company at a fair price than a fair company at a wonderful price," he liked to say. That is exactly the line that will separate our two files.

Act II — What the two scores read

This is where the method stops being a matter of culture and becomes a matter of measurement. Two questions are put to every company, and they are deliberately kept apart.

Can it hold up? That is what the health score estimates — how unlikely it is that a business falls apart. It is read in the balance sheet and the income statement, never in the share price.

The first two rows are twins. Every row below them is opposed.

M/I Homes earned 23.7 cents on every dollar of equity, ran a 15% margin, and covered its interest 283 times — which is to say it had no debt problem at all. Teladoc was losing money on its operations: negative margin, negative earnings per share, and negative interest cover, which does not mean "tight" or "thin": it means operating profit was not paying the interest, and the money had to come from somewhere else.

Is it cheap? That is what the Value score estimates. For M/I Homes it was wide and backed by something solid: nothing in a price of 0.77 times net assets required the business to improve. It only had to keep going. For Teladoc, the upside assumed a turnaround — that the losses would stop, that the margin would return. That is not a discount, it is a wager.

One thing this reading never does is issue a licence. The method does not certify that a business will hold; it says how unlikely a collapse looks given the accounts available that day, and leaves the decision to you.

What became of the two files

Here are the prices, on the date of analysis and today. They are public, and verifiable on any chart.

One hundred dollars in the first was worth 267; in the second, eight were left.

What this table shows is not a talent for stock picking: it is the mechanical consequence of a reading made on the day the accounts were filed. The line that separated the two files — one made money, the other lost it — was legible to everyone, for free, on the date it was written. It simply was not being looked at, because both stocks carried the same price on their assets.

✦ What these two files demonstrate is a reading — and it can be reproduced. The accounts were filed, the ratios can be recomputed, and the line that separated the two businesses was legible to anyone on the day. That is exactly what the club does every morning, across the whole universe rather than on two files. Every situation is its own, and past performance is not a guide to future performance.

The method, without the recipe

What we can say fits in a few lines. Our universe is U.S. small and mid caps, from $200 million to $5 billion — because that is where, statistically, value stays the least picked over, and therefore the most often mispriced. Every morning, every company is scored on the two questions above. Scores are read as a rank within that day's cohort, never against a fixed threshold: score levels depend on the market regime, and a threshold that made sense in 2015 does not describe today's market.

Those scores feed two rankings. Avoid gathers the lowest decile of health — what the method flags as fragile, and the list we hold to be the most useful of the two: it is where a file like Teladoc's would have sat. The Shortlist brings together the sound and cheap small and mid caps — health that holds and a high Value score. A stock only leaves a list once it has crossed back over the boundary with room to spare, so that it does not flicker in and out from one morning to the next.

What we will not say is how these elements are weighted, at what thresholds a stock enters or leaves, nor the calibration that assembles them. Not out of coyness: calibration is precisely what separates a list of known criteria — those have belonged to everyone for a century — from a method that produces a result. It is re-examined at each closed year and kept only if it still holds.

What we publish instead of a scoreboard

You may have noticed what this issue does not contain: no annualised return, no track record, no portfolio, no comparison to an index.

That is a choice, and it follows from the requirement set out above. We publish what you can redo. The filings of M/I Homes and Teladoc are lodged; their prices are public; every ratio in this issue can be recomputed from documents anyone can consult. You have just checked a method without once having to take our word for it.

An in-house scoreboard cannot be audited. You would have to trust us on the universe kept, on the entry and exit dates, on the files quietly dropped along the way. That is asking for exactly what this letter refuses to ask of you.

So we publish the rule instead. What puts a company into a ranking, what takes it out, on which data and on what date — then, every morning, the result of that rule applied to the entire universe. You can check it against the filings, as you have just done, twice. That is a higher bar than a performance figure, not a lower one.

In a word

If you take one thing from this issue, let it be this: cheap is not a conclusion, it is a question.

M/I Homes and Teladoc carried the same price on their assets, three hundredths apart. One made money, the other lost it — and that line, legible on the day, was worth one hundred and sixty-seven percent on one side and ninety-two on the other.

The method does not work because it is brilliant. It works because it is disciplined. It reads the accounts, it separates what can break a business from what can lift a stock, it refuses to let one answer hide the other, and it says so on the day it reads them.