Investment legends like Buffett and Chris Bloomstran teach is that, for value investors, the returns come from compounding over long periods of time. The wisdom of their words certainly took a long time to compound in my head: only recently, after 20 years of reading, did the penny finally drop. What follows reverse engineers the compounding engine step by step, so you can save a few decades.
Christopher Bloomstran, the founder and Chief Investment Officer of Semper Augustus, writes great annual letters. I’ve been reading them religiously since 2011. My style is value investing, and like many others I grew up on Berkshire Hathaway letters and aim to find great compounders to invest in. But nowadays when people ask me what investment related ideas to study, I recommend they study Chris.
Not that the Berkshire letters are bad. But Chris digs deeper into the accounting. That may not sound like a compelling reason to prefer his writing. It’s never boring though, and it connects many dots that help me understand the financial mechanics in a business I don’t always grasp from the more generalized comments Buffett wrote. Bloomstran is a big Berkshire fan, arguably its best analyst, and the conglomerate is invariably a big Semper Augustus position. Like Buffett and Munger, Bloomstran’s MO is to identify great businesses and hold them for a long time to let compounding work at full force. If this investment style fits you, like it does for me, I can’t think of a better teacher than Chris.
Despite reading Chris for a long time, it wasn’t until his most recent 2025 letter that I grasped how compounding really works. I never truly saw the actual financial machinery producing the outsized returns. The missing insight was the different role of equity as measured by book value, and as measured by market value. What I hadn’t realized is that having shareholders reinvest dividends is not at all equivalent to keeping money in the business. Let’s unpack this by pulling the covers off the compounding engine.
”
It is remarkable how much long-term advantage people like us have gotten by trying to be consistently not stupid, instead of trying to be very intelligent.
”
―
Step 1 – Understanding profitability, or how powerful the business engine is.
The sentence in the annual letter that stood out for me was this comment (p. 15)
“You’ve likely heard us say a million times that profitability is properly measured against the capital of the business, not against sales.”
I had heard him say it many times, but it had never clicked. I stopped reading and started thinking, then scribbled notes that eventually became this text.
Not coincidentally, the sentence appeared in a section discussing retailers. Why is that relevant? To understand that, let’s compare the numbers of two very different retailers: Dino Polska, a Polish supermarket chain selling groceries and staples, and LVMH, a global seller of luxury products.
Dino Polska has sales of 33.6B PLN and net income of 1.56B, a profit margin of 4.6%. LVMH does 3 times that. It realizes 11.2B EUR of net income on sales of 80.8B EUR, a profit margin of 13.9%. But now consider Bloomstran’s metric of net income compared to total capital employed in the business. It reorders the ranking. Dino Polska total capital is 8.2B, consisting of 8.7B in equity and net cash (or negative debt) of 0.5B, for a total capital base of 8.2B. LVMH also doesn’t use a lot of debt. They run on 80.5B EUR in total capital, about 80% of which is equity. Expressed on this basis, Dino Polska’s profit margin is a 19.0% return on total capital, while LVMH realizes 13.9%. Buffett understood this too. It’s part of why Berkshire owns a railroad.
The reason both metrics point to a different profitability champion is the clock speed of the business. It’s an overly simplified view, but imagine the total capital in the business is used to buy inventory. Dino Polska only earns a little on each sale of diapers or detergent. But it sells and replenishes inventory 4.5 times per year. LVMH earns more on each sale, however those fine cognacs spend a lot of time slowly aging in barrels, and the elegant leather handbags too patiently wait on shelves for a buyer, an entire year on average. What you make on each sale matters, but sales velocity does too.
Step 2 – Retained Earnings, or redirecting the exhaust to power a turbo.
Dino and LVMH take different paths to earning their income, and follow equally different paths in what they do with it. Management has a couple of redeployment options: use the money to pay shareholders a dividend, buy back shares, or reinvest it in the business. Money spent on dividends and buybacks is not available for reinvestment in the business, like Retained Earnings are. This is why some of the best compounders never pay dividends. For this reason, Bloomstran looks at Retained Earnings and wants that to be a sizable fraction of net income.
If management reinvests retained earnings on behalf of the shareholders, the logical question is what return they can expect on this investment. Before we tackle that question, let’s see how Dino Polska and LVMH allocate their profit. (We’ll simplify here by treating net income as equal to cash available for reinvestment.)
LVMH hands out while Dino Polska retains. Compared to 11.2B in net income, LVMH paid 7.1B in dividends: about two thirds. Dino Polska does not send cash back to its shareholders. There is no dividend. 100% of net income is Retained Earnings for reinvestment in the business. (It shares this allocation policy with Berkshire Hathaway, which has never paid a dividend.) Exhaust can’t turbo-charge an engine if it isn’t retained.
Step 3 – Return On Equity, or the turbo driving the compounding engine.
The reason retained earnings drive the compounding effect is they are added to equity and can therefore be expected to earn the Return On Equity with Equity defined as Book Value. That is not the case for newly invested money, including for example reinvested dividends. Even if dividends are not taxed, which in many jurisdictions and for many types of shareholders they are, shareholder money coming in earns a different return: Return On Equity with Equity defined at Market Value. Another look at our two exhibits illustrates this.
Both companies have nice returns on book value equity: 16.3% for LVMH, and 18.0% for Dino Polska. Assuming both businesses can maintain that level of financial performance, those are the respective returns their shareholders can expect on the retained earnings. Coincidentally, both shares traded at a Price to Book ratio of 4.7 at the end of last Fiscal year.
Price to Book matters. A Dino Polska zloty retained in the business will return 18.0% a year from now. But reinvesting an LVMH dividend euro will only return 16.3% divided by 4.7, or about 3.5%, as the money needs to pay the market price for equity. The difference between retaining and reinvesting: eighteen percent, or three and a half.
The difference is stark, if you trust the numbers. Book Value is malleable and accounting tricks can distort it. Chris knows this. He is no naïve consumer of financial reporting, but critically applies qualitative corrections to his analyses. Read the 2015 letter and its detailed valuation of Berkshire Hathaway for another masterclass. It discusses how practices like (supposedly one off) asset write-downs distort the picture. We shouldn’t take financial statements at face value but correct them for distortions or window-dressing. But this doesn’t invalidate the principle that a high amount of retained earnings are the driving force of the compounding engine. A more realistic assessment of Book Value tells us what return to expect. And if distortions are so excessive the corrected numbers get us to reject an investment, so much the better.
It’s not like LVMH management doesn’t know all of that. They could retain more, but their business is more mature with limited opportunity for investment and growth. Dino Polska still has plenty of runway. If LVMH shareholders want to reinvest their dividend in a compounder they might prefer Dino Polska, but then too they’ll need to initially buy those shares at market value. Compounding is wonderful for continuing shareholders, not for new ones.
Step 4 – Consider the share count.
Owning a growing business is nice, unless you need to share it with an increasing number of others. Money spent on share buybacks is not available for reinvestment as retained earnings. Many companies issue shares to employees, and buy those shares in the open market, typically at a price exceeding book value, to keep the share count stable. Nevertheless the cost of that is a drain on the retained earnings available for compounding. Share buybacks below book value are beneficial to the continuing shareholder as they boost our engine, while buybacks above book value weaken it. Disciplined capital allocators know the difference. LVMH and Dino Polska fall in that category. Neither company seems to manipulate book value, and both managements have kept share count stable in recent years. Chris and Warren would approve.
As aspiring individual value investors, we are told that a long time horizon can give us an edge over professional investors with constant short term pressure to keep up with the market, or with constraining investment mandates. We are advised to try and emulate the Berkshire method, at our more modest scale. Easier said than done. I admire Buffett as much as anyone, and his writings have taught me a great deal. But they can read like the chronicles of an ace pilot. Bloomstran’s letters on the other hand read like hands-on advice from the experienced engine mechanic. It feels like Chris is showing us the trade secrets, and teaching us what to care about and look for. If you too want to be a better value investor and improve your results, sign up to the school of Semper Augustus. Lessons are free and there’s no limit on students. It took this slow student fifteen years. But the story of compounding, it turned out, is a tale of two equities.
”
Now we come to Howard himself, Gary Cooper. Ayn Rand was one of Cooper’s biggest fans from the time she emigrated from Russia and worked in Hollywood as an extra. She was of course thrilled beyond belief when he agreed to play Howard. There is a photograph of the short Rand gazing up at the chiseled, handsome Cooper, and she’s practically drooling. After Rand worked – I can’t remember if it was months or years – on Howard’s big speech in the courtroom, Cooper told her after he finished filming it that he never understood the speech. I’m fairly certain he didn’t understand the rest of the role either and that he had never read the book. A more glorious-looking, charismatic man to play Howard you couldn’t have found, but did he understand this role the way he understood Lou Gehrig? I doubt it. Did Rand, for all her artistic integrity care? I doubt it. In the end, that great philosopher, that giant intellectual Ayn Rand was, in reality, a woman like any other.
”
―
Building a Berkshire by the Book
Credits
Words
> Stefan Verstraeten
Ideas
> Christopher Bloomstran’s 2025 annual letter for Semper Augustus.
>
Photo
> Engine Room in the basement of Bourse de Commerce museum in Paris, France. Art installation by Laura Lamiel. Photo by Stefan Verstraeten.
Video
>
