Pieter Slegers — I'm Buying The Whole Business
A break from stock selection: after reading 7,121 pages of Berkshire, Buffett and Watsa material, the case that owning whole cash-generating businesses — financed with debt and topped up with insurance float — is the difference between compounding at 10% and 20% a year. Ends with a survey soliciting $250,000 minimum commitments to do it together.
One-line take: the first issue in this archive that is not about buying shares at all. The premise is stated as a personal turning point — "What if you would buy a majority stake in a (small) listed company yourself, just like Warren Buffett did with Berkshire? And then you compound from there?… this time it 'clicked' for me." The evidence is a rehabilitation of Buffett's self-declared worst trade: he paid $14.86 a share against $32.30 of assets — a 0.46x multiple of book — and the textile business he called a mistake still returned $20.3 per share of cumulative operating profit over 1965-1974, more than the entry price. Layered on top is the float argument, spelled out in beginner's terms (a $1,500 annual car premium against an $8,000-$10,000 claim once every 17-18 years) and then to its logical end: "If Berkshire Hathaway would use it's operating profit and float to just copy the S&P 500, by definition it will outperform the index because they have 'free money' to invest… That's exactly why I think Berkshire will keep outperforming going forward." The mechanics are then made reproducible: a downloadable Excel model, and a worked example on Security National Financial ($SNFCA) — a Salt Lake City life-insurance / mortuary / mortgage company at 0.6x book, 3.3x EV/EBITDA, 6.5x earnings and 3.1x free cash flow — showing EUR 4.22bn after 20 years on the EBITDA framing versus EUR 606m from just owning stocks (a 6.96x edge, 21.2% CAGR), or EUR 2.46bn versus EUR 704m on the P/E framing (3.49x, 17.1% CAGR), both assuming 50% acquisition debt at 5%. The commercial intent is explicit and new: a Google Form asking readers whether they want to buy a company together or invest in a fund that does it for them, "the minimal amount to participate would be $250.000." Two cautions the letter does not raise about its own numbers: every result is levered 50/50 with tax-deductible interest (the edge is as much the debt as the multiple), and SNFCA is used purely as an arithmetic input — there is no argument that this business is acquirable, no control premium, and no quality assessment of the kind the rest of this archive insists on.
1. Stocks & names mentioned
A methodology issue — only three names appear, and only one carries figures. Stance follows the post's own framing: Berkshire is argued Positive on the float mechanism; Security National is a worked arithmetic example, presented as statistically cheap but with no quality assessment and no recommendation, so it is Neutral; Fairfax appears only as a body of reading. Research legend: QT Qualtrim · SA Seeking Alpha · STK Stock Analysis. Written post with no timestamps — the At link opens the article.
| Ticker | Name | Research | View | What he said | At |
|---|
| BRK.B | Berkshire Hathaway | QT · SA · STK · FA | Positive | An explicit forward call, argued from the float rather than from the portfolio: "If Berkshire Hathaway would use it's operating profit and float to just copy the S&P 500, by definition it will outperform the index because they have 'free money' to invest in the index. That's exactly why I think Berkshire will keep outperforming going forward." The famous "dumbest stock I ever bought" is re-argued as a good one on the archive's own arithmetic — $14.86 paid against $32.30 of assets (0.46x), and $20.3 per share of cumulative textile operating profit returned over 1965-1974 on $418.5m of revenue and 1,017,547 shares. The float chart runs from ~$26bn in 2000 to ~$175bn at 4Q 2025 (~$124,000 per A share). No valuation, no price, no position — the name is the proof of the mechanism, not a pick. | read ↗ |
| SNFCA | Security National Financial Corporation | QT · SA · STK · FA | Neutral | The archive's first mention, and used as the worked example of the whole thesis — not as a recommendation. "A perfect example to make this concept clear? Let's take the listed stock Security National Financial Corporation." A 1965-founded Salt Lake City company in life insurance, cemetery and mortuary services, and mortgages — "As you can see, it's a very boring business. And it's cheap: Price-to-book 0.6x, EV/Sales 0.6x, EV/EBITDA 3.3x, P/E 6.5x, P/FCF 3.1x." Fed into the published Excel model at 50% debt / 5% interest / 5% growth / 25% tax over 20 years, it produces EUR 4,219,897,217 versus EUR 606,080,470 from simply investing (6.96x edge, 21.2% CAGR, 30.3% earnings yield on EUR 90.09m of equity) on the EBITDA framing, and EUR 2,458,392,436 versus EUR 704,032,870 (3.49x, 17.1% CAGR, 15.4% earnings yield on EUR 104.65m of equity) on the P/E framing. Note what is absent by this archive's own standards: no moat, no ROIC, no Quality Score, no management assessment, no float-quality test — and the insurance business here is a life insurer whose float is long-dated and reserved, not the P&C float the article celebrates. | read ↗ |
| FFH.TO | Fairfax Financial Holdings | QT · SA · STK · FA | Neutral | Named only as source material — "All shareholder letters of Prem Watsa of Fairfax (772 pages)", one of the three bodies of reading behind the article, alongside Adam Mead's Berkshire history (1,337 pages) and Buffett's letters and transcripts (5,012 pages). No stance, no figures, and no reference to the 30-share position bought on 16 August or to its role in the 1 September upweighting. Its presence is nonetheless the tell: the float argument in this issue is the same one that justified buying Fairfax three weeks earlier, now generalised from a share purchase to an outright acquisition. | read ↗ |
Three things the issue does not say about its own model. (1) The edge is leverage as much as price. Every headline figure assumes 50% of the purchase price is debt at 5% with tax-deductible interest and a 25% tax rate; the "just investing" comparator is unlevered. A like-for-like unlevered run is not shown. (2) The stress test is on, and disclosed only in the assumptions image — a year-7 shock halving earnings/EBITDA and refinancing at 9%. The results survive it, which is the strongest part of the case and is never mentioned in the prose. (3) The prose and the model are two different examples. The narrative uses a $5m business throwing off $1m of free cash flow ($54.2m versus $171.0m over 25 years); the model uses EUR 54.6m of EBITDA and EUR 32.2m of net earnings — SNFCA-scale, not $5m-scale. The concept is unaffected but the numbers are not comparable, and the article presents them as one continuous argument.
2. Talking points
The premise, and where it came from
- Framed as a post-Berkshire-AGM obsession: "The only thing I could think about during the Berkshire weekend was the following: What if you would buy a majority stake in a (small) listed company yourself, just like Warren Buffett did with Berkshire? And then you compound from there?"
- "And while we all know that Warren Buffett did this, this time it 'clicked' for me. So that's why I went deep. Really deep."
- The stated research base: Adam Mead's The Complete Financial History of Berkshire Hathaway (1,337 pages), all of Buffett's shareholder letters and public transcripts (5,012 pages), and all of Prem Watsa's Fairfax letters (772 pages) — 7,121 pages. "Luckily you don't have to as I did so for you."
Rehabilitating Buffett's "dumbest stock"
- Buffett in 2010: "The dumbest stock I ever bought was — drum roll here — Berkshire Hathaway." The counter-argument: "I'm quite convinced that without this investment, Buffett wouldn't have become as successful as he is today."
- The entry: accumulated 1962-1965, first at $7.6 a share, average $14.86. "Berkshire Hathaway was a classic cigar butt stock" — working capital alone was $19 a share, and cash, receivables and inventories were $20.8 million.
- The napkin: price paid $14.86, value of assets $32.30, multiple paid 0.46 — "Warren Buffett paid less than half of the book value of the company."
What the textile business actually returned
- From Mead's Table 3.16 (1964-1974): total textile revenue $418.5 million, total operating profit $20.7 million, on 1,017,547 shares outstanding.
- Per share: cumulative revenue $411.3, cumulative operating profit $20.3. "Paid $14.86 per share. Got $20.3 back in 10 years."
- The verdict is deliberately modest: "It's definitely not the best investment in the world, but it's also not terrible." The point is that a dying business bought at 0.46x assets still returned more than its price — the downside case, not the upside one.
Float, explained from scratch
- "Float is the money an insurance company holds between collecting your insurance premium and paying out a claim. Customers pay upfront but claims come later."
- The worked example: a $1,500 annual car policy against an average accident once every 17-18 years costing $8,000-$10,000. "Until the car accident takes place, Berkshire Hathaway can invest these premiums in stocks and bonds."
- "If you do well as an insurance company and you are profitable… You receive the float for free. It's free money that is not yours." The conditional — if you underwrite profitably — is stated once and then dropped; the rest of the argument treats float as costless.
- Berkshire's float chart: roughly $26bn (2000) → $57bn (2010) → $113bn (2019) → $161bn (2022) → ~$175bn (4Q 2025), or about $124,000 per A share.
The structural claim about Berkshire
- "If Berkshire Hathaway would use it's operating profit and float to just copy the S&P 500, by definition it will outperform the index because they have 'free money' to invest in the index. That's exactly why I think Berkshire will keep outperforming going forward."
- Illustrated with a share-price chart: Berkshire above +1,000% since 2000 against roughly +300% for the S&P 500.
- Worth noting how unusually strong this is for the archive — a forward outperformance claim made with no price, no multiple and no expected-return calculation, in a letter series that normally refuses to buy anything without all three.
10% versus 20% — the compounding arithmetic
- The setup: a "very stable boring company" generating $1 million of free cash flow, bought outright for $5 million. "Under the assumption that the company remains constant, you will recoup your entire investment in 5 years."
- "This would mean that from year 6, you will have an extra million to invest every single year (the cash flow of the business). If you can do this for 20 years, the difference gets ridiculous."
- Situation 1 (invest $5m for 25 years at 10%): $54.2 million. Situation 2 (buy the company, recoup in 5 years, invest the cash flows): $171.0 million. "That's three times as much profit! It's the difference between compounding at 10% per year versus compounding at 20% per year."
- "And if you can add insurance float to the equation, it gets even better."
The SNFCA worked example
- Security National Financial Corporation: founded 1965, Salt Lake City, three segments — life insurance, cemetery and mortuary services, mortgages. "As you can see, it's a very boring business."
- The statistics: P/B 0.6x, EV/Sales 0.6x, EV/EBITDA 3.3x, P/E 6.5x, P/FCF 3.1x. "Let's just say that we could acquire the company for 5x it's cash flow. In that case we could recoup our investment within 5 years and use the cash flows after year 5 to keep buying other companies or to just invest in the stock market."
- No moat test, no ROIC, no management assessment, no discussion of whether a controlling stake is actually available. This is the archive's usual process running in reverse: the cheapness comes first and the quality question is not asked at all — the exact failure mode confessed to a week earlier in Part I ("Almost every time I made a buy decision because I thought the company was somewhat quality but definitely cheap, it was a mistake in hindsight").
The model, and what is in the assumptions
- A downloadable Google Sheet ("You need to download the Google Spreadsheet in Excel before the formulas work"), with an Assumptions tab and a Summary tab.
- Shared inputs: 10% market return, 5% interest on acquisition debt, 50% debt, 5% earnings growth, 25% tax rate, 20-year hold, debt strategy set to "pay down first".
- Stress scenario, switched on: a year-7 shock halving earnings/EBITDA, with the refinancing rate resetting to 9%. This is the most defensible feature of the model and the prose never mentions it.
- The debt is doing a large share of the work — "0% = all equity, 80% = max leverage" — and the comparator ("just investing") is unlevered. No sensitivity table for the debt share is shown.
The two framings, and why they disagree
- EBITDA framing (EUR 54.6m EBITDA at 3.3x): strategy EUR 4,219,897,217 at 20 years versus EUR 606,080,470 from just investing — a 6.96x edge, 21.2% CAGR, on EUR 90.09m of equity invested at a 30.3% earnings yield.
- P/E framing (EUR 32.2m net earnings at 6.5x): strategy EUR 2,458,392,436 versus EUR 704,032,870 — a 3.49x edge, 17.1% CAGR, on EUR 104.65m of equity at a 15.4% earnings yield.
- The explanation given is simply the entry multiple: "The reason for this is that the multiple you pay differs." The deeper reason — that EV/EBITDA is pre-interest and pre-tax while P/E is post both, so a levered buyer double-counts the benefit in the EBITDA framing — is not addressed.
The five conclusions
- "Buffett's 'biggest mistake' became one of his greatest investments ever thanks to the power of long-term compounding."
- "The secret of Warren Buffett? Insurance float + buying entire companies."
- "Buying an entire cash-generating business can create much higher returns than simply investing in the stock market."
- "Strong cash flows can be reinvested into new acquisitions, creating a powerful compounding machine over time."
- "Finding one undervalued business with recurring cash flows can be enough to build exceptional long-term wealth."
The ask — a new commercial direction
- The issue closes with a Google Form: "Are you interested in buying an entire company together? Or in investing in a fund that does this for you?" — two options, plus a required free-text field.
- "Please note that the minimal amount to participate would be $250.000."
- This is a material widening of the product. Everything archived here to date is research and a model portfolio; this is a solicitation of interest in a private acquisition vehicle or fund at a $250k minimum, aimed at the paid list. No structure, jurisdiction, fee, or regulatory framing is disclosed — it is explicitly only a survey of interest.
- Read against the 1 September policy issue (the book to shrink from 21 names to 15-20, "developed countries only", a stricter bar), the direction of travel is consistent: fewer, larger, more concentrated positions — and now, potentially, positions of control.
3. In plain English
BRK.B — Berkshire Hathaway Positive
Berkshire is used here to explain a specific advantage, not as a stock pick. An insurance company collects premiums today and pays claims years later. In between it holds a large pile of other people's money, called "float", and it can invest that money. If the insurance side roughly breaks even, that money is effectively free to borrow.
Slegers' argument is arithmetic: if Berkshire simply put its profits and its float into an index fund, it would still beat the index — because it is investing borrowed money that costs nothing. Berkshire's float has grown from about $26 billion in 2000 to roughly $175 billion at the end of 2025. That is why he says Berkshire should keep outperforming from here.
He also revisits the purchase Buffett calls his worst-ever. Buffett paid $14.86 a share for a dying textile mill whose assets were worth $32.30 a share — less than half of book value. Over the following decade the mill threw off $20.3 a share of operating profit, more than the purchase price. The lesson taken is not that textiles were good, but that buying an entire business cheap enough is hard to lose on, even when the business itself is poor.
SNFCA — Security National Financial Corporation Neutral
Security National is a small Salt Lake City company that sells life insurance, runs cemeteries and funeral homes, and originates mortgages. Slegers picks it because it is dull and statistically cheap: it trades below the accounting value of its own assets, at about six and a half times profits and about three times the cash it generates.
It is a demonstration, not a recommendation. He asks what would happen if you bought the whole thing rather than a few shares — paying half in cash and half with a loan at 5%, then using the company's own cash flow to pay down the debt and afterwards to buy more assets. Over twenty years his spreadsheet turns that into roughly seven times more money than simply owning stocks on one set of assumptions, and roughly three and a half times more on another. Both runs include a deliberate disaster in year seven, where profits halve and the loan is refinanced at 9%.
Two things are worth holding onto. Most of the advantage comes from the borrowed half of the purchase price, and the comparison it beats — "just investing" — uses no borrowing at all. And nothing in the analysis asks whether this is a good business, whether anyone could actually buy control of it, or what price a control buyer would have to pay above the market quote. By the standards this newsletter applied to itself a week earlier, that is the exact mistake it promised to stop making.
FFH.TO — Fairfax Financial Holdings Neutral
Fairfax appears here only as reading material — 772 pages of Prem Watsa's shareholder letters, one of the three sources behind the article. There is no view expressed and no numbers given.
It is still worth noting, because Fairfax is the Canadian company built on exactly the mechanism this issue is about: underwrite insurance, hold the float, invest it. Slegers bought a position in it three weeks earlier and named it for an increase a week earlier. This issue is that same idea taken one step further — from owning shares in a company that does this, to owning a company outright and doing it yourself.
Summary derived from the archived Compounding Quality post (text and transcribed tables in transcript.txt) for personal study. Not investment advice. © Compounding Quality / Pieter Slegers for source material.