The Atomic Framework
At Atomic Moat Research, you and I are not gambling. We are engineering.
We view companies the way a private buyer views an acquisition. No ticker symbol. No price chart. Just the question a rational owner of the entire business would ask: what will this produce in cash over its lifetime, and is today’s price interesting relative to that? Everything else is noise.
Our methodology is built on the premise that a great business operates like a nuclear reactor — stable structure, efficient fuel consumption, the ability to generate enormous amounts of usable energy without melting down. The physics metaphors are not decorative. They are precise.
The Philosophy: Defying Gravity
In physics, perpetual motion is impossible. In finance, it is the Holy Grail.
The standard law of the stock market is mean reversion — the financial equivalent of gravity. High returns attract competition. Margins compress. Growth slows. Eventually, every star burns out.
We hunt for the anomalies. Companies with moats so structurally deep that competition cannot close them. Businesses that get stronger as they get bigger — self-sustaining reactions that compound capital for decades while the rest of the market tries to pick the next quarterly winner.
We do not care about the next quarter. We care about the next decade.
Two principles govern everything that follows.
Wonderful businesses at fair prices. Charlie Munger’s correction of Graham: stop looking for cheap businesses and start looking for excellent ones. A cigar butt gives you one last puff. A franchise compounds.
The private owner standard. The correct valuation anchor is what a rational buyer of the entire business would pay — not what a momentum trader bid for it last Tuesday. This single reframe eliminates most of the noise that passes for investment analysis.
The Four Filters
Every company passes through four filters in sequence. A failure at any stage ends the analysis. We never skip to valuation because the numbers look attractive. The filters exist to protect against exactly that temptation.
Filter I — The Circle of Competence
Can we describe, in plain language, how this business makes money, why customers stay, and what it will likely look like in ten years?
This comes first. Before the balance sheet, before the income statement, before the valuation. Because a DCF on a business you do not understand produces precise-looking nonsense.
The test is specific. Buffett’s formulation: “a part interest in an easily-understandable business whose earnings are virtually certain to be materially higher five, ten and twenty years from now.” Not probably higher. Not likely higher. Virtually certain. This standard eliminates most businesses immediately and is more demanding than any probabilistic framework.
Simple does not mean boring. Visa is simple. Baltic Classifieds is simple. A pharmaceutical biotech betting on a single Phase 3 trial is not simple — it is a binary event dressed in a lab coat.
If the answer to “how does this business make money” requires more than thirty seconds to explain to an intelligent non-specialist, we walk away. The circle boundary is there for a reason.
Filter II — The Reaction Chamber
What specifically prevents competitors from taking this company’s customers?
This is the core of the reactor. The reaction chamber generates the energy. The containment field keeps it from leaking away.
We begin with two tests before any formal scoring.
The Grizzly Bear Test. Would we rather wrestle grizzlies than compete with this business? If well-capitalised, intelligent competitors would find it prohibitively difficult to take meaningful share, the moat is real. If the honest answer is “we’d probably win eventually” — the moat is not what it appears.
The Patterson Test. Write one sentence explaining why a well-funded competitor cannot take meaningful share within five years. The sentence must name a specific structural barrier. “Customers really like us” fails. “We hold the only regulatory approval, obtained after a 12-year clinical programme, and switching triggers re-certification costs that kill the economics of migration” passes. If the sentence cannot be written clearly, the moat has been asserted, not proven.
Moats, ranked by durability:
Statutory / regulatory monopoly. Works council software embedded in German law. Intraoperative device approvals. Government-granted exclusivity. Hardest to replicate. Nearly permanent.
High switching costs. The customer would lose data, retrain staff, reintegrate systems, or face regulatory audit to leave. ERP, clinical software, asset finance platforms. Not permanent but extremely sticky.
Network effects. Value increases with scale. More participants make the network harder to leave. Classifieds monopolies, payment rails, professional networks. Durable as long as the #1 position is held.
Brand pricing power. Prices raised without meaningful volume loss. Durable but requires sustained reinvestment.
Cost / scale advantage. Structurally cheaper than any rival can replicate. Verify it is structural, not cyclical.
IP / patents. Meaningful only when combined with other moats. Alone, this is a weak moat that expires.
Two additional checks before leaving this filter.
Is the moat widening or narrowing? Signs of widening: pricing power increasing, churn decreasing, switching costs deepening. Signs of narrowing: competitors closing the gap, gross margins declining, customer concentration increasing.
The Technology Vector. Is technological change working for this moat, against it, or passing by it? Specifically: is the dominant competitive factor the technology connecting buyers and sellers, or the relationships and trust between them? Technology platforms where relationships are the technology are far less vulnerable to disruption than platforms where technology is merely the delivery mechanism.
Filter III — The Nucleus Check
Are these people good operators AND honest stewards of shareholder capital?
The nucleus holds the atom together. An unstable nucleus — management that allocates capital poorly, reports dishonestly, or prioritises their own enrichment over the business — destroys everything the reaction chamber generates.
Two tests run first.
The Retained Earnings Test. For every dollar retained by this business, is at least one dollar of market value being created? This requires ROIC on incremental capital above 15% sustained over time. Without this, retained earnings are a drain dressed as growth.
The Bloopers Test. Read the last three annual reports. Do the strategic decisions described read like obvious extensions of a winning formula — easy layups where the right answer is clear and the downside limited? These are bloopers. Good businesses present management with bloopers. Alternatively, do they read like agonising tradeoffs where capital must keep flowing with no conviction about the outcome? Those are agonies. Airlines, for example, present management with agonies. The Baltic Classifieds of the world present bloopers. The pattern is a structural moat signal that financial metrics cannot capture.
What we look for: owner-operators with skin in the game. Capital allocation that prioritises reinvestment at high ROIC over empire-building acquisitions. Shareholder letters that discuss failures honestly. Compensation structures aligned to 5–10 year outcomes. Debt used conservatively.
Special positive signal. When a CEO voluntarily conducts a substantial share buyback during a period when the stock has been ignored or sold off, this is one of the clearest management quality signals available. They are demonstrating, with their own company’s cash, that they believe the market is wrong. That is the opposite of every incentive pointing toward caution.
Red flags: excessive executive compensation relative to company size. Opaque segment reporting. Frequent large acquisitions at elevated prices. Management that takes credit for good results and blames macro for bad ones. Related-party transactions at non-arm’s-length terms.
Filter IV — Energy Output
Do the financials prove the moat is real, and is the price consistent with the thesis?
A reactor that glows bright but powers nothing is useless. We measure actual usable energy leaving the facility — not the accounting version of it.
The Geiger Test (balance sheet first). Before everything else: if a company cannot survive a nuclear winter on its own cash reserves, we walk away. Survival is the prerequisite for compounding. We check liquidity, leverage (net debt / earnings ≤ 3–5×), debt maturity timing, and goodwill quality. Net cash is preferred. Net debt is acceptable only for highly predictable businesses generating consistent free cash flow.
The Reaction Chamber (income statement). Revenue is a headline. Margins tell the story. Gross margin ≥ 40% and stable or rising over 10 years signals pricing power and moat reality. Declining gross margin is the first sign of a containment leak — competition eroding the moat. We label margins precisely: gross margin, EBIT margin, operating margin, EBITDA margin are four different numbers and we never confuse them.
Energy Output (cash flow). Profit is an accounting opinion. Cash is a fact. We measure owner earnings — the cash a buyer of the entire business could extract annually without impairing it.
Owner Earnings = Net Income
+ Depreciation & Amortisation
− Maintenance Capex
− Stock-Based Compensation
− Working Capital changes (if significant)This is Buffett’s 1986 definition. It is not EBITDA, which excludes maintenance capex and ignores dilution. Munger’s formulation: “Every time you see EBITDA, substitute the words ‘bullshit earnings’.” Stock-based compensation is always subtracted. Always. It is real dilution regardless of how management chooses to describe it.
We scrutinise FCF / net income conversion. If free cash flow consistently tracks well below reported earnings, something is being hidden in the working capital or the depreciation schedule.
The Valuation (last). We reach this step only after the qualitative filters have passed. The valuation confirms whether the price is consistent with the thesis already formed — it does not generate the thesis. If the model is needed to build conviction, the conviction was never there.
We build three scenarios (bear at 1/3 of historical growth, base at 1/2, bull at 85%), use owner earnings as the base, apply a 12% discount rate as the opportunity cost hurdle, and require a margin of safety proportionate to uncertainty: 15% minimum for statutory monopolies, 25–30% for high-quality compounders, 35%+ for anything with execution risk. If the gap between current price and conservative intrinsic value is large enough that imprecision in the estimate does not matter, the price is interesting.
The Critical Mass Scale
We do not issue buy or sell recommendations. We assess the state of the reaction.
★☆☆☆☆ — RADIOACTIVE Status: Avoid
The Audit: The thesis is broken or was never valid. The balance sheet is toxic, margins are collapsing, management is diluting shareholders to fund their own compensation, or the moat has been asserted without evidence. This is a meltdown in progress. Walking away is the correct action.
★★☆☆☆ — UNSTABLE Status: Pass
The Audit: High risk, inadequate reward. The reaction chamber is leaking — competition is eroding margins, management quality is questionable, or the valuation assumes a perfection the underlying business cannot deliver. Speculative at best. Dangerous if leveraged.
★★★☆☆ — STABLE ISOTOPE Status: Watchlist
The Audit: A high-quality business at a fair price. The reactor is running cleanly, the moat is intact, management passes the bloopers test — but the current price offers no meaningful margin of safety. We monitor these. We wait for Mr. Market to have a bad day.
★★★★☆ — FISSION Status: High Conviction
The Audit: The moat passes the Patterson Test and the Technology Vector is favourable. Management is demonstrably reinvesting at high ROIC. The price offers a genuine margin of safety. Owner earnings are growing and the FCF conversion confirms the reported earnings are real. The reaction has started.
★★★★★ — CRITICAL MASS Status: Core Holding
The Audit: The rare setup. An anomaly that defies the gravity of mean reversion. The moat is structural, the nucleus is stable, and the energy output is genuine — owner earnings compounding year after year with no meaningful threat in sight. These are the holdings carried for a decade while the rest of the market tries to outguess the next quarter. Generational compounders. We concentrate into these and hold.
Disclaimer
The metaphors are for visualization; the math is for verification.
Atomic Moat Research is a financial publisher, not an investment advisor. The content herein is for informational and educational purposes only. It represents the opinions of the author and does not constitute a recommendation to buy or sell any security. Financial markets are high-risk environments. We conduct deep due diligence, but we are not infallible. All readers are encouraged to perform their own independent research and consult with a licensed professional before deploying capital.