The Investment Manual

A guide to the concepts used in every per-company equity research note on Compoundex.

Pillar 0 · The Business in Plain Terms

In this section we set out what the business does, in two voices.

Plain version. Here we explain the business as if speaking to a seven-year-old child or a friend who has never heard of the company. No jargon. Concrete examples. We do not assume the reader knows what a CPU is, what HBM3e memory is, or what a hyperscaler is. The test is whether a non-finance reader closes the section able to explain the business to someone else.

Technical version. Here we explain the same business in finance and industry terms. The number of reportable segments, how revenue is disclosed at sub-levels, the manufacturing or operating model, the names of the principal customers and suppliers, the CEO and tenure. The test is whether a practitioner closes the section with a working mental model of the cash machine.

Pillar 1 · Circle of Competence

In this section we evaluate whether the business sits inside our circle of competence. Buffett's first filter. If we cannot explain it, we cannot value it.

Business segments. Segments are the divisions a company uses to organise its operations and report revenue. They matter because each one usually behaves differently. One may be cyclical while another is secular. One may carry a wide moat while another is commoditised. Aggregating segments into one company total hides the most important information about the business.

How the company earns a dollar of profit. Every dollar of operating profit comes from somewhere. The highest-margin dollar is generated by the product, customer, or geography with the strongest pricing power. The lowest-margin dollar is generated by the commodity-like end of the business. The mix of high-margin and low-margin dollars is what determines the blended margin, and the mix shift is what determines the future of the blended margin.

Competitors. Competition is the rivalry between firms for the same customer dollar. Understanding the competitive structure of an industry is the difference between knowing what a business is today and knowing what it can become. We focus on the top three competitors because three is the smallest number that captures the structure of most industries. For each one, we identify where they sit, the specific axis on which they compete (price, performance, software, distribution), and what their incentive structure does to our business.

Three main customers. Customers are who actually pays. Identifying them tells us where the recurring revenue actually comes from and what would happen if any one of them changed providers. For business-to-business companies we name the specific buyers. For consumer-facing companies we characterise the customer cohort by their economic profile and what brings them back.

Industry economics. These are the structural forces that shape how the entire industry works. Five dimensions matter most.

• Cyclicality. Whether revenue swings with the macro economy. A cyclical business looks cheap at the top of the cycle and expensive at the bottom. The inverse is the trap.
• Capital intensity. Whether the business requires ongoing capital investment to maintain its competitive position. High capital intensity ties up cash that could otherwise return to shareholders.
• Customer concentration. How concentrated the customer base is. If the top two customers account for one quarter of revenue, the loss of one is material.
• Supplier concentration. How concentrated the supplier base is. If a single supplier provides ninety percent of a critical input, the business carries that supplier's geopolitical and operational risk.
• Regulatory and geopolitical. Which regulators have jurisdiction and what risks they create. Export controls, tariffs, banking rules, antitrust review, sanctions.

Verdict. The closing judgment on whether the business sits inside our circle of competence well enough to proceed to a valuation. If the answer is no, the analysis stops here.

Pillar 2 · Business Quality and Durable Moat

In this section we measure whether the business has a durable competitive advantage.

ROIC. Return on Invested Capital. The percentage return a business earns on every dollar reinvested into operations. The formula is net operating profit after tax divided by invested capital, where invested capital equals debt plus equity minus cash. Buffett's threshold for high-quality compounders is roughly fifteen percent. Below the cost of capital, the business destroys value on every dollar reinvested. Above the cost of capital, the business compounds shareholder wealth at the rate of the spread. Comparing ROIC to WACC is the cleanest single test of whether a business is creating or destroying value.

Gross margin trajectory. Gross margin equals revenue minus cost of goods sold, divided by revenue. The trajectory matters more than the level. A widening gross margin signals improving pricing power, a mix shift toward higher-margin products, or operating leverage. A narrowing gross margin signals commoditisation or input-cost pressure. We track the three-year direction and the expected direction over the next three years.

Moat under Dorsey's framework. Pat Dorsey, in The Little Book That Builds Wealth, organises competitive advantage into four categories. Each is graded for the business at hand.

• Intangible assets. Brand, patents, regulatory licences, network ownership. Intangibles create pricing power and survive industry shocks. A wide intangible moat is what allows a company to charge prices that competitors cannot match.
• Switching costs. What it costs the customer to leave. Time, money, integration risk, retraining. Switching costs lock in revenue and protect pricing across cycles.
• Network effects. The product becomes more valuable as more people use it. Two-sided networks like payment cards, marketplaces, and exchanges create the strongest version of this advantage.
• Cost advantage. Structurally lower unit cost than peers. Walmart on supply chain, TSMC on scale, Costco on procurement.

Moat trajectory. Widening, stable, or narrowing. Moats are not static. New entrants, regulatory change, technology shifts, and customer behaviour can widen or narrow a moat over time. The trajectory is what matters for the next ten years of compounding.

Verdict. Wide, narrow, or none. The verdict sets the required margin of safety in Pillar 7. Wide-moat businesses earn a lower required cushion. Narrow-moat businesses require more. Businesses with no moat are speculations at any price.

Pillar 3 · Management Quality and Capital Allocation

In this section we test whether the people running the business are good operators and disciplined allocators of capital.

CEO and operational track record. The Chief Executive Officer is the single largest off-balance-sheet asset of the business. We name the CEO, state how long they have been in the role, and recount the specific operational achievements that prove they can execute. Long-tenure CEOs who have weathered a full cycle reveal more about quality than recent appointees. Generic praise for management does not count. Numbers and milestones do.

Capital allocation. Capital allocation is how management deploys the cash the business generates. There are four main uses: reinvestment in operations, acquisitions, dividends, and share repurchases. Two of these reveal the most about discipline.

• Buybacks. Repurchases of shares done at prices below intrinsic value create shareholder value. Repurchases above intrinsic value destroy it. The test is the average price paid versus intrinsic value at the time of the buyback.
• M&A. Acquisitions reveal management style most directly. The strategic logic, the price paid, and the operational discipline after closing are graded separately. A good deal at a high price can become a bad deal. A bad deal at any price destroys capital.

Insider ownership. The percentage of shares outstanding held by the CEO and senior management. Buffett-tier owner-operators (Berkshire, Constellation Software) hold above twenty percent. Standard large-cap executives hold half a percent to two percent. The level itself is less important than the trajectory and the open-market buying or selling activity.

Compensation ties. The formula management is paid against. Compensation that vests on stock-based-compensation-adjusted EPS gives management an incentive to issue more stock and inflate the per-share metric. Compensation that vests on operating income, free cash flow, or relative total shareholder return aligns management with long-term shareholders. The 10-K Compensation Discussion and Analysis discloses this.

Verdict. Excellent, good, adequate, or weak. CEO succession risk is named explicitly because it is the single largest unmodelled risk in many cases.

Pillar 4 · Financial Health and Balance Sheet

In this section we test whether the business survives a bad year. The metrics differ depending on whether the business is industrial or a bank, because the two business models work differently.

For industrial and technology businesses

Net debt to EBITDA. Net debt equals total debt minus cash. EBITDA is earnings before interest, taxes, depreciation, and amortisation, a rough proxy for cash earnings. The ratio measures how many years of cash earnings it would take to clear the debt. Below two times is conservative. Above four times is risky. A negative ratio means the business holds more cash than debt: a fortress balance sheet.

Interest coverage. Operating income divided by interest expense. Measures how many times the business can service its debt obligations from operating cash. Above five times is safe. Below two times is dangerous.

Debt maturity wall. The schedule of when the debt comes due. Concentrated maturities create refinancing risk if interest rates rise or credit markets close. A laddered profile spread across many years is safer than a single large tranche due in twelve months.

Liquidity buffer. Cash on hand relative to annual operating needs. If the business loses access to credit markets entirely, how long does it survive on its own balance sheet?

Goodwill as percent of total assets. Goodwill is the premium paid above book value in past acquisitions. High goodwill (above thirty percent of total assets) raises the risk of future impairment charges that can shock reported earnings.

For banks and lenders

CET1 ratio. Common Equity Tier 1 capital divided by risk-weighted assets. The cleanest single measure of a bank's loss-absorption capacity. The US regulatory minimum is four point five percent. Large banks operate near eleven to thirteen percent. Above twelve percent is fortress for a deposit-funded bank.

Funding mix. Where the bank gets its money to lend. Deposits, debt, securitisations, or other sources. Deposit-funded banks are the most stable because customer deposits are sticky and cheap. The funding mix is a structural cost advantage or disadvantage.

Net Interest Margin (NIM). The spread between what the bank earns on its loans and what it pays for its funding, divided by interest-earning assets. A higher NIM is structurally better. NIM compression is a key risk to watch in any lender.

Net Charge-Off Rate (NCO). The percentage of loans the bank actually wrote off as uncollectible in a given period, annualised. The credit cycle metric. Premium card books should not see NCO sustained above three point five percent. Personal loan books carry higher acceptable thresholds.

Loan loss provisions. Reserves the bank sets aside against expected future loan losses. Under CECL (Current Expected Credit Loss accounting), banks must reserve for the full life of the loan at origination. Rising provisions signal a deteriorating credit cycle.

Tangible book value. Total equity minus goodwill and other intangibles. The hard equity backing the loan book. Banks are often valued as a multiple of tangible book.

Pillar 5 · Earnings Quality

In this section we test whether the earnings the company reports are real. Cash earnings and accounting earnings can diverge for legitimate reasons or aggressive ones. Different metrics apply to industrial businesses and to banks.

For industrial and technology businesses

FCF over net income. Free cash flow divided by reported net income. FCF equals operating cash flow minus capital expenditure. A ratio above one means cash earnings exceed accounting earnings. Below one means accounting earnings are stronger than cash. Persistent ratios below zero point eight raise quality concerns.

OCF over net income. Operating cash flow divided by net income. Excludes capital spending. Above one point five is healthy. Below one is a warning sign that working capital or non-cash gains are inflating reported earnings.

SBC as percent of revenue. Stock-based compensation paid to employees, as a percentage of revenue. The Compoundex methodology charges SBC as a real cost in the DCF and never adds it back. When the company issues new shares to employees, existing shareholders are diluted. SBC of four to five percent of revenue is normal for technology. Above eight percent is a quality flag.

Audit opinion. The independent auditor's view on the financial statements. A clean opinion from a Big Four firm is the baseline. Material weaknesses or qualifications disclosed in Item 9A of the 10-K are red flags.

For banks and lenders

Fair-value loan marks. Some lenders hold loans at fair value rather than amortised cost. Fair-value accounting recognises gains as marks improve, before any cash actually changes hands. This can lift reported GAAP earnings above cash earnings when marks are favourable. The quality of these gains depends on the assumptions inside the marks.

Gain-on-sale revenue. Revenue recognised when originated loans are sold to investors rather than held to maturity. The gain is real cash but it pulls future earnings forward. Banks with high gain-on-sale revenue can appear to grow earnings faster than the loan book actually supports.

Provision rate. Loan loss provisions as a percentage of average loans. Persistently low provisions during benign credit cycles can mean either disciplined underwriting or aggressive accounting. We compare to peers.

Stock-based compensation. Same convention as industrial. Charged in full, never added back.

Pillar 6 · Valuation

In this section we estimate what one share of the business is worth. Industrial businesses are valued by discounted cash flow. Banks are valued by residual income on tangible equity. Both produce a value per share that we compare to the market price.

For industrial and technology businesses

DCF. Discounted Cash Flow. The model projects unlevered free cash flow for five years, applies a terminal value beyond, and discounts everything back to today at the cost of capital. The output is one estimate of intrinsic value per share.

UFCF. Unlevered Free Cash Flow. Operating profit after a normalised tax rate, plus depreciation and amortisation, minus capital expenditure, minus the change in net working capital. The cash the business produces before any lender is paid.

WACC. Weighted Average Cost of Capital. The blended return investors expect across equity and debt, weighted by capital structure. The cost of equity is built using CAPM: risk-free rate plus beta times equity risk premium. The cost of debt is the after-tax interest rate the company pays on its borrowings.

Terminal growth rate. The rate at which cash flow grows forever beyond the explicit forecast. Capped at long-run US GDP (about three percent) for businesses without a wide moat. No company can compound above the economy forever.

Exit multiple. A market multiple applied to terminal-year EBITDA, as if the business were sold at the end of the forecast. Averaged with the Gordon perpetuity to avoid leaning on a single assumption.

Three scenarios. Bear, Base, Bull. Each scenario flips the segment growth and margin assumptions. The Base case follows analyst consensus in the near years and decelerates to terminal growth. The Bull case credits operational potential we have conviction in. The Bear case assumes the thesis fails.

My Case. The analyst's working view. Usually Base on most lines, tilted to Bull on the highest-conviction drivers. The figure the analyst would bet on, not the consensus answer.

Reverse DCF. Starts with the spot price and asks what assumptions the market is already pricing in. If the spot price requires terminal growth of five percent and a terminal EBIT margin of thirty-five percent, the market is paying for a bull thesis.

Owner-earnings yield versus Graham hurdle. Owner-earnings yield equals free cash flow divided by market capitalisation. The Graham hurdle is the risk-free rate plus the equity risk premium, the minimum return an investor should demand from any equity position. Clearing the hurdle means the price offers an acceptable cash return today. Falling short means the case rests on future growth.

Triangulation. The DCF base, the reverse DCF, and the owner-earnings yield are compared. Agreement strengthens conviction. Disagreement is a flag.

For banks and lenders

Residual income on tangible equity. Net income minus a charge for the cost of equity applied to tangible book value. The model projects residual income forward and discounts back at the cost of equity. The output is a value per share.

Cost of equity. Risk-free rate plus beta times equity risk premium. For a bank the discount rate is the cost of equity, not WACC. Deposits and borrowings are operating funding, not capital structure.

Return on Tangible Equity (ROTE). Net income divided by average tangible book value. The bank equivalent of ROIC. Above twenty percent is high quality.

Price-to-tangible-book. Market capitalisation divided by tangible book value. A useful sanity check. Premium banks trade at multiples of tangible book.

Three scenarios. Same Bear, Base, Bull convention as industrial. For a bank the switches are revenue growth, net interest margin, cost of equity, terminal growth, and credit-loss assumptions.

Pillar 7 · Margin of Safety

In this section we compare the intrinsic value from Pillar 6 to the market price and grade the cushion. This is the most important pillar.

Fair Value MOS. Margin of Safety. The formula is intrinsic value minus spot price, divided by intrinsic value. Positive when intrinsic exceeds spot. Negative when spot exceeds intrinsic. A positive MOS expressed as a percentage tells the buyer how much the analysis can be wrong before the position loses money.

Required MOS bands. The threshold scales with business quality.

• Wide-moat business in stable industry. Approximately thirty percent.
• Narrow-moat business or cyclical industry. Thirty-five to forty percent.
• Bank or lender. Thirty-five to forty-five percent because credit risk compounds uncertainty.

Margin of Safety Table. Four rows: Bear, Base, My Case, Bull. Each row shows the intrinsic value, the spot price, the MOS percentage, and a one-line interpretation. The Base row is highlighted gold as the visual anchor because Base is the decision row.

Munger placement. Charlie Munger described four quadrants where a business can sit on the quality-versus-price grid.

• Wonderful business at fair price. The prize.
• Fair business at wonderful price. Works only with a deep discount and a sell-at-fair plan.
• Wonderful business at rich price. The discipline is to wait.
• Fair business at fair price. Avoid.

Closing quote. A short Buffett, Graham, or Munger quotation that reinforces the verdict. The quote frames the discipline.

Risk Management and Watchlist

In this section we list the five thesis-killers. Each risk uses two sub-labels.

Why fatal. What breaks the thesis if the risk fires. The level of specificity is operational: a number, a date, a competitor, a product. Not why the risk is unpleasant. Why it ends the case.

Monitor. Where to watch for the signal. A specific 10-Q line, a competitor's earnings call, a government register, a Bloomberg field. The Monitor label converts an abstract worry into an operational discipline.

Final Thesis

In this section we close the analysis. Three short paragraphs of prose.

Current quality. What the business is today. What is intact, what is impaired, what is recovering.

What to focus on next quarter. The binary read. A specific metric and a specific threshold that, when it reports, will confirm or invalidate the thesis.

Holding period and sell discipline. How long we plan to hold, and what would make us sell. The sell triggers are stated in advance, in writing. A price wobble alone is never a reason to act.

Closing sentence. The action at the current price and the entry or exit zones.

Website Distillation

Every note closes with ten labelled topics that compress the entire analysis into a short, scannable summary. Business, Competitive position, Moat, Management, Balance sheet, Earnings quality, Valuation, Margin of safety, Risk Watchlist, Verdict. Each topic is one short paragraph.

This is independent education and methodology, not investment advice. Compoundex is not licensed or regulated by the CNMV or any financial authority, and nothing here is a personalised recommendation to buy or sell any security. Always do your own analysis.