Deep research. Multidisciplinary thinking. Cash-flow valuation. Probabilistic judgement. Patience.
Company Anatomy does not begin with a target price.
It begins with the business.
The objective is to understand how a company makes money, what determines its long-term economics, how much cash it can reasonably generate for its owners, what could impair that cash generation, and how the value implied by those economics compares with the price available in the market.
Our approach can broadly be described as multidisciplinary, long-term value investing.
It draws from the principles of investors and thinkers including Warren Buffett, Charlie Munger, Philip Fisher, Benjamin Graham, Howard Marks, Ray Dalio and others—but the final analysis and judgement are our own.
There is no single formula.
Different businesses require different mental models.
1. Stay within the circle of competence
Research starts by asking whether we can understand the business well enough to form a defensible view.
We prefer businesses where we can reasonably understand:
- how revenue is generated;
- who the customer is and why they buy;
- the competitive structure of the industry;
- the economics of each incremental unit of growth;
- the capital required to operate and expand;
- the principal risks to the business; and
- the variables most likely to determine long-term value.
If the important variables cannot be understood with reasonable confidence, the appropriate conclusion may simply be “too difficult.”
Depth is preferred to breadth.
The objective is not to have an opinion on every listed company.
2. Understand the economic machine
Financial statements describe what happened.
Our research tries to understand why it happened and whether it can continue.
Each company is examined through a multidisciplinary latticework of mental models drawn from areas such as:
Accounting
Revenue quality, margins, working capital, cash conversion, capital intensity and balance-sheet strength.
Economics
Supply and demand, industry structure, economies of scale, pricing power and competitive dynamics.
Strategy
Moats, distribution, switching costs, customer behaviour, barriers to entry and reinvestment opportunities.
Psychology
Incentives, social proof, customer behaviour, management behaviour, overconfidence and other behavioural biases.
Probability
Ranges of outcomes, base rates, uncertainty and asymmetric risks.
Technology
Potential disruption, productivity improvements, changing cost structures and new competitive threats.
Operations and systems thinking
Bottlenecks, constraints, feedback loops, operating leverage and second-order effects.
The purpose of these models is not complexity for its own sake.
It is to avoid analysing a multidimensional business through a single financial lens.
3. Deep research before valuation
We prefer primary evidence wherever reasonably available.
Research may include:
- annual reports and financial statements;
- exchange filings;
- earnings-call transcripts;
- investor presentations;
- credit-rating reports;
- regulatory and government data;
- industry publications;
- competitor disclosures;
- historical company communications; and
- other credible publicly available sources.
Important claims are traced back to their source wherever practical.
The research report should make it possible for a reader to distinguish between what the evidence says and what the analyst believes it means.
4. Study management through history, not presentation
Management quality cannot be assessed from one investor presentation or conference call.
We therefore attempt to reconstruct the historical record.
Past annual reports, earnings calls, capital-allocation decisions, guidance and management commentary are reviewed to ask:
What did management say it would do?
What actually happened?
How did management behave when circumstances became difficult?
Was capital allocated rationally?
Were shareholders treated as partners?
Did explanations change after poor outcomes?
We pay particular attention to the relationship between management words, incentives and actions.
A long historical record can reveal patterns that a single year’s numbers cannot.
5. Use checklists to reduce avoidable mistakes
Research is supported by structured checklists influenced by the work of Philip Fisher, Buffett, Munger and other investors, together with lessons accumulated through practical experience in investing, marketing, finance and operating businesses.
The checklist may examine areas such as:
- customer proposition;
- market size and competitive structure;
- pricing power;
- unit economics;
- return on incremental capital;
- reinvestment runway;
- management incentives;
- capital allocation;
- accounting quality;
- balance-sheet resilience;
- disruption risk;
- customer concentration;
- regulatory risk;
- governance;
- competitive advantage; and
- potential sources of permanent capital impairment.
A checklist does not make the decision.
It reduces the probability that an important question is forgotten.
6. Follow the cash
Ultimately, the economic value of a business depends on the cash it can generate for its owners over time.
A central part of our valuation work is therefore estimating future free cash flow.
Rather than extrapolating reported earnings mechanically, we examine the underlying drivers:
Revenue → margins → taxes → reinvestment → working capital → capital expenditure → free cash flow.
We then ask:
How much cash might this business reasonably generate over its economic life?
and:
What price is the market asking us to pay for that stream of cash today?
Depending on the nature of the business, reports may use discounted cash-flow analysis, owner-earnings approaches, reverse valuation, comparable valuation or other appropriate methods.
No valuation model is treated as an oracle.
A model is only as good as the assumptions underneath it.
7. Think in distributions, not single-point forecasts
The future is uncertain.
We therefore prefer ranges of outcomes to false precision.
Where useful, research may employ:
- bull, base and bear scenarios;
- sensitivity analysis;
- reverse DCF;
- probability-weighted outcomes;
- Monte Carlo simulation; and
- other distribution-based techniques.
Monte Carlo analysis does not make an uncertain future certain.
Its purpose is to expose how valuation changes when several uncertain variables move simultaneously.
Instead of saying:
“The business is worth exactly ₹X.”
we are more interested in understanding:
What range of outcomes is plausible?
Which variables matter most?
What assumptions are already embedded in the current market price?
How much room is there for us to be wrong?
8. Invert the thesis
Before becoming comfortable with an investment thesis, we actively search for ways it could fail.
Every substantive report should ask:
- What is the strongest argument against our conclusion?
- Which assumption would destroy the valuation if it proved incorrect?
- What could permanently impair free cash flow?
- What could weaken the competitive advantage?
- Is management allocating capital poorly?
- Are we extrapolating unusually favourable conditions?
- What are we assuming because other investors assume it?
- What evidence would make us change our view?
This process of inversion is intended to combat confirmation bias and reduce the risk of becoming emotionally committed to a thesis.
The objective is not to prove that we are right.
It is to discover where we may be wrong.
9. Price, value and margin of safety
A great company is not automatically a great investment.
Price matters.
After estimating the economics of the business across a range of reasonable outcomes, we compare those estimates with the market valuation.
We look for situations where there appears to be a meaningful difference between price and conservatively estimated value, after accounting for uncertainty and downside risk.
The greater the uncertainty, the greater the margin of safety we would ordinarily require before becoming comfortable with a conclusion.
Valuation is therefore not merely about estimating upside.
It is equally about understanding the consequences of being wrong.
10. Judgement remains part of the process
Investing cannot be completely reduced to an algorithm.
Two analysts can examine the same financial statements and reach very different conclusions because the difficult questions often concern durability, competitive behaviour, management quality and future economics.
Our analysis therefore combines quantitative work with judgement developed through experience in investing, finance, marketing and business.
That judgement is explicitly treated as judgement—not disguised as fact.
Where possible, reports distinguish between:
FACT
Information supported by a cited source.
ESTIMATE
A calculation or forecast based on stated assumptions.
ANALYST JUDGEMENT
Our interpretation of what the available evidence means.
Separating these categories makes the reasoning easier to challenge and update.
11. Long-term thinking and patience
Our research horizon is generally measured in years rather than quarters.
Quarterly results matter when they reveal something about the long-term economics of a business.
They matter less when they merely create short-term volatility without changing intrinsic value.
The approach therefore emphasises:
deep understanding over activity;
patience over constant prediction;
business economics over market narratives;
and
decision quality over the appearance of precision.
A durable analytical advantage, if one exists, is more likely to come from understanding a smaller number of businesses unusually well than from reacting faster to every piece of market information.
There is no assurance that this approach will outperform the market.
12. Research is a living hypothesis
A published view is not a permanent commitment.
New evidence can change:
- the business economics;
- management assessment;
- competitive position;
- probability assumptions;
- free-cash-flow estimates;
- valuation; or
- the investment conclusion.
When material evidence changes, the thesis should change.
Intellectual consistency does not mean defending an old opinion.
It means applying the same reasoning process to new facts.
Our research principles
- Business-first
- Cash-flow focused
- Probabilistic
- Historically grounded
- Patient and long-term
- No assurance of returns
Use of artificial intelligence
AI and analytical software may be used to assist with activities such as organising information, analysing datasets, reviewing documents, modelling scenarios, coding, calculations and challenging hypotheses.
They do not replace the accountability of the Research Analyst.
Material sources, assumptions, calculations and conclusions remain subject to analyst review and judgement, and the extent of AI use is disclosed where applicable.
What this methodology does—and does not—do
The methodology is designed to improve the quality, transparency and robustness of investment analysis.
It does not eliminate uncertainty.
Research is based on publicly available information, known observations, assumptions and analysis available at the time of publication. Estimates and valuations may prove incorrect as circumstances change.
Investment recommendations, where made in individual research reports, are subject to the specific rationale, valuation, risk factors, disclosures and conflicts-of-interest disclosures contained in those reports.
Investments in securities are subject to market risk. Research recommendations do not provide any assurance or guarantee of returns.
Aditya Siddharth Mehta
SEBI Registered Research Analyst
Registration No. INH000029768