Choosing the Right Financial Model for the Decision You Face
The most expensive modelling mistake is not an arithmetic error. It is building the wrong model, correctly. A model built to persuade an investor and a model built to satisfy a lender answer different questions, are structured differently, and fail each other's tests. This guide is about choosing before you build.
- Match the model to the decision, not to the template you already have
- What makes a model decision grade
- The three questions a reviewer asks before reading a number
- Where a model stops and a signed professional opinion starts
A model is an argument, and the audience decides its shape
Everything in this guide follows from one idea: a model is built for a reader.
Before anything else, answer three questions. Who will rely on this? What will they do differently depending on the answer? And what would make them reject it? Those three answers determine the model's shape more than the industry does.
An investor is buying the upside and testing whether you understand your own business. The model needs unit economics, a credible growth mechanism and a clear picture of what the money buys. A lender is buying the downside and testing whether you can service debt in a bad year. The model needs a debt schedule, cover ratios and a stress case. A board approving a budget is testing accountability and needs the plan expressed in the same lines it will later be reported against. The same business, the same facts, three different models.
Which model, for which question
| The question | The model | What it turns on |
|---|---|---|
| Can we afford the plan we have written? | Three statement operating model | Working capital and the cash conversion cycle, not the profit line |
| What happens if the plan does not hold? | Scenario and sensitivity layer over the operating model | Which two or three assumptions actually move the answer |
| Can this project service its debt? | Project finance model | The debt schedule, the cover ratios and the construction period |
| What is this business plausibly worth? | Discounted cash flow, with a comparables cross-check | The terminal value and the discount rate, which usually carry most of the answer |
| Does this acquisition work? | Merger model, or a leveraged buyout model where debt drives the return | Synergies you can actually name, and the financing structure |
| How much do we raise, and what does it cost us? | Investor model with a cap table | Burn, runway and the dilution arithmetic across rounds |
A single business often needs more than one of these. What it does not need is one model asked to do all six jobs at once.
What decision grade actually means
Every assumption is visible and changeable
One assumptions sheet. No input typed inside a formula. If a reader cannot find the growth rate in under a minute and change it without breaking the model, the model is not finished.
The three statements tie
Balance sheet balances in every period. Cash flow reconciles to the movement in the cash line. Retained earnings roll forward correctly. A model that does not tie is not a model with a small error; it is a model whose output means nothing.
The regulatory inputs are dated
Every tax rate, threshold, statutory ratio and deadline in the model carries a note saying where it came from and when it was checked. This is what stops a model silently describing a rule that has been replaced.
It survives someone else opening it
The test is whether your own finance team can run the model in six months without calling the person who built it. Anything else is a dependency, not a deliverable.
The first ten minutes of a serious review
An experienced reviewer does not start with the answer. They start by breaking the model. Change revenue growth to zero and see whether the balance sheet still balances. Set the interest rate to something absurd and see whether the model errors or quietly returns a number. Trace one output back through the chain to the assumptions sheet and see how many steps it takes.
Then they look for the three failure signatures. Hardcoded numbers inside formulas, which mean an assumption is invisible. Circular references left running with iterative calculation switched on, which mean the answer depends on where the calculation started. And inconsistent row formulas across a time series, which is how a manual override from three months ago becomes permanent.
None of this is about arithmetic. It is about whether the model can be trusted by someone who did not build it, which is the only test that matters once the model leaves your desk.
Where the model shape follows the sector
| Sector | What drives the model | The line that usually breaks it |
|---|---|---|
| Real estate and property development | A project timeline with milestone based collections and construction outflow | Collections timing, which is a function of sales velocity and not of revenue recognition |
| Banking and financial services | The balance sheet is the business: asset growth, spread and provisioning | Credit cost, which is the assumption that moves the answer most and is estimated least |
| Healthcare and pharmaceuticals | Capacity, occupancy and payer mix, or for pharmaceuticals a product pipeline with probability weighting | Capital expenditure phasing against the point at which capacity earns |
| Manufacturing and supply chain | Capacity utilisation, contribution margin and working capital | Inventory and receivable days, which is where the cash actually goes |
| Technology and e-commerce | Cohort behaviour: acquisition cost, retention, contribution per customer | Retention, because a small change compounds over the forecast period |
The sector changes which lines matter. It does not change the requirement that the three statements tie.
Where a model stops
A model is an input to professional judgement. It is not a substitute for it.
There is a line between analysis and a signed professional opinion, and it is worth stating precisely because it is routinely blurred in marketing material.
Under the Companies (Registered Valuers and Valuation) Rules, 2017, a registered valuer's report must state the nature and sources of the information relied upon, and any caveats or limitations in it must not operate to limit the valuer's responsibility for the report. The Model Code of Conduct in Annexure I goes further: a valuer may not disclaim the duty of care, except to the extent that assumptions rest on statements of fact provided by the company, its auditors or its consultants, or on information in the public domain that the valuer did not generate. Rule 8(2) adds that where a valuer takes an input from another registered valuer, liability for the resulting valuation stays with the first valuer.
The practical consequence is straightforward. Management projections can be an input to a valuation, and the report must say so. But handing a professional a model does not transfer responsibility for the model's assumptions to them, and it does not convert the model into a report.
There is a standard for this, and it is worth knowing what it does and does not allow. A chartered accountant in practice may examine prospective financial information under SAE 3400, the standard on the examination of prospective financial information. The distinction it draws first is between a forecast, built on management's best-estimate assumptions, and a projection, built wholly or partly on hypothetical assumptions.
What the practitioner may then say is narrower than most people expect. The report gives negative assurance on the assumptions, ordinarily that nothing came to the practitioner's attention causing them to believe the assumptions do not provide a reasonable basis for the forecast. Alongside it sits a positive opinion on preparation and presentation: whether the information has been properly prepared on the disclosed assumptions and presented under the applicable reporting framework. For a projection the conclusion is framed against whether the hypothetical assumptions are consistent with its stated purpose, and distribution may have to be restricted.
What the practitioner may not say matters more. They do not certify that the forecast is accurate, do not state or imply that the projected results will be achieved, do not describe the engagement as an audit of future results, and do not assume responsibility for management's assumptions. The report must warn that actual results are likely to differ, possibly materially, because anticipated events frequently do not occur as expected. Management remains responsible for preparing the information and for identifying, supporting and disclosing the assumptions.
| Engagement | Governing standard | Assurance provided |
|---|---|---|
| Examination of a forecast or projection | SAE 3400 | Assurance, in the specific form described above |
| Agreed-upon procedures on model inputs or calculations | SRS 4400 | None. Factual findings only |
| Compilation of prospective information | SRS 4410 (Revised), where applicable | None |
| Valuation using forecast cash flows | The applicable ICAI Valuation Standard | A valuation conclusion, not assurance that the forecast results will occur |
| Preparation or review of a model as advisory work | Engagement terms and the ethical requirements | None, unless a separate assurance engagement is undertaken |
Naming the engagement correctly is the first control. Most disputes about what a model was supposed to be start here.
Our own modelling work is advisory. It is the last row of that table and we describe it as analysis, which is what it is. Where a client needs something a reader can rely on, that is a separate engagement under one of the rows above, and it is scoped and priced as one.
Where valuation sits, and why it is not on these pages
A model and a valuation report are different things, and confusing them is expensive. A model is an analytical tool you own and keep using. A valuation report is a signed deliverable, and Indian law is specific about who may sign one and when you must have one. We cover that separately and in depth on our valuation pages, starting with who can sign a valuation report in India. These pages stay with the model itself.
Questions about choosing a model
Can one model do everything?
One well built operating model can carry a scenario layer, a simple valuation output and a management reporting pack without strain. It cannot also be a project finance model, because the debt schedule drives everything in the latter and is a consequence in the former. Two models is usually the honest answer.
How detailed should the forecast be?
As detailed as the decision requires and no more. A five year monthly model with forty revenue lines is not more accurate than a five year quarterly model with four; it is only harder to check and slower to change. Detail should follow materiality, not ambition.
How far out should a model run?
Long enough to reach a steady state, which for most operating businesses is three to five years. Project finance models run the length of the debt, and sometimes the concession. Beyond that, a forecast is arithmetic dressed as foresight.
Where to go next
Financial Modeling Services
Back to the main page: what we build, how an engagement runs, and how to reach us.
Three Statement Models, Forecasting and Budgeting
The operating core: how the three statements tie, how to choose drivers that a business actually manages, what the accounting basis does to a forecast, and the integrity checks that catch most errors in under ten minutes.
Scenario, Sensitivity and Risk Analysis
Turning one number into a defensible range: scenario architecture, what sensitivity does and does not tell you, covenant and debt service headroom, and how to put a range in front of a board without it collapsing back into a point estimate.
Fundraising Models: Investor Financials, Cap Table and Dilution
The model as an investor reads it: burn and runway, the cap table, convertible instruments under Indian exchange control, employee option pools and the tax that attaches to them.
Project Finance and Infrastructure Models
A different model shape: bankability rather than profitability, the Reserve Bank's project finance regime as it stands since October 2025, debt sizing and cover ratios, and where viability gap funding fits.
Send an enquiry
Tell us the decision and the audience, and we will tell you which model it needs, including when the answer is that you do not need one.
Position as at 18 September 2026. Reviewed every six months.
This page is general information, not professional advice. A financial model is only as good as the rules it assumes will still be there when the forecast period arrives, and most of those rules moved recently. The Income-tax Act 2025 replaced the 1961 Act on 1 April 2026 and renumbered every section. The Reserve Bank's project finance regime was rewritten with effect from 1 October 2025. The external commercial borrowing framework was rewritten in February 2026. The treatment of a share buyback changed twice inside eighteen months. Take professional advice before acting on anything on this page. We are happy to be that adviser, but we do not act on a web page, ours or anyone else's, without one.