There's a specific moment in every investment banking interview and every client presentation where the DCF walk-through either builds or destroys credibility.
It's not the moment where someone challenges your WACC. It's not the terminal value question. It's the moment thirty seconds in, when the interviewer or the MD realises whether you're narrating a model or explaining an investment thesis.
Most candidates narrate the model. They start with revenue projections, explain each line item mechanically, discuss the discount rate calculation, add the terminal value, and arrive at a range. The person across the table nods politely, writes nothing down, and moves on.
The analysts who get offers and the practitioners who hold the room do something different. They tell the story the DCF is constructed to test. The model is evidence for a thesis, not the thesis itself.
Why Most DCF Walkthroughs Fail
The standard DCF walkthrough fails for a structural reason, not a knowledge reason.
Candidates who prepare extensively for DCF questions spend their preparation time on the mechanics: how to project revenue, how to calculate WACC, how to determine the terminal growth rate. They know the components. What they haven't prepared is the narrative architecture that makes those components cohere into an argument.
When you walk through a DCF component by component — revenue assumptions, EBITDA margins, capital expenditures, working capital, discount rate, terminal value — you're doing financial archaeology. You're showing someone the bones of the model.
What they're actually evaluating:
- Do you know which assumptions are load-bearing and which are secondary?
- Can you defend your most important assumptions with specific reasoning?
- Do you understand where the model is sensitive versus where it's stable?
- Can you explain why your output is useful despite its limitations?
The candidate who walks through a DCF in the correct order but treats every line item with equal emphasis has failed this evaluation even if every number is technically correct.
The three questions your DCF walkthrough must answer, in this sequence:
- What business story is this model testing?
- What are the two or three assumptions that actually drive the output?
- Where is the model most uncertain, and how should the audience hold that uncertainty?
Every well-constructed DCF walkthrough organises itself around these three questions. Everything else is detail.
Step 1: Lead With the Thesis, Not the Model
The most common opening to a DCF walkthrough is: "So I started by projecting revenue over a five-year period using a combination of top-down and bottom-up assumptions..."
This is the wrong opening. You've started inside the model. You haven't told anyone why this model exists or what business question it answers.
The correct opening tells the audience three things in under 90 seconds:
What company/situation this is: One sentence that establishes the business model and the relevant value driver. "TechCorp is a B2B SaaS company with 85% recurring revenue and has been growing at 35% annually for three years with gross margins expanding toward 72%."
What the DCF is testing: Every DCF tests a specific business hypothesis. Name it explicitly. "The DCF tests whether that growth rate moderates to 20% by Year 5 while gross margins expand to the 73-75% range that comparable scaled SaaS companies have achieved. The question is whether the current market price is consistent with that trajectory or whether it implies something more aggressive."
What the audience should listen for: Tell them which assumptions matter before you walk through them. "The model output is most sensitive to the Year 4-5 margin assumption and the revenue growth deceleration path. Those are the two places where reasonable analysts can disagree substantially."
Why this opening works:
It gives the audience a framework for listening. They know what the big questions are before you get into the details. When you reach the margin assumption, they already know it's load-bearing — so they pay attention. When you reach the working capital assumption, they already know it's secondary — so they file it away instead of interrupting to challenge it.
The opening also immediately signals analytical maturity. You're not reciting a model — you're running an analysis.
Step 2: Identify the Load-Bearing Assumptions Before Walking Through All of Them
A five-year DCF model typically has forty to sixty individual assumptions. Revenue growth rates, margin assumptions, working capital percentages, capital expenditure intensity, depreciation and amortisation, tax rates, beta, risk-free rate, equity risk premium, terminal growth rate, exit multiple.
Most of these assumptions have minimal impact on the output. Some of them — typically two to four — drive the vast majority of the valuation range.
The analyst who walks through all forty assumptions in sequence has made every assumption appear equally important. The practitioner who identifies the four load-bearing assumptions and allocates presentation depth accordingly has demonstrated something much more valuable: judgment about what matters.
How to identify load-bearing assumptions before the presentation:
Run a sensitivity table. Change each major assumption by its plausible range and observe the change in enterprise value. The assumptions that move the output by more than 10% with a plausible one-standard-deviation change are load-bearing.
In most DCFs, the load-bearing assumptions are:
Revenue CAGR (Years 1-5): In growth companies, small changes in revenue trajectory compound significantly. The difference between 22% and 28% growth over five years is not 6 percentage points — it's the difference between significantly different revenue bases in Year 5, which flows through to EBITDA, free cash flow, and terminal value simultaneously.
Terminal EBITDA margin: The terminal margin assumption determines both the terminal year cash flow and — if you're using an exit multiple — the multiple's application base. It's often the single highest-impact assumption in the model.
WACC: A 100 basis point change in WACC in a typical growth company DCF moves enterprise value by 15-25%. This is both impactful and contested.
Terminal value assumption: In most five-year DCFs, 60-70% of enterprise value is in the terminal value. The terminal value methodology dominates the output even though it's often discussed last.
The counterintuitive presentation order:
Most candidates walk through the model in the order it was built — revenue to EBITDA to cash flows to discount rate to terminal value to output. The correct presentation order is different: most impactful assumption first, least impactful last.
This matters because audiences lose attention over time. The critical assumptions — the ones that will be challenged, the ones that differentiate your analysis — deserve presentation when attention is highest.
Step 3: Defending Key Assumptions With Specific Evidence
This is where most candidates diverge from practitioners. When asked to defend a revenue growth assumption, the candidate says: "I assumed 22% growth because the company has been growing at 35% historically and I expect some moderation due to market saturation."
That is not a defence. That is a description of the assumption with a general rationale. A defence is specific.
What a specific defence looks like:
"I assumed 22% for Year 3 based on three data points. First, the five publicly traded direct peers with comparable ARR and market penetration showed a median deceleration from peak growth to 48-month growth of approximately 12-15 percentage points. TechCorp's deceleration from 35% baseline to 22% by Year 3 is in line with this peer experience.
Second, management's own guidance for the next fiscal year is 28%, which implies they expect some deceleration from the trailing rate. Extrapolating that deceleration trajectory puts Year 3 in the 22-24% range.
Third, the total addressable market they're penetrating is approximately $18B, and at $580M in current ARR they're at 3.2% penetration. At 22% growth, they reach 5.7% by Year 3 — well within the achievable range for a best-in-class SaaS company. I'd become more sceptical of growth persistence above 25% if penetration exceeds 8-10%."
The three sources of assumption support that practitioners use:
Historical peer precedent: What have comparable companies actually achieved at similar business stages? This is the most credible external reference. Peers who've done it already are better evidence than theoretical market models.
Management guidance: What is the company saying publicly? Management guidance is not a reliable predictor of exact outcomes, but it constrains the reasonable range. Assuming significantly above or below guidance requires explaining the divergence.
Structural logic: What market, competitive, or operational mechanism would cause this assumption to hold or not hold? For a margin assumption, the structural logic might be the unit economics of incremental customers, the leverage in the cost structure, or the comparison to fully-scaled peers.
The candidate who says "27% because peers are around that level" has provided peer precedent only. The practitioner who says "27% because peers at comparable scale average 27%, management has guided to 25-28%, and the cost leverage in R&D and S&M supports this range" has provided all three tiers — and that's what earns credibility.
Step 4: Navigating the WACC Discussion Without Losing Credibility
WACC is the section where technically prepared candidates most commonly lose credibility. Not because they don't know how to calculate it — because they present it as if it's more precise than it is, and experienced interviewers and practitioners immediately see through false precision.
The WACC credibility trap:
"My WACC is 9.23%. I used a beta of 1.34 based on a two-year weekly regression of the company against the S&P 500, a risk-free rate of 4.35% (current 10-year Treasury), an equity risk premium of 5.5%, a pre-tax cost of debt of 6.2%, a 25% tax rate, and a 70/30 equity/debt capital structure consistent with the company's current mix."
This presentation has two problems:
Problem 1: False precision. The WACC to two decimal places implies a precision the underlying inputs don't support. Beta estimation varies by 20-30% depending on the lookback period, frequency, and comparison index. The equity risk premium is a long-running academic and practitioner debate — reasonable estimates range from 4.5% to 6.5%. A 9.23% WACC expressed to two decimal places is not more accurate than a 9% WACC — it's less honest about its own uncertainty.
Problem 2: No acknowledgment of the range. WACC is a range, not a point estimate. Presenting it as a single number and defending that number as if it's precise is not intellectually honest.
The correct WACC presentation:
"My base case WACC is approximately 9%, and I've tested the model in the 8-11% range. The two inputs that drive most of the uncertainty are the equity risk premium — which I've set at 5.5% based on Damodaran's India market estimate but which reasonable analysts put anywhere from 4.5% to 6.5% — and the beta, where the two-year regression gives 1.3 but the five-year regression gives 0.9. I've used 1.1 as a blended estimate and I can walk through why if useful."
This presentation does three things the false precision approach does not: it acknowledges uncertainty honestly, it identifies the specific drivers of that uncertainty, and it demonstrates that the analyst has already thought about the sensitivity — rather than waiting to be challenged.
The WACC question that separates candidates:
"What happens to your WACC if the company raises additional debt to fund a buyback?"
The correct answer: The pre-tax cost of debt is typically lower than the cost of equity (reflecting debt's seniority in the capital structure). Taking on more debt — if the company can service it — lowers the blended WACC because you're replacing expensive equity with cheaper debt. However, as leverage increases, the cost of both debt and equity rises (financial distress risk), and at some leverage level the WACC starts increasing again. The optimal capital structure exists somewhere in the middle — and finding it is one of the analytically challenging aspects of capital structure advisory.
Step 5: Presenting the Output as a Range, Not a Point Estimate
The single most common DCF presentation error, made by experienced analysts and junior analysts alike: presenting the DCF output as a single number.
"The DCF value is ₹2,400 Cr."
This is wrong in a way that's worse than a numerical error. It's wrong in its presentation of certainty. The DCF output is not ₹2,400 Cr. It is "₹2,000-2,800 Cr, with the range driven primarily by the terminal margin assumption (23-29%) and WACC (8.5-10.5%)."
The sensitivity table is not optional:
Every DCF walkthrough should conclude with a sensitivity table showing the output across the plausible range of the two most impactful assumptions. The table does two things:
First, it shows where the model is stable and where it moves — which is more information than any point estimate.
Second, it demonstrates that you understand your model. An analyst who can present a sensitivity table without being asked has signalled that they understand DCF as a probabilistic framework, not a calculator.
What the table should look like:
| WACC ↓ / Terminal Margin → | 23% | 26% | 29% |
|---|---|---|---|
| 8.5% | ₹2,100 | ₹2,500 | ₹2,900 |
| 9.5% | ₹1,850 | ₹2,200 | ₹2,550 |
| 10.5% | ₹1,650 | ₹1,950 | ₹2,250 |
This table shows the full range (₹1,650-2,900 Cr), which is significantly more honest than any single number in the middle. It also immediately reveals which assumption has more impact — in this case, terminal margin moves the output more than WACC across the plausible ranges.
The Limitation Acknowledgment: What Builds Trust
The final element of a strong DCF walkthrough is what most analysts skip because it feels like admitting weakness: explicitly acknowledging what the model doesn't capture.
In practice, this acknowledgment builds rather than destroys credibility — because experienced practitioners know every model has limitations. An analyst who names them demonstrates awareness. An analyst who presents a model as comprehensive invites the MD to find the limitations and raise them first.
The limitations worth acknowledging:
- "This DCF doesn't incorporate the potential impact of the regulatory review pending with CCI — if it requires a business unit divestiture, the terminal value assumptions change materially."
- "The working capital assumptions are based on the company's historical pattern, but if they enter a rapid international expansion phase the absolute working capital drag will be higher than my model shows."
- "My terminal value uses a 4% perpetuity growth rate, which is slightly above India's long-run GDP growth assumption. I'm embedding a modest productivity premium, and I'd revise that assumption if you have a different view on their long-run competitive position."
Each of these statements demonstrates that the analyst has thought past the model output to the real-world conditions that would make it wrong. That's the mark of an analyst who can be trusted to present analysis to a client — because they won't be surprised by a question the model doesn't answer.
Closing: From Model Narrator to Analytical Practitioner
The gap between narrating a DCF and presenting one that holds the room is not a technical gap. It's a presentation and judgment gap that stems from misunderstanding what the DCF walkthrough is actually testing.
Thesis-led opening. Load-bearing assumption identification. Specific evidence-based defence. Honest WACC range. Sensitivity table as the output, not a point estimate. Explicit limitation acknowledgment.
This structure takes longer to prepare than a mechanical walkthrough. It takes the same or less time to deliver. And it produces a fundamentally different impression on the person across the table — the impression of an analyst who understands what the model is actually doing, not just one who built it.
At Meritshot, the Investment Banking programme trains DCF walkthrough skills through practitioner review — students present analyses to experienced bankers who have reviewed hundreds of pitch books and DCFs in live deal contexts. The feedback is the same feedback you would receive from a VP at 10 PM: not whether the numbers are right, but whether the argument is coherent, the load-bearing assumptions are identified, and the uncertainty is presented honestly. That's the standard that matters.
Explore the Meritshot Investment Banking Programme →
This article was written by the Meritshot content team. Meritshot trains professionals in Data Science, AI Engineering, Full Stack Development, Investment Banking, and Cyber Security through hands-on, practitioner-led programmes.
