Building the Forecast Period
Learn how to construct a driver-based 5-year forecast — revenue, margins, reinvestment, and working capital — that produces the cash flows for your DCF.
Analyst Objective
Build a realistic, driver-based forecast that flows cleanly into unlevered free cash flow for your DCF.
What You'll Learn in This Lesson
1Introduction — What Is the Forecast Period?
The forecast period is the engine of the DCF — it produces the Unlevered Free Cash Flows (UFCF) used to calculate intrinsic valuation. Typically, investment bankers forecast **5 years**, although certain industries require **7–10 years**.
A strong forecast is not just a set of numbers — it is aforward-looking narrative about how a business grows, invests, and matures.Your job in IB or PE is to model a realistic growth and reinvestment trajectory that aligns with both historical performance and industry expectations.
DCF Forecast Timeline
Historical | Forecast Period (DCF) | Terminal Value --------------|------------------------------------|------------------- 2019 2020 2021| 2022 2023 2024 2025 2026 | TV (post-forecast)
Throughout this lesson, you will build the full 5-year projection block that powers Lessons 5 and 6. By the end, you’ll have the confidence to build driver-based forecasts exactly like a real analyst.
2Choosing the Forecast Horizon (5 Years vs 7–10 Years)
Most DCFs use a 5-year forecast. The logic is simple: beyond 5 years, assumptions become too speculative, so we hand off the valuation to the Terminal Value (Lesson 6).
However, some businesses require longer forecasts. This usually occurs when:
- Margins take many years to normalize (e.g., biotech, hardware manufacturing)
- Growth decays slowly (e.g., SaaS, subscription businesses)
- CapEx cycles are long (e.g., airlines, heavy industrials)
- The business is early-stage with negative UFCF that improves over time
Analyst Tip — How Bankers Justify Horizon Choice
Bankers always justify forecast duration by referencingindustry maturity, historic volatility, and management guidance. A 5-year forecast is only “incorrect” if the business model demands a longer path to stabilization.
Example: Choosing Between 5- and 7-Year Forecast
SaaS Business (high growth):• 5-year forecast insufficient because margin expansion lasts longer
• Typically use 7-year forecast
Industrial Business (stable, cyclical):
• 5-year forecast common • Terminal value captures remaining steady-state cash flows
In practice, analysts almost always choose a 5-year horizon unless the company explicitly requires a longer stabilization path.
3Revenue Forecasting — Growth, Volume, and Pricing
Revenue is the starting point of the forecast period. Every other line item — margins, expenses, CapEx, and working capital — ultimately depends on how you project the top line. A good revenue build is grounded in:
- Historical growth rates and volatility
- Unit / volume trends and pricing power
- Industry growth and competitive dynamics
- Management guidance and company strategy
Top-Down Approach
Start from market size and share. Useful in early-stage or new-market situations. Example: TAM × Market Share × Average Price.
Bottom-Up Approach
Build from units and pricing. Example: Subscribers × ARPU, Stores × Sales per Store, Units Sold × Price per Unit. Preferred in banking models.
Anchor on the Last Historical Year
Start from 2022A revenue — the base for your forecast.
Revenue Drivers
Every forecast starts from a **driver view** of the business — either top-down (market share) or bottom-up (units × price).
Tapering Growth Path
13% → 12% → 10% → 8% — higher in early years, then gently decelerating as the business matures.
Early forecast years carry higher growth; later years taper toward a steady-state range used for the terminal value.
Example — Tapering Growth (More Realistic)
Year 2022 2023E 2024E 2025E 2026E ----------------------------------------------------- Revenue (M) 400 452 507 557 601 Growth % — 13.0% 12.0% 10.0% 8.0%
Growth decelerates as the business scales. This pattern is common in SaaS, consumer, and many growth businesses.
In interviews, if asked to “project revenue,” never say you just hardcode growth. Mention at least one driver: units, stores, customers, pricing, or ARPU. That’s how actual bankers think.
4COGS & Gross Margin Forecast — Cost Structure of the Business
After revenue, the next key component is Cost of Goods Sold (COGS) and the resulting Gross Margin. Most investment banking models forecast COGS as a percentage of revenue, letting gross margin expand or contract over time based on the story.
Gross Margin Expands When:
- Fixed costs are leveraged over growing revenue
- Pricing power increases (strong brand, low competition)
- Product mix shifts toward higher-margin offerings
Gross Margin Compresses When:
- Input costs rise (commodities, freight, labor)
- Competition forces price cuts or promotions
- Sales mix shifts toward lower-margin products
Baseline: 60% COGS → 40% Gross Margin
Start from the historical cost structure in 2022A.
Gross Margin Story
Analysts usually forecast COGS as a % of revenue. Small changes in COGS % can create meaningful expansion or compression in gross margin.
- COGS % down → gross margin expands (good).
- COGS % up → gross margin compresses (bad).
- Always stress-test margins with cost-shock scenarios.
In the base case, COGS % drifts down and gross margin expands. In the compression case, a cost shock reverses some of that progress.
Example — Forecasting COGS and Gross Margin
Year 2022 2023E 2024E 2025E 2026E ----------------------------------------------------------- Revenue (M) 400 452 507 557 601 COGS % of Revenue 60.0% 59.0% 58.0% 57.5% 57.0% COGS (M) 240 267 294 320 342 Gross Profit (M) 160 185 213 237 259 Gross Margin % 40.0% 41.0% 42.0% 42.5% 43.0%
Note how gross margin gradually improves as the company scales. This pattern reflects operating leverage in production and supply chain.
In a modeling test, if you’re asked to “sensitize gross margin,” you’ll usually adjust COGS % of revenue by ±100–200 bps and observe the effect on EBITDA and UFCF. Being comfortable with this relationship is extremely important.
5Operating Expenses (SG&A, R&D, S&M) — Forecasting Opex Structure
Operating expenses are typically forecasted as apercentage of revenue, with the assumption that the company gains some degree of operating leverage as it scales. These costs include:
- SG&A — overhead, admin, sales support
- R&D — product development, engineering
- S&M — advertising, marketing, sales
SG&A
Usually semi-fixed. Declines as % of revenue over time due to scale efficiencies.
R&D
Important for software/tech/biotech. Often stable % of revenue or grows slowly.
S&M
Key driver in consumer & subscription. Often % of revenue declines as brand matures.
Starting Point: High Opex % of Revenue
In the base year, SG&A, R&D, and S&M consume a large share of revenue.
Opex Leverage Story
Because SG&A, R&D, and S&M are partly fixed, they usually grow **slower than revenue**. Their combined % of revenue drifts down over time — this is operating leverage.
- SG&A often declines the fastest (back-office scale).
- R&D tends to be flatter as % of revenue.
- S&M falls as the brand and customer base mature.
Total opex % of revenue trends down each year, even as revenue grows — this is why EBITDA margins expand in scalable business models.
Example — Forecasting Operating Expenses as % of Revenue
Year 2022 2023E 2024E 2025E 2026E ----------------------------------------------------------- Revenue (M) 400 452 507 557 601 SG&A % of Rev 22.0% 21.0% 20.0% 19.0% 18.0% SG&A (M) 88 95 101 106 108 R&D % of Rev 8.0% 8.0% 7.5% 7.5% 7.0% R&D (M) 32 36 38 42 42 S&M % of Rev 12.0% 11.0% 10.5% 10.0% 10.0% S&M (M) 48 50 53 56 60
Note how each cost line scales more slowly than revenue. This is **operating leverage**, which increases EBITDA margins over time.
In modeling tests, if they tell you “SG&A decreases by 100 bps each year,” that means:SG&A% next year = prior year's SG&A% − 1%.
6EBITDA → EBIT Forecast — Operating Profitability Over Time
Once you’ve forecasted revenue and operating expenses, you can calculateEBITDA and EBIT (Operating Income). These metrics reflect how profitable the core business is before interest and taxes.
EBITDA (Earnings Before Interest, Taxes, Depreciation & Amortization)
Reflects operating performance before non-cash items. Often used as a proxy for cash flow.
EBIT (Operating Income)
EBITDA minus depreciation & amortization. Used in DCF because it reflects true operating profitability.
Example — Forecasting EBITDA & EBIT Margins
Year 2023E 2024E 2025E 2026E --------------------------------------------------------- Revenue (M) 452 507 557 601 EBITDA Margin % 18.0% 19.0% 20.0% 21.0% EBITDA (M) 81 96 111 126 Depreciation (M) 24 26 28 30 EBIT (M) 57 70 83 96 EBIT Margin % 12.6% 13.8% 14.9% 16.0%
Notice how EBIT margins expand over time due to operating leverage and stable D&A. This is extremely common in scalable business models like SaaS, fintech, and mature consumer brands.
In practice, depreciation is either tied to CapEx or modeled as a percent of revenue. Both approaches are valid in DCF forecasting as long as consistency is maintained.
7Depreciation & Amortization Forecasting — Building Non-Cash Charges
Depreciation & Amortization (D&A) connects your forecasted income statement to both the cash flow statement and PP&E assumptions. DCFs require a reasonable estimate of future D&A because:
- D&A reduces EBIT (affects NOPAT)
- D&A is added back as a non-cash expense
- D&A often correlates with CapEx
- D&A forecasts influence long-term margin outlook
Method 1 — % of Revenue
Simple and commonly used. Works well for SaaS, services, and asset-light companies.
Method 2 — % of Prior PP&E
Better for asset-heavy businesses. Directly ties D&A to the capital base.
Method 3 — “Roll-Forward Light”
Approximate the PP&E roll-forward without building a full schedule. Common for DCFs that don’t require GAAP-perfect balancing.
Method 1: D&A as % of Revenue
Simplest approach — multiply forecast revenue by an assumed D&A % based on history.
D&A Modeling Approaches
In practice, you pick the method that best matches the business model andmateriality of D&A. Bankers care more about a sensible story than a perfect GAAP schedule.
In interviews, it's enough to say that you usually model D&A as a **% of revenue** or a **% of PP&E**, and that a full PP&E roll-forward is only needed for accounting-heavy tasks.
Example — D&A Forecast
Year 2023E 2024E 2025E 2026E -------------------------------------------------- Revenue (M) 452 507 557 601 D&A % of Revenue 5.3% 5.1% 5.0% 5.0% D&A (M) 24 26 28 30
D&A grows gradually with revenue. A stable or slightly declining % of revenue is very common.
In most DCF tests, you do NOT model a full PP&E schedule. A D&A % of revenue or a D&A % of last year’s PP&E is more than enough.
8CapEx Forecasting — Reinvestment Needs of the Business
Capital Expenditures (CapEx) represent the company’s investment in long-term assets — servers, equipment, manufacturing capacity, real estate, stores, etc.
CapEx is one of the **most important — and most subjective — assumptions** in a DCF forecast. The two primary forecasting methods are:
CapEx as % of Revenue
Simple. Works for asset-light companies, consumer brands, or SaaS.
CapEx as % of Prior PP&E
Better for asset-heavy companies: airlines, industrials, logistics.
Example — CapEx Forecast
Year 2023E 2024E 2025E 2026E -------------------------------------------------- Revenue (M) 452 507 557 601 CapEx % of Revenue 8.0% 7.8% 7.5% 7.5% CapEx (M) 36 40 42 45
CapEx trends more slowly than revenue, reflecting reinvestment needs but not scaling proportionally with growth.
Industrial and airline models often require CapEx for many years due to heavy reinvestment cycles. SaaS models often show the opposite pattern — CapEx intensity declines as the business scales.
9Working Capital Forecasting — AR, AP, Inventory, Accruals
Working capital determines how much cash is tied up in daily operations. Instead of guessing ΔNWC directly, analysts almost always forecastworking capital ratios:
- AR Days
- Inventory Days
- AP Days
- Accruals & other WC items
Working Capital Shock: AR Increases
AR ↑ 8 → customers take longer to pay, so cash goes down.
Rule of thumb: if more cash is tied up in working capital (ΔNWC > 0), UFCF goes down. If working capital is released (ΔNWC < 0), UFCF goes up.
Days Metrics (Analyst Formulas)
AR Days = (AR / Revenue) × 365Inventory Days = (Inventory / COGS) × 365
AP Days = (AP / COGS) × 365
Example — Converting Days to Dollars
Metric Value Calculation ----------------------------------------------- AR Days 45 days AR = (45/365) × Revenue Inventory Days 50 days Inventory = (50/365) × COGS AP Days 40 days AP = (40/365) × COGS
Example — ΔNWC From Forecasted Days
Year 2025E 2026E ---------------------------------- Revenue (M) 557 601 COGS (M) 320 342 AR Days 45 47 Inventory Days 50 52 AP Days 40 41 AR (M) 69.0 77.5 Inventory (M) 43.8 48.7 AP (M) 35.1 38.5 ΔAR +8.5 ΔInv +4.9 ΔAP +3.4 ΔNWC = −ΔAR − ΔInv + ΔAP = −8.5 −4.9 + 3.4 = −10.0
ΔNWC = +10 (cash outflow). This directly reduces UFCF in the DCF.
Working capital forecasting is often the part students skip, but it has a huge impact on UFCF. Mature businesses typically show stable days metrics. High-growth companies often experience worseningworking capital needs.
10Full 5-Year Driver-Based Forecast — Bringing Everything Together
Now we combine everything: revenue, COGS, operating expenses, D&A, CapEx, and working capital into a full 5-year forecast block. This is the exact modeling structure used in professional DCFs and IB case tests.
Year 1: Build FCFF from EBIT (2023E)
Start with EBIT, tax it to get NOPAT, then add D&A and subtract CapEx and ΔNWC.
2023E FCFF Build
EBIT = 57.0M
NOPAT = EBIT × (1 − 25%) =42.8M
+ D&A = 24.0M
− CapEx = 36.0M
− ΔNWC = 10.0M
FCFF = 20.8M
This is the **same FCFF logic** you used in earlier lessons — now applied across the full 5-year forecast block.
FCFF by Year (M)
Built from: EBIT, D&A, CapEx, ΔNWC
5-Year Forecast Table (Fully Driven)
Year 2022A 2023E 2024E 2025E 2026E ---------------------------------------------------------------------- Revenue (M) 400 452 507 557 601 Growth % — 13.0% 12.0% 10.0% 8.0% COGS % of Revenue 60.0% 59.0% 58.0% 57.5% 57.0% COGS (M) 240 267 294 320 342 Gross Profit (M) 160 185 213 237 259 Gross Margin % 40.0% 41.0% 42.0% 42.5% 43.0% SG&A % of Revenue 22.0% 21.0% 20.0% 19.0% 18.0% SG&A (M) 88 95 101 106 108 R&D % of Revenue 8.0% 8.0% 7.5% 7.5% 7.0% R&D (M) 32 36 38 42 42 S&M % of Revenue 12.0% 11.0% 10.5% 10.0% 10.0% S&M (M) 48 50 53 56 60 EBITDA (M) 81 96 111 126 126 EBITDA Margin % 20.3% 21.2% 21.9% 22.6% 21.0% D&A (M) 24 26 28 30 30 EBIT (M) 57 70 83 96 96 EBIT Margin % 14.3% 15.5% 16.4% 17.2% 16.0% CapEx (M) (36) (40) (42) (45) (45) ΔNWC (M) — (10) (12) (10) (8) Tax Rate % 25% 25% 25% 25% 25% ---------------------------------------------------------------------- UFCF Inputs Ready → go to Lesson 5: WACC & Discounting
This forecast block is the foundation of your DCF. In Lesson 5, we apply WACC to discount these cash flows.
11Realistic vs Unrealistic Forecasts — Analyst Judgment
Forecasting is part art, part science. It is easy to build unrealistic projections that show explosive growth and margin expansion. Bankers scrutinize forecast assumptions heavily.
Red flags that make a forecast unrealistic:
- Margins expand too quickly
- CapEx is unrealistically low relative to growth
- ΔNWC flips from positive to negative without explanation
- Revenue CAGR massively exceeds industry levels
- Operating expenses shrink too rapidly as % of revenue
Forecasts that bankers consider credible often show:
- Gradual, not explosive, margin expansion
- CapEx tied logically to revenue or PP&E growth
- Stable AR, AP, and inventory days
- Tapering growth rates
- Evidence of operating leverage
A good sanity check is: Does this forecast tell a believable story?Analysts often simplify that to: “Would a CFO sign off on this?”
12Interview Prep — Forecasting Questions & Answers
These forecasting questions appear frequently in investment banking interviews, modeling tests, and private equity screens.
2. Top-down vs bottom-up revenue forecasting?
3. What drives gross margin expansion?
4. Why do COGS % often decline as revenue grows?
5. What causes SG&A to shrink as % of revenue?
1. Walk me through how to project revenue.
6. When does SG&A rise as % of revenue?
7. Why do EBITDA margins expand over time?
8. How do you forecast depreciation?
9. Why might CapEx exceed D&A?
10. What are AR days?
12. How do you convert AR days to dollars?
13. What industries require longer forecast periods?
14. Why do growth rates taper?
15. What makes a forecast unrealistic?
11. How do AP days affect cash flow?
16. Why not forecast 20 years?
17. How do you forecast inventory?
18. What does negative ΔNWC imply?
19. How does revenue mix affect forecasting?
20. What does “operating leverage” mean?
If you can confidently explain forecasting drivers (revenue → margins → EBIT), you will outperform 90% of candidates in IB first-rounds and superdays.