Financial AnalystInterview Answers

A DCF values a business by projecting free cash flows over a forecast period (typically 5–10 years) and discounting them back to present value using the Weighted Average Cost of Capital (WACC). WACC blends the cost of equity (using CAPM: risk-free rate + beta × equity risk premium) and after-tax cost of debt, weighted by capital structure. I also calculate a terminal value — usually using a Gordon Growth Model — which represents cash flows beyond the forecast period. The sum of discounted FCFs plus terminal value gives enterprise value. The most sensitive assumptions are the WACC and terminal growth rate.

I start with the income statement — projecting revenue, COGS, and operating expenses to get to EBIT and net profit. The balance sheet and cash flow statement are then linked to the P&L through working capital assumptions (DSO, DPO, DIO), capex schedules, and debt movements. The cash flow statement reconciles beginning and ending cash using operating, investing, and financing activities. All three statements must balance and reconcile — I use checks throughout the model to catch errors. Clean structure, clear assumptions, and audit trails are essential.

I work bottom-up for revenue — by product, channel, or customer segment — then overlay top-down strategic assumptions. For costs, I separate fixed and variable components and model them to the revenue drivers. I build flexibility into the model with scenario analysis (base, upside, downside). I challenge assumptions with historical trends and external benchmarks. I also ensure forecasts are realistic and include the finance team, commercial leads, and operations in the process — a forecast built in isolation will be ignored.

I look at revenue growth and quality of revenue, gross margin and margin trends, EBITDA and operating leverage, free cash flow conversion, working capital efficiency (DSO, DPO, inventory turns), return on invested capital (ROIC), and debt metrics (leverage ratio, interest coverage). For SaaS or subscription businesses I focus on ARR, churn, NRR, CAC, and LTV. The metrics I prioritise depend on the business model and the decision at hand.

I modelled three acquisition scenarios with different assumptions about integration costs, synergy timing, and exit multiples. My analysis showed that the acquisition the board was favouring looked attractive on an EV/EBITDA basis but had a negative NPV under realistic integration assumptions. I presented sensitivity analysis that highlighted the key value drivers. The board paused the deal, renegotiated the price, and achieved a significantly better outcome. Financial analysis should be the foundation for major capital decisions.

I lead with the conclusion and recommendation, then support with evidence. I translate financial language into business terms — instead of "EBITDA margin compression," I say "for every dollar of revenue we're retaining fewer cents in operating profit." I use clear visuals — waterfall charts for variance analysis, scenario tables for sensitivity — and keep technical detail in appendices. I also prepare for questions and know my numbers well enough to handle challenges in the room.

I've built models extensively in Excel — advanced functions, dynamic ranges, VBA where appropriate. I've also used Anaplan, Adaptive Planning, and Power BI for integrated planning and reporting. I can extract and manipulate data from ERP systems including SAP and Oracle. I prefer tools that create a single source of truth and reduce manual reconciliation — but I always ensure I understand the underlying logic, regardless of the tool.

I'm transparent about data limitations and clearly flag assumptions I've made. I use sensitivity analysis to show how outputs change under different assumptions, so decision-makers understand the range of outcomes, not just a single number. I prioritise finding the most reliable proxies when direct data isn't available. I also establish a clear view on what additional data would most reduce uncertainty and recommend how to obtain it. Honest analysis with stated uncertainty is more useful than false precision.