A target company presents a five-year forecast projecting 20% revenue growth, 300 basis points of margin expansion and Rs 400 crore of cumulative free cash flow. The model is well-constructed, internally consistent and supported by a strategy presentation with market research, competitive analysis and named growth initiatives.
The question is not whether the forecast is ambitious. It is whether the forecast is credible. And credibility is determined not by the quality of the model but by the accuracy of management’s prior forecasts: when this management team has made similar projections in the past, how close did the projections come to actual results?
This is the back-test. It is the simplest analytical tool in due diligence. It is also the most revealing, because it answers a question that no amount of modelling sophistication can address: does this management team deliver what it promises?
The Methodology
Revenue back-test
For each of the last three to five years, obtain the revenue forecast that management presented at the beginning of the year (from the budget, the board pack, the investor presentation or the lender’s information memorandum). Compare to the actual revenue at year-end.
Year 1 forecast: Rs 350 crore. Actual: Rs 310 crore. Variance: -11%. Year 2 forecast: Rs 420 crore. Actual: Rs 365 crore. Variance: -13%. Year 3 forecast: Rs 500 crore. Actual: Rs 445 crore. Variance: -11%.
Pattern: systematic over-projection of approximately 12%. The management team consistently forecasts revenue it does not deliver. The over-projection is not random (which would produce both positive and negative variances). It is directional: always optimistic, never conservative.
The current forecast of Rs 600 crore should be read by the buyer as approximately Rs 528 crore (Rs 600 crore x 88%) after applying the historical accuracy discount.
Margin back-test
Management projects EBITDA margin expansion from 14% to 18% over five years. Has management ever projected margin expansion before? Did it materialise?
Year 1: Projected 15% EBITDA margin. Actual: 14.2%. Year 2: Projected 16% margin. Actual: 13.8%. Year 3: Projected 17% margin. Actual: 14.5%.
Pattern: management has projected margin expansion in every forecast and has never delivered it. The margin has fluctuated between 13.8% and 14.5% regardless of the projection. The current projection of 18% has no historical support.
Working capital back-test
Management projects working capital days declining from 85 to 70 over three years, releasing Rs 40 crore of cash. Historical data shows working capital days increasing from 75 to 85 over the prior three years. The projection contradicts the trend without identifying a specific, funded initiative that would reverse it.
Capex back-test
Management projects Rs 50 crore of annual capex. Actual capex over the last three years has averaged Rs 72 crore. The projection understates the capital the business historically requires by 30%. The understated capex produces an overstated free cash flow projection.
Cash flow back-test
The cumulative effect: revenue over-projection of 12%, margin expansion that has never materialised, working capital improvement that contradicts the trend and capex understatement of 30%. The free cash flow forecast, which depends on all four assumptions simultaneously, may overstate actual cash generation by 40% to 60%.
A buyer who pays for the forecasted cash flow pays for cash that the historical back-test says will not arrive.
Why Systematic Bias Matters More Than Individual Misses
A management team that forecasts revenue of Rs 400 crore and delivers Rs 380 crore has missed by 5%. That is a reasonable forecasting error. Markets are unpredictable. Customers are variable. Circumstances change.
A management team that has over-projected by 10% to 13% in every year for five consecutive years has not made forecasting errors. It has demonstrated a systematic bias. The bias is structural: it reflects the way the management team approaches projections, whether due to optimism, incentive alignment (bonuses tied to targets that are set from forecasts), growth narrative maintenance (investors expect to see ambitious projections) or an organisational culture that penalises conservative estimates.
The systematic bias will apply to the current forecast with the same probability as it applied to the prior ones. The back-test does not prove the current forecast is wrong. It establishes the calibration: how much should the buyer adjust management’s projections based on their track record?
The Credibility Assessment
In Northrop Management Private Limited’s due diligence practice, forecast credibility is assessed across four dimensions.
Dimension 1: Historical accuracy. The back-test. The single most objective measure of management’s forecasting reliability. A management team that has delivered within 5% of forecast for three consecutive years has earned credibility. A team that has over-projected by 12% has earned a discount.
Dimension 2: Assumption granularity. Are the projections decomposed into testable components? “Revenue will grow 20%” is an assertion. “We will add 15 new customers at Rs 2 crore average revenue, retain 90% of existing customers and achieve 3% price increases” is a set of hypotheses that can each be tested against evidence. Granularity improves credibility because each component can be independently verified.
Dimension 3: Sensitivity disclosure. Has management modelled what happens if growth is 5% lower, if margins compress, if working capital expands, if capex is higher? A forecast that includes only the base case tells the board what management expects. A forecast that includes downside scenarios tells the board what management is prepared for. The first demonstrates optimism. The second demonstrates rigour.
Dimension 4: Consistency with evidence. Are the assumptions consistent with market growth rates, competitive dynamics, the company’s own capacity, its customer pipeline and its historical capability? A company projecting 20% revenue growth in a market growing at 8% must demonstrate a specific, credible mechanism for capturing 12 percentage points of share gain. Without the mechanism, the projection is aspiration, not forecast.
The Credibility-Adjusted Forecast
The output of the Forecast-to-Actual Test is a credibility-adjusted forecast: management’s projections modified by the historical accuracy discount.
Management forecast: Rs 600 crore revenue, 18% EBITDA margin, Rs 108 crore EBITDA.
Historical accuracy discount: 12% revenue reduction, margin capped at demonstrated range (14.5%).
Credibility-adjusted forecast: Rs 528 crore revenue, 14.5% EBITDA margin, Rs 76.6 crore EBITDA.
At 10x EBITDA: management’s forecast supports a Rs 1,080 crore valuation. The credibility-adjusted forecast supports Rs 766 crore. The Rs 314 crore difference is the credibility gap: the amount the buyer pays for performance that management has historically not delivered.
The board should see both valuations. The negotiation should occur within the range between them.
Ashish Chaudhary, frames the back-test discipline directly: “Do not simply ask whether management’s current forecast looks reasonable. Ask: how accurate have similar forecasts been historically? A management team that has consistently over-projected by 12% is telling you to discount the current forecast by 12%. The back-test makes this discount explicit. The board deserves to see it before approving the price.”
Questions for the Boardroom
- For each of the last three to five years, how did the target’s actual results compare to the forecast management presented?
- Is the forecasting variance random (both positive and negative) or systematic (consistently in one direction)?
- If we apply the historical accuracy discount to management’s current forecast, what does the credibility-adjusted valuation look like?
- Has management ever delivered the margin expansion it is currently projecting, and if not, what evidence supports the expectation that it will this time?
- Does the investment committee memo present both management’s forecast and the credibility-adjusted forecast, or only the base case?
Practical Checklist
Request: Prior-year budgets, forecasts and investor/lender presentations for the last three to five years. Actual results for the same periods.
Calculate: Forecast-to-actual variance for revenue, margins, capex, working capital and cash flow, by year and in aggregate.
Identify: Systematic bias versus random error. Improving or deteriorating accuracy. Specific components where bias is largest.
Apply: Historical accuracy discount to the current forecast. Calculate the credibility-adjusted forecast.
Present: Both the management forecast and the credibility-adjusted forecast to the investment committee, with the credibility gap quantified in enterprise-value terms.
Closing Implication
A forecast is a promise. The back-test measures how often management keeps its promises. The board that sees only the promise pays for optimism. The board that sees the promise alongside the track record pays for evidence. The difference, in enterprise-value terms, is frequently the largest single adjustment in the entire diligence.