The borrower submitted a business plan projecting Rs 80 crore revenue by Year 3, 18% EBITDA margin, Rs 14.4 crore EBITDA and a DSCR of 1.9x. The lender sanctioned the facility on these projections. The projections were well-modelled, internally consistent and supported by a management presentation describing market opportunity, competitive positioning, product pipeline and growth initiatives. The credit committee was impressed.
Year 3 has arrived. Actual revenue: Rs 52 crore (35% below projection). Actual margin: 13% (5 percentage points below). Actual EBITDA: Rs 6.8 crore (53% below). Actual DSCR: 0.9x. The borrower cannot fully service the debt from operations.
The projection missed. The question the lender should ask is: could the miss have been predicted at the time of appraisal? The answer, in most cases, is yes, because the borrower’s own history of projecting, which was available at the time of appraisal but was not examined, would have revealed that this borrower consistently projects outcomes it cannot deliver.
Decomposing the projection into testable components
The aggregate projection (Rs 80 crore by Year 3) is not testable as a single number. It becomes testable when decomposed into its components: how many customers, at what average revenue per customer, with what retention rate, in which segments, at what prices, with what market share.
The borrower projected Rs 80 crore from: 120 customers at Rs 67 lakh average revenue, with 95% retention of existing customers and 25 new customers per year at Rs 40 lakh average initial revenue.
Each component is independently testable against the borrower’s own history:
Historical customer acquisition: 12 new customers per year (the projection assumed more than double the historical rate, with no documented explanation for why the acceleration would occur).
Historical retention: 82% (the projection assumed 95%, which the borrower had never achieved in any prior year, and which is in the top decile of the industry benchmark).
Historical average revenue growth per customer: 6% per year (the projection implied 15% growth to reach the Rs 67 lakh average from the current Rs 48 lakh, requiring each customer to nearly triple their spending over three years).
Every component of the projection exceeded the borrower’s demonstrated capability. The customer acquisition assumption required 108% acceleration. The retention assumption required 13 percentage points of improvement. The per-customer revenue assumption required 150% above the historical growth rate. No single component had been achieved before. The projection required all three simultaneously.
The credibility-adjusted projection
A credibility-adjusted projection, using the borrower’s demonstrated (not aspirational) performance on each component, produces: Year 3 revenue of Rs 50 crore to Rs 55 crore (consistent with the actual Rs 52 crore), EBITDA of Rs 6.5 crore to Rs 7.2 crore (consistent with the actual Rs 6.8 crore) and a DSCR of 0.85x to 0.95x (consistent with the actual 0.9x, which is below the facility’s covenant).
The credibility-adjusted projection would have shown the credit committee, at the time of appraisal, that the borrower’s demonstrated capability supported approximately Rs 52 crore of revenue and a DSCR below 1.0x. The facility would have been sized, priced and structured differently. The committee would have either sanctioned a smaller facility (sized for Rs 52 crore rather than Rs 80 crore) or required the borrower to explain, with documented evidence, what specific changes would cause the historical performance to accelerate.
Institutionalising the back-test
The projection back-test should be a standard appraisal step for every new borrower. The appraiser retrieves the borrower’s prior projections (from earlier lending relationships, from investor presentations, from prior CMA submissions, from internal business plans the borrower is willing to share) and compares them to actuals.
A borrower with a history of over-projecting by 30% should have its current projection discounted by 30%. A borrower with a history of delivering within 5% of projection should have its current projection accepted with higher confidence. A borrower with no prior track record of projecting should have its projection benchmarked against industry achievement rates for companies of similar size, sector and age.
The back-test does not predict the future. It calibrates the projection against the borrower’s own track record of projecting, which is the most reliable predictor of whether the current projection will be achieved. A borrower who has over-projected for three consecutive years is not a borrower who will suddenly begin hitting targets. The pattern is the evidence. The evidence should inform the sanction.
Northrop applies this test in credit appraisal reviews and independent loan monitoring. [email protected] | +91 92899 25657.