The expected value of an ODD review per manager, its break-even points, and why depth should scale with the ticket.
Operational due diligence is worth its cost when the loss it is expected to prevent exceeds what the review and the monitoring cost over the fund's life. On illustrative assumptions, a $50.0m commitment, a 3.0 per cent chance of a material operational failure, a 20 per cent loss when it happens and a review that catches 60 per cent of them, ODD avoids an expected $180,000 for a cost of $46,000: 3.9 times its cost. The same review on a $10.0m ticket returns 0.8 times, which is an argument for a lighter, proportionate review, not for none.
Worked in full in Operational Due Diligence in Private Equity by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The question is posed per manager, over the life of one fund commitment. Costs are internal time at a fully loaded rate plus out-of-pocket items. The loss side uses three inputs every ODD head has an opinion on and few write down: how often a manager suffers a material operational failure (misappropriation, valuation misstatement, a failed administrator, a sanctions or fraud event), how much of the commitment such a failure costs, and how many of those failures a proper review would have caught or prevented. All three are illustrative; the point is to make them explicit.
| Input | Value |
|---|---|
| Commitment to the fund | $50.0m |
| Initial review: 80 hours at $150 an hour | $12,000 |
| On-site travel | $6,000 |
| Background, litigation and regulatory checks | $4,000 |
| Monitoring: 16 hours a year for 10 years | $24,000 |
| Total cost over the fund's life, undiscounted | $46,000 |
| Incidence of a material operational failure | 3.0% |
| Loss as a share of commitment when it happens | 20% |
| Share of failures ODD detects or prevents | 60% |
Expected loss = commitment × incidence × loss given failure
= 50.0m × 3.0% × 20% = $300,000
A 3.0 per cent incidence is one manager in 33 over a fund's life. The expected figure hides the shape: in 32 cases nothing happens, and in one the investor loses $10,000,000.
Avoided loss = expected loss × detection = 300,000 × 60% = $180,000
Net benefit = 180,000 − 46,000 = $134,000; benefit / cost = 3.9
Excel: =B1*B2*B3*B4/B5 with commitment, incidence, loss, detection and total cost in B1:B5
In basis points of commitment, the programme costs 9.2 and is expected to save 36.0. For a team running 12 new managers a year, that is $552,000 of cost against $2,160,000 of expected losses avoided.
Each input can be solved for the value at which the review just pays for itself, holding the others. These are the figures to put in front of a committee that thinks ODD is a cost centre.
On a $50.0m ticket the review pays for itself even if it catches only about one failure in six. The variable that most often makes it uneconomic is the size of the commitment.
| Commitment | Incidence 1% | Incidence 3% | Incidence 5% |
|---|---|---|---|
| $10.0m | 12,000 (0.3x) | 36,000 (0.8x) | 60,000 (1.3x) |
| $25.0m | 30,000 (0.7x) | 90,000 (2.0x) | 150,000 (3.3x) |
| $50.0m | 60,000 (1.3x) | 180,000 (3.9x) | 300,000 (6.5x) |
Where the multiple is below 1.0x, small tickets at low incidence, the full review is uneconomic on expected value. That is where proportionality comes in. A desk-based review of 40 hours with the same checks, no travel and monitoring every other year costs $22,000 over the life of the fund. If it catches 45 per cent of failures rather than 60, on a $10.0m ticket at 3 per cent it avoids $27,000: 1.2 times its cost, against 0.8 for the full review. Depth should scale with the ticket and with the risk rating, which is what tiering does.
Expected value is the right test for the budget, not for the decision on a single manager. An emerging manager with no independent administrator can carry an incidence several times the average, and one veto finding is enough to decline whatever the arithmetic of the average says.
The common mistake is to compare the cost of the review with the expected loss, $46,000 against $300,000, and declare the review worth 6.5 times its cost. The review does not remove the loss; it removes the share it catches. At 60 per cent detection the ratio is 3.9, and an untested review, one that reads policies without checking that the controls operate, may detect far less. The point is made in money in what an untested callback control costs: a control on paper protects a fraction of what a tested one does. The opposite mistake is to drop ODD on small tickets because the full review fails the test, when a lighter review passes it.
The rating discipline behind detection, and why averaging scores hides the findings that matter, is worked on the companion page and its free workbook.
Count the whole life, not just the initial review. In this illustrative case the initial review costs $22,000 (80 hours at $150, travel and background checks) and monitoring adds $24,000 over a ten-year fund: $46,000 per manager, or 9.2 basis points of a $50.0m commitment.
Solve the expected value for each input. With $46,000 of cost, 20 per cent loss given failure and 60 per cent detection, a $50.0m commitment breaks even at an incidence of 0.77 per cent, about one manager in 130; at 3.0 per cent incidence the break-even ticket is $12.8m. Below that, use a lighter, proportionate review.
No, they should get a proportionate review. On a $10.0m ticket a full review avoids an expected $36,000 against a $46,000 cost, 0.8 times. A desk-based review costing $22,000 that catches 45 per cent of failures avoids $27,000, 1.2 times its cost, and still applies the veto findings that should stop a commitment outright.
This article is one calculation from Operational Due Diligence in Private Equity. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: private equity and private markets → · All 1340 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.