On this page
What is an earn-out?
An earn-out is the part of an acquisition price that the buyer pays after closing, and only if the acquired business reaches performance targets written into the purchase agreement. Kohers and Ang (2000)5 describe the deal as a two-part payment: “a front-end payment at the time of the merger, and a deferred payment, or earnout, which is contingent on the target’s ability to achieve certain measurable prespecified performance standards”. In the US contracts Cain et al. (2011)2 read, the deferred part was most often tied to the target’s profit or sales, measured once a year over one to three years, and paid in cash, shares or both.
When does it help, and when does it hurt? The studies below find earn-outs used most where the seller knows more about the business than the buyer can check, where the value rests on owner-managers, and where the target can still be measured on its own after closing, and acquirers in such deals earned more at announcement. None of them measures whether an earn-out repays its costs, or whether it keeps anyone. This essay’s reading: it is most likely to go wrong when the buyer plans to merge the target so fully that its results can no longer be told apart, because the payment formula then becomes a quarrel about whose costs and whose customers count.
In this essay, an earn-out means a contingent part of the purchase price whose amount depends on the acquired business’s own performance after closing. The systematic review by Dahlen (2024)3 separates it from two neighbours. A contingent value right, in the form the review describes, protects target shareholders against a fall in the acquirer’s share price, so its value moves with the market as well as with the business. A performance commitment, prevalent in China, settles the whole price up front and has the seller pay compensation if the targets are missed. A third instrument, not in the review’s list, is not an earn-out either: a closing-accounts adjustment corrects the price to the working capital or net debt on the closing date, so it settles a past balance, not future performance.
Figure 1An earn-out in sequence, from signing to the last payment
Five stages and an end condition. Shares are from US contracts from 1994 to 2003; the arbiter is from one published contract.
Source: Kohers and Ang (2000); Cain et al. (2011); Datar et al. (2001); layout the author's.
How the promise is booked is a separate question, and this essay does not answer it. Cain et al.2 (2011) note that since SFAS 141(R), effective in fiscal 2009, a US acquirer records an earn-out at fair value as part of the purchase price at the acquisition date. Which later changes reach the income statement, how IFRS 3 treats the same promise, and whether a payment tied to the seller staying counts as price or as pay are questions for an accounting guide.
Why would a buyer pay part of the price later?
Kohers and Ang (2000)5 name two problems in their title: agreeing to disagree and agreeing to stay. The first is a valuation gap. In a single cash payment, “The bidder, being the less-informed party, bears the risk of paying too much”, while a seller who is right about the business may never receive what it is worth. An earn-out lets both sides sign on the agreed part and leave the disputed part to the results. It also sorts sellers, but only partly: “only targets that are truly superior or those that are overly optimistic will have the confidence to self-select”, and “Time will distinguish between these two types of targets”.
The second problem is that the value may leave with the people who run the business. Linking the deferred payment to the owner-manager’s continued employment, Kohers and Ang5 argue, “serves as a bonding mechanism to induce the manager to stay”. Datar et al. (2001)4 frame the same pair as adverse selection, where the seller has private information about the business, and agency costs, where its value depends on the seller’s effort after the sale. Their survey of acquirers in 117 earn-out deals from 1996 and 1997 gives the motives in practice: of the 27 who could and would name a motive, 14 cited valuation differences, 7 keeping the owner, and 6 both.
The acquirers who cited valuation said the targets had “no track record,” “little tangible assets,” or a value that depended heavily on future growth; the earn-out gave the entrepreneur an opportunity to “put his money where his mouth was.” The same authors are plain about the price. The seller now carries business risk it wanted to sell, including “the uncertainty associated with the competence of the acquiring company in managing the operations of the target”. On their reading, in many cases the value an earn-out creates will not be enough to pay the seller for bearing that risk.
Which deals carry earn-outs?
The large samples agree on the kind of deal more than on how common earn-outs are. Datar et al.4 (2001) found earn-outs in 4.1% of 39,706 completed acquisitions in the disclosed-value file of Securities Data Company (SDC) from 1990 to 1997, with targets in 157 countries: about 7% of acquisitions of private companies and under 1% of public ones. They expected the true share to be “far higher”, because deals with undisclosed terms drop out of the database. Cain et al. (2011)2 counted 990 earn-outs among 25,213 acquisitions by publicly traded US corporations from 1994 to 2003, or 3.9%, rising from 3.1% in 1994 to 6.8% in 2003.
The UK figure is of another order. Barbopoulos and Sudarsanam (2012)1 studied 4,788 bids by acquirers listed on the London Stock Exchange from 1986 to 2008 and found an earn-out in 1,251, or 26.1%; 1,045 of those targets were private. They point to the much larger share of private targets in UK deals as one reason. The shares come from different databases, filters and countries, so they are not a trend line; what holds across all four samples is where earn-outs appear.
They appear where the target is hard to value or hard to keep. Fewer than 2% of the targets in Cain’s2 sample were publicly traded; 74% were private companies and 23% subsidiaries of public firms. Kohers and Ang5 modelled the choice in private and subsidiary deals. The odds of an earn-out were almost 70% higher for high-tech targets and about 86% higher for service firms. They were about 72% higher for private targets than for subsidiaries, and about 30% lower when buyer and target shared an industry. Smaller acquirers, and deals large relative to the acquirer, used them more. Datar’s4 team adds a contrast: shares follow a different logic, because “financing considerations are a more important factor in the use of stock”.
| Study | Deals, country, period | Earn-out share | Where earn-outs were used | What acquirers earned | What it cannot show |
|---|---|---|---|---|---|
| Kohers and Ang (2000) | 938 completed earn-out deals in SDC, US targets (about 18% with foreign acquirers), 1984 to mid-1996 | Not reported by the authors; Cain et al. put it at 5.6% | Private, high-tech and service targets, cross-industry deals, smaller acquirers | Private targets, announcement day: 1.348% with an earn-out against 0.867% for cash and 0.868% for stock (difference significant at 10% and 5%); no abnormal performance over 3 years | Payout and retention come from 110 deals, with no comparison group for retention |
| Datar et al. (2001) | 39,706 completed deals in SDC’s disclosed-value file, targets in 157 countries, 1990 to 1997 | 4.1%; about 7% of private and under 1% of public targets | Private and subsidiary targets, cross-industry deals, services, industries with few deals | Not studied | The authors expect the true share to be higher; survey answers from 27 acquirers |
| Cain et al. (2011) | 990 earn-out deals by US public acquirers, 1994 to 2003 | 3.9%, from 3.1% in 1994 to 6.8% in 2003 | 74% private targets, 23% subsidiaries, under 2% public | Not studied | Contract terms read for 447 to 535 deals; expected payout simulated for 127 contracts with prior-year data |
| Barbopoulos and Sudarsanam (2012) | 4,788 bids by UK listed acquirers, 1986 to 2008 | 26.1% (1,251 bids) | Private and subsidiary targets, diversifying deals, intangible-rich industries | Five-day return 1.48% against 1.07% without an earn-out; for private targets 1.49% against 1.22%, not significant; no long-run outperformance in calendar-time portfolios | Listed acquirers only; the “optimal” use behind its long-run gains is classified in-sample |
| Dahlen (2024), review | 64 papers, 1970 to 2023 | Not a sample | Listing status and target industry are the robust determinants | Reads the studies as positive at announcement and in the long run; research skewed towards short-term returns | Costs, dispute rates and actual payouts are largely unmeasured |
Figure 2Earn-out use and acquirer returns: four samples, one pattern
Samples from 1984 to 2008, mostly US and UK. Earn-out shares are not comparable across rows, because the databases, filters and countries differ.
Source: Kohers and Ang (2000); Datar et al. (2001); Cain et al. (2011); Barbopoulos and Sudarsanam (2012); Dahlen (2024).
What does an earn-out contract measure, and for how long?
Cain et al. (2011)2 read the contract terms in SEC filings. The maximum earn-out averaged $21 million, with a median of $5 million, and if paid in full it would average 33% of the transaction value. In 90% of the contracts the payment depended on the target’s performance alone. Accounting profit was the measure in 52% of the cases where it could be identified, sales in 32% and non-financial milestones, such as clinical trials, FDA approval or specific customer contracts, in 12.2%. The period averaged 2.57 years, and 77% measured performance annually. Most payments rose in a straight line or in steps up to a cap.
The headline is a ceiling, not a forecast. Among targets with prior-year figures, the median had to improve its performance by 18% before any payment was due, while over a quarter needed no improvement at all. For the 127 contracts with prior-year figures, a Monte Carlo simulation that centres each target’s future performance on its prior-year level put the expected payment at 13.8% of transaction value on average; the figure SDC records for all 990 deals averaged about 33%. About 19% of the simulated targets had an expected payoff of zero. The authors conclude that the SDC figures “appear to represent an upper bound, rather than an expectation of the potential earnout payoffs in most acquisitions”. Kohers and Ang (2000)5 found the same gap in payments actually made, which the section on returns reports.
Figure 3Earn-out headline against expected payment, as a share of transaction value
Different samples: the maximum covers 447 deals, the simulation 127. The simulation assumes no growth.
Source: Cain et al. (2011), Tables 2 and 6.
The sample contract printed by Datar et al. (2001)4 shows what each clause decides. The buyer paid $957,000 in cash and $750,000 in a promissory note at closing, and up to $1,350,000 more over six years: sixty percent of the pretax income above a yearly minimum, capped each year. The company’s accountants determine pretax income under generally accepted accounting principles, net of a charge for interest on loans from the buyer and “a reasonable charge representing direct costs incurred by the Purchaser on behalf of and for the benefit of the Company”. The buyer may cancel the payments if the seller materially breaches named terms of his employment agreement, is convicted of a felony, is debarred from the company’s US government contracts or is dismissed for wilful refusal to perform. A dispute over pretax income goes to arbitration by three certified public accountants.
Each of those clauses answers a question that otherwise surfaces after closing: what counts as profit, which of the buyer’s costs the target must carry, what happens if the manager breaches the employment agreement or is dismissed for cause, and who settles a disagreement. Kohers and Ang (2000)5 found the measures in their contracts to be almost always profit-related or revenue growth, and only once an outside factor, so earn-outs are “not commonly used for hedging exogenous risk factors”. They insure the buyer against the target, not against the market.
Do acquirers gain from paying this way?
At the announcement, on these samples, yes. Kohers and Ang (2000)5 measured an average abnormal return of 1.356% on the announcement day for 449 acquirers using earn-outs. For private targets the day’s return was 1.348% with an earn-out, against 0.867% for cash and 0.868% for stock, a difference significant at the 10% level against cash (p = .0727) and the 5% level against stock (p = .0493). Where information asymmetry was high, the gap was wider over two days: in high-tech private targets the two-day return was 2.56% with an earn-out and 1.01% without, in service firms 2.85% against 0.89%. For subsidiaries bought from a parent there was no difference. A 3-year calendar-time test found no abnormal performance afterwards, so the gain was not reversed.
The UK evidence is weaker in the places that matter most. Barbopoulos and Sudarsanam (2012)1 found a five-day return of 1.48% for earn-out bids against 1.07% for the rest, a significant difference. For private targets, where the theory predicts the largest benefit, it was 1.49% against 1.22% and the difference was not significant; at the median, earn-out bids beat cash bids, 0.87% against 0.40%, and trailed mixed offers, 0.87% against 1.01%. In cross-border deals the earn-out added nothing. The authors, and Dahlen’s3 review after them, read the UK evidence as long-run gains; on the paper’s own calendar-time test, earn-out acquirers “do not outperform non-earnout acquirers in any of the three periods”. The long-run gains rest on regressions and on bids that the authors’ model, fitted to the same sample, classifies as the optimal choice.
Figure 4Earn-out announcement returns against deals paid otherwise
Announcement returns only, not realised deal value. Windows differ: one, two or five days.
Source: Kohers and Ang (2000); Barbopoulos and Sudarsanam (2012).
Sellers also received more on paper. In Kohers and Ang’s5 private targets, the median offer was 5.60 times book value with an earn-out, against 2.20 for cash and 2.30 for stock. That price counts the full earn-out, and “the actual premium paid would be less than the stated premium amount in many cases”. Their own check runs the other way too: with an earn-out of a third of the price and nothing paid, the median would still be 3.70 times book. In 79 deals where payments could be traced, 72 paid something, 36 paid in full and 7 paid nothing. Assuming 30% where the amount could not be traced, the authors estimate that about 62% of the stated amounts were paid. In the seven deals that paid nothing, the acquirer’s two-day return at announcement had been −1.75%, not significant, against 3.98% where all was paid.
Retention is the weaker half of the record. Of 56 target managers Kohers and Ang5 could follow, 37, about 66%, stayed beyond the earn-out period, and 36 of those 37 received part or all of the payment. Of the 19 who left, 6 received nothing. The study has no comparison group of deals without earn-outs, so it shows that staying and being paid went together, not that the earn-out kept anyone. Dahlen (2024)3 records that “management retention is often discussed in practice but receives only limited attention from scholars”. He counts two attempts to measure it: one finds earn-outs meant to keep managers booked at their maximum value, the other, with a related study, finds higher announcement returns where managers are likely to stay. Both read intentions and expectations, not whether anyone stayed.
Where do earn-outs break?
Where the measured unit stops being separable, on the argument the studies make; none of them measures it. Datar et al. (2001)4 list the mechanism among the costs of an earn-out: “As the firms become more integrated, joint costs arise whose allocation can lead to disputes by creating measurement problems.” They close by naming, as a question for further research, the trade-off a deal team has to price: “The existence of an earnout may also constrain the acquirer’s ability to obtain benefits from integrating the target and acquirer operations because the arrangement creates a need to evaluate the target’s performance as a separate entity.”
The samples fit that argument without testing it. Kohers and Ang (2000)5 describe earn-out deals as mostly cross-industry, “often with little integration occurring between the bidder and target after the merger is completed”, with targets often run as separate subsidiaries. Cain et al. (2011)2 conclude that “earnout periods are shorter when it is more likely that acquiring firms have the ability to distort the performance measure on which the earnout is based”; the proxy behind it is the volatility of stock returns in the target’s industry, which they read as noise in the performance signal, not integration. Datar’s4 survey fits the argument too: where the buyer ran the target, 2 of 11 earn-outs rested on net income and 5 of 11 on objective, non-accounting goals, against 7 and 2 of 15 where the seller ran it. The authors read this as sellers avoiding measures the buyer could shape. The numbers are small and come from one survey.
Figure 5Earn-out measures by who runs the target after the sale
Shares of each group's contracts, 15 and 11 in all: a pattern, not an estimate.
Source: Datar et al. (2001), Table 1.
On this essay’s reading, this is where the earn-out collides with post-merger integration: the synergy plan and the earn-out formula ask for opposite things from the same unit. The one study of earn-outs and synergies that Dahlen’s3 review finds, on Chinese deals, reports a significantly positive relation that varies with the size of the earn-out; this essay does not hold it.
The second failure runs the other way, through the seller. A limited window, Datar’s4 team warns, “may give the seller/manager an incentive to take actions [such as reducing research and development (R&D) expenditures] which maximize short-term profits at the expense of long-run profits”. Whether sellers do this is untested in the held sources: Dahlen (2024)3 lists cut R&D spending among the open questions, and the published tests of earnings management that the review counts are few, and four of five study the Chinese market.
The third failure is administrative. Tracing payments in annual reports, Kohers and Ang (2000)5 found “there is no standard method for reporting earnout payments”; some acquirers pooled several earn-outs, and some were themselves acquired during the earn-out period. An earn-out that rests on customer relationships carried by the seller’s people is also exposed to the attrition the essay on the retention number describes; whether an earn-out holds those relationships is a question the held studies do not test.
What do the studies leave unanswered?
How often earn-outs end in court. No held source measures a dispute rate. Dahlen’s3 review lists it as an open task, “Researchers could begin by measuring the share of earnout deals negotiated in court due to legal disputes”, and adds that “no research has attempted either a classification of the costs involved or a quantification of the respective costs”. Rates quoted on law-firm and advisory pages are not supported by anything held here.
Whether the findings travel. The four primary samples cover deals announced between 1984 and 2008, three of them US or UK and one a worldwide file. Three tie their tests to listed or publicly traded acquirers, Dahlen lists private equity buyers among the questions still open, and none of the four rests on a German, Austrian or Swiss sample. The review reports later and local evidence this essay does not hold: a rise to as much as 15% of North American deals, which the review dates before the financial crisis in one place and to the 2010s in another; from an international study by Viarengo, Gatti and Prencipe, earn-outs in 27% of UK transactions in its sample and 7% of German ones; and a German survey of finance professionals, by the review’s author and two colleagues, in which more than half used earn-outs in more than half of their deals. The last two count different things, deals in a database and the practice of those who answered.
The return findings are market reactions, mostly over a few days, and Dahlen finds the field “skewed towards short-term abnormal returns”. Firms choose earn-outs where they expect them to work, so the comparisons with cash and stock deals are not experiments. And whether earn-outs are paid, and keep people, in today’s deals is open: the payment and retention figures above come from one subsample of 110 deals announced between 1984 and mid-1996, and Dahlen notes that no study has followed up on actual payouts since. A deal team can use the evidence for the direction, where earn-outs fit and where they strain, but not for a probability that a given earn-out will pay or hold.
How could a deal team stress-test the measured target?
What follows is this essay’s synthesis, not a tested method. The sources say where earn-outs strain; the test turns each strain into a question asked before signing. Take an invented software unit with revenue of 10 million last year. The buyer pays 14 million at closing and up to 6 million more: 1 million for each million of second-year revenue above 11 million, capped at 17 million of revenue. The headline price is 20 million, of which 30% is contingent.
If revenue grows 10% a year, second-year revenue is 12.1 million and the earn-out pays 1.1 million, so the deal costs 15.1 million, not 20. Reaching the cap needs revenue 70% above last year. That is the distinction between headline and expectation, worked on one deal with growth added. Now add the integration plan: in month four the buyer moves the unit’s sales team into its own. From then on, the unit’s revenue depends on which salesperson books which deal, and the 6 million rides on an allocation rule the seller of the business does not control. The acquisition thesis has to say which of the two the deal is buying.
Figure 6An earn-out decision field for one measured target
Author's reading, untested. The corners are prompts for judgement, not a scored model.
Source: Author's decision field grounded in Datar et al. (2001) and Cain et al. (2011); untested.
Figure 7An earn-out stress test for one measured target
One row per earn-out tranche. Fill it in with the integration plan open, not after it.
Source: Author's decision worksheet grounded in Datar et al. (2001), Cain et al. (2011) and Kohers and Ang (2000); reader inputs only, untested.
The worksheet asks six things of each tranche. Is the measure one the seller’s team still controls, or one the buyer’s cost allocations will shape? How much growth does the threshold demand, and what does the formula pay if the business simply grows as it has? Which step in the integration plan merges the measured unit, and is the earn-out period shorter than the time before that step? What happens if the manager leaves, the unit is sold or the buyer is itself acquired? And are the accounting policies, the cost charges and the arbiter written down before signing, as they were in the contract Datar’s4 team printed?
A tranche that fails a column has three honest repairs. Measure something the integration will not touch, such as revenue from named products or a milestone. Delay the integration step until the window closes. Or move that part of the price back to closing and negotiate it there. Each repair costs something the deal was meant to buy, which is the point of asking before signing. Whether these repairs reduce disputes is untested.
References
- Barbopoulos, L., & Sudarsanam, S. (2012). Determinants of earnout as acquisition payment currency and bidder's value gains. Journal of Banking & Finance, 36(3), 678–694. https://doi.org/10.1016/j.jbankfin.2011.10.007
- Cain, M. D., Denis, D. J., & Denis, D. K. (2011). Earnouts: A study of financial contracting in acquisition agreements. Journal of Accounting and Economics, 51(1–2), 151–170. https://doi.org/10.1016/j.jacceco.2010.05.001
- Dahlen, N. (2024). Earnouts in mergers and acquisitions: A systematic literature review of a contingent payment mechanism. Management Review Quarterly. Online First, published online 2024. https://doi.org/10.1007/s11301-024-00429-w
- Datar, S., Frankel, R., & Wolfson, M. (2001). Earnouts: The effects of adverse selection and agency costs on acquisition techniques. Journal of Law, Economics, and Organization, 17(1), 201–238. https://doi.org/10.1093/jleo/17.1.201
- Kohers, N., & Ang, J. (2000). Earnouts in mergers: Agreeing to disagree and agreeing to stay. The Journal of Business, 73(3), 445–476. https://doi.org/10.1086/209649
Updated 1 change
- The reference list now names the edition read for each source published online before its journal issue: Dahlen (2024), Online First, published online 2024.