Verisk Still Has to Try to Buy AccuLynx

Most merger agreements contain a calendar. If the government has not finished its antitrust review by a date the parties picked at signing, either side can walk away, and that date is the buyer's insurance against the regulator. Verisk Analytics bought the insurance last July, when it agreed to pay $2.35 billion in cash for AccuLynx, a roofing-software company, with an outside date of November 26, 2025 and one extension to December 26. The Federal Trade Commission did not finish. On December 26 Verisk sent its termination notice, and on December 29 it announced that it would buy back the $1.5 billion of bonds it had sold to pay for the company. On Friday the Delaware Supreme Court, in a two-page order, told Verisk that it still has to try to buy AccuLynx, and has to close if the FTC ever says yes. The insurance did not pay out because of an email a Verisk product executive sent on August 5, 2025, six days after the deal was announced, to a competitor of AccuLynx, about a software integration. The email broke no promise, and the trial judge found "virtually no evidence suggesting that Verisk intended to scuttle the deal." It just cost Verisk the exit, which is a strange thing for an email to do.

Start with why the deal looked easy. Verisk makes Xactimate, the software insurance adjusters and contractors use to price the damage after a storm or a fire. AccuLynx makes the software roofers use to run their businesses. The two did not compete and neither sold anything to the other, so both sides concluded, in the trial court's words, that the merger "presented only minimal antitrust risk," and priced that conclusion into the calendar: AccuLynx proposed an outside date ninety days out, Verisk proposed six months, and they met at four months plus one.

Now the email. Verisk lets other software companies plug into Xactimate, and historically it has given everyone the same standard plug. In late 2024 ServiceTitan, which competes with AccuLynx, asked for a better one, live access to Xactimate's pricing data inside its own estimating tool, which the opinion calls an Enhanced Integration. The two sides were far apart on price (Verisk wanted about $20 per estimate, ServiceTitan offered $4), the AccuLynx rumor reached the market, and five days before signing Verisk's team decided to "pause on the new methods and capabilities and offer the baseline ecosystem play we offer to everyone." A week after signing, the vice president running the project wrote to ServiceTitan:

Now that the [AccuLynx] acquisition has been announced publicly, more information has been coming to [another colleague] and me. Considering this new information, we will not be able to continue down the path of placing Xactimate's pricing data directly into [ServiceTitan's] estimating solution.

This is an ordinary business email, and if you are an antitrust lawyer you have already winced, because "we won't give you the good integration now that we're buying your competitor" is, in miniature, the harm a vertical merger review exists to find. If you are a product executive at a data company, it was a Tuesday.

Under the Hart-Scott-Rodino Act the agencies have thirty days after a filing to clear a deal or send a "second request" for documents, which stops the clock until thirty days after the parties comply. Verisk and AccuLynx filed on August 19. On September 5, the same day Verisk's president of property and restoration solutions presented the business to FTC staff, ServiceTitan's general counsel told the FTC's staff attorney by phone that Verisk had been negotiating a bespoke integration and had "changed their minds" after the announcement. The agency built a "market reset" theory out of that call, that Verisk would favor AccuLynx with a richer integration and starve its rivals, and spent six weeks asking Verisk for the evidence it already had. On September 22 Verisk withdrew and refiled its notification, giving the agency another thirty days, to October 22. On October 14 the staff asked three times, in three ways, whether an AccuLynx competitor had ever asked for more than the standard integration and been refused, and the same executive who had presented in September said he "could not recall." Two days later Verisk wrote that it had "found no evidence that Verisk ha[d] declined to pursue an integration requested by a competitor of AccuLynx." On October 22 the FTC's chairman told AccuLynx's lobbyist that a competitor had been about to integrate with Xactimate "and Verisk cut the discussions because of this deal," and the second request went out that day.

A regulator that asks a question whose answer it already holds is not asking a question. It is finding out whether you will tell the truth, and Verisk's people, the court found, had forgotten it rather than hidden it.

Verisk's outside lawyers found the ServiceTitan emails on October 30 while culling documents for the second request and disclosed them on November 4. The court found that Verisk had "misrepresented and omitted information," that "its missteps were not intentional," and that the combination was "admittedly confounding." Verisk then collected nearly four million documents under a "quick look" agreement, paid close to $8 million in legal fees to review them, handed the agency more than 400,000, and took its one extension. On December 24 the FTC said it would want full compliance anyway, which Verisk's chief executive testified could take "18 to 24 months." On December 26 Verisk terminated.

Now the clause. The termination right in Section 9.1(f) lets the buyer terminate after the outside date unless its own "knowing and willful breach" prevented a closing condition from being satisfied, and nobody argued Verisk had done that. The right is also subject to the last sentence of Section 9.1, which binds both sides:

Notwithstanding anything else contained in this Agreement, the right to terminate this Agreement under Section 9.1 shall not be available to a party (a) if such party is then in material breach of its obligations hereunder or (b) if such party's failure to fulfill its obligations or to comply with its covenants under this Agreement, or other willful conduct, has been the primary cause of, or primarily resulted in, the failure to satisfy any condition to the obligations of the parties hereunder.

Two words in there, "willful conduct," are the whole case. Verisk said "willful" means culpable; AccuLynx said it means on purpose. Vice Chancellor Bonnie David sided with AccuLynx. Section 9.1(f) says "knowing and willful breach," and Section 9.2 defines a Willful Breach as one undertaken "with the intent of causing a breach of this Agreement." If the drafters had wanted the last sentence to require wrongdoing, they had a defined term for wrongdoing one section away and did not use it; they wrote "other willful conduct," which "ratchets down the required mental state" to "any voluntary and intentional conduct that caused a condition to fail." Ending the ServiceTitan talks counted "even if doing so was not wrongful and Verisk did not expect its decision would raise competition concerns." The sentence also demands a "primary cause," and the court found one: more likely than not, "but for the August 5 Termination Email, the FTC would not have been seriously concerned about a potential 'market reset,'" and Verisk's own expert, a former director of the FTC's Bureau of Competition, had testified that the email was "the reason why the FTC needed to investigate." The court did not need to decide whether Verisk breached its efforts covenants, or whether the FTC's theory is any good. Verisk did a thing on purpose, the thing was the main reason a condition failed, and the sentence says that is enough.

The remedy is specific performance, an order to keep trying: use commercially reasonable efforts to get HSR clearance, and close "if the FTC approves the transaction." Verisk had agreed in Section 11.12 "not to raise any objections to the availability of the equitable remedy of specific performance," so its objections were not available. It also owes AccuLynx $3.85 million of direct costs with interest, a number it never contested because its post-trial brief did not answer the request. "Any opposition is therefore waived." Verisk said it "strongly disagree[d]," appealed on August 18, and lost on October 23 before all five justices, who affirmed "on the basis of and for the reasons stated" below and added nothing. Its statement that day said it would proceed as ordered and that "there is no guarantee that the transaction will be completed."

The number is always wrong. The $1.5 billion of notes Verisk sold in August 2025 carried their own calendar, a promise to buy them back at 101 cents on the dollar if the agreement was terminated. Verisk bought them back on January 6, 2026, booking a $33.9 million loss on the redemption and $12.2 million of deal costs on a termination that did not happen, and if the FTC clears the deal it will need the $2.35 billion again.

I do not know whether the FTC will clear it; the second request Verisk walked away from in December is the one it must now answer in full. I do know what the parties think of the FTC's theory, because the opinion records both of them agreeing it was "unusual, unsupported by any real-world examples or empirical evidence, and likely would not have provided a basis to stop the deal." That concern has so far cost the deal more than a year, a trial, an appeal, and a bond redemption. A competitor whose bespoke integration had just been cancelled picked up the phone, and an agency that has yet to bring a case against this deal found a theory it could investigate for as long as it liked. That is most of what merger enforcement is now: a calendar, and the agency holds it.

So look at what each side bought at signing. Verisk got a short calendar, an efforts standard of "commercially reasonable" instead of a promise to do whatever it took, a list of "Burdensome Condition[s]" that the standard "will in no event require," so it could never be made to divest a business or litigate against the government, and no reverse termination fee; the phrase "termination fee" appears nowhere in the agreement, so if the calendar ran out, walking was free. A free walk makes a buyer indifferent to delay, because delay is what pays off the option. AccuLynx got the last sentence of Section 9.1, and the two words in it made Verisk the insurer of its own ordinary decisions for as long as the deal stayed open. Verisk told the court it never expected the email to raise competition concerns, and the court answered that even if so, it did not matter, which is the point. The sentence reached conduct Verisk never thought about, which is what insurance is for, and Verisk, reading the calendar and not the sentence, had written the policy on itself.

Columns are the views of their authors. They are not legal advice and not the position of the firm.