Summary

  • ICE initially offered US$175 per MarketAxess share, a 33.6% premium to the 27 May close. The signed deal pays US$167, described as approximately a 33% premium to the 29 July close.
  • The premium barely changed, but the cash price fell US$8, or about 4.6%. Each percentage uses a different market-price denominator, so the premium cannot show whether absolute consideration was preserved.
  • The path was US$175, then US$165 from ICE, a US$170 counter from MarketAxess and a US$167 final proposal. MarketAxess later sought a special dividend; ICE refused additional monetary value.
  • The final price sits inside both management cases used for the fairness analysis. Buyer synergies, financing and regulatory protections remain separate receipts, and the acquisition has not closed.

The premium stayed; the cheque shrank

On 27 May, Intercontinental Exchange, Inc. made a non-binding proposal to acquire MarketAxess Holdings Inc. for US$175 a share in cash. MarketAxess's new preliminary merger proxy says that price represented a 33.6% premium to the US$130.94 closing price that day.

The agreement signed on 29 July pays US$167 a share. The same proxy describes that as approximately a 33% premium to MarketAxess's 29 July close. It is tempting to read those percentages as evidence that stockholders kept substantially the same bargain. They did not keep the same dollar price.

US$175 minus US$167 is US$8. Relative to the initial proposal, the reduction is about 4.57%. Both premiums can still sit near 33% because they divide different offers by different unaffected share prices. The denominator moved with the market. A premium is useful for comparing an offer with one reference price; it is not a record of how the offer changed through negotiation.

This distinction is not an accusation that the final price is unfair. It is a measurement boundary. The headline premium answers, “How far above this day's market price?” The cash consideration answers, “How much will one share receive if the deal closes?” Those are different questions.

Four prices reveal the negotiation

The proxy supplies the path that the announcement compressed. ICE's May proposal asked for a 60-day exclusive negotiation period. MarketAxess did not grant exclusivity and remained free to engage elsewhere. From roughly March through July it held preliminary discussions with about six other parties, including five financial sponsors and one strategic party, but none submitted a written indication of interest or a price.

By 17 July, ICE had moved down to US$165. It cited market conditions, due diligence, a higher cost of capital and its assessment of the value available from the transaction. MarketAxess's share price had fallen from US$130.94 on 27 May to US$113.80 on 17 July. The company responded that it would continue at US$170. ICE then raised its proposal to US$167.

At a 21 July board meeting, the advisor noted that analyst target prices had fallen more than 35% year to date while EPS estimates had declined only about 3%. The proxy characterises that divergence as a rerating of the earnings multiple. The figures do not prove which forecast was right. They show why an apparently stable premium can accompany a lower absolute bid: the market reference and the price investors were willing to place on each dollar of expected earnings had changed.

MarketAxess then tried another route. It asked for a one-time special dividend before closing, intended to deliver extra value without changing the US$167 merger consideration or the stated premium. ICE declined a higher price, a special dividend and other additional monetary value. The proxy does not disclose a proposed dividend amount. Inventing one would turn a negotiation record into a false valuation.

Two standalone cases do not resolve to one value

The board did not evaluate only a premium. Management prepared a Base Case and a lower Sensitivity Case for MarketAxess as a standalone company, without the merger or its costs. The sensitivity case assumed half the incremental market-share gains of the base case for several bond markets in 2027 and 2028, a steeper decline in fees per million and a lower post-stock-compensation EBITDA-margin path.

The distance compounds. In the Base Case, 2031 revenue reaches US$1.222 billion, EBITDA after stock compensation US$623 million and unlevered free cash flow US$398 million. The corresponding Sensitivity Case figures are US$1.095 billion, US$539 million and US$342 million. These are management projections prepared for the transaction process, not promises, external consensus or ICE guidance.

J.P. Morgan's discounted-cash-flow analysis produced a range of US$166.51 to US$230.35 per share for the Base Case and US$145.14 to US$200.06 for the Sensitivity Case. US$167 lies within both. It is only US$0.49 above the low end of the base range, but a lower endpoint is not a minimum fair price. The advisor said its analyses were reference points, not appraisals or predictions, and considered them as a whole.

That is why neither side can be reduced to one convenient number. Selecting US$230.35 would ignore discount-rate, terminal-growth and operating assumptions. Selecting US$145.14 would do the same in the other direction. The relevant public receipt is that the final offer crossed both analytical ranges while remaining exposed to sharply different standalone paths.

ICE's synergy belongs to a different ledger

ICE's transaction presentation gives the buyer's case. It describes approximately US$6.0 billion of equity value and US$5.7 billion of enterprise value, targets US$100 million of run-rate cost synergies and presents the price as about 10.6 times enterprise value to adjusted EBITDA on a fully synergised basis. ICE expects gross leverage of approximately 3.4 times at closing and targets 3.0 times or less within 18 to 24 months.

None of that increases the US$167 cheque. The target's stockholders receive fixed cash, not a continuing share of achieved savings. The synergies could be valuable to ICE and still be absent from target-holder consideration. They could also arrive late, cost more to produce or differ from the target. The presentation supplies an incentive and monitoring plan, not a completed result.

Financing has the same boundary. ICE's second-quarter Form 10-Q discloses a 364-day senior unsecured bridge commitment of up to US$6.2 billion. It is backup financing if permanent debt is not obtained, and the commitment can be reduced to zero by a term loan, senior notes and an amended revolving facility. US$6.2 billion is not a disclosed draw, final acquisition cost or settled funding mix.

MarketAxess's operating record also resists a simple failure story. Its second-quarter Form 10-Q and results showed US$218.4 million of revenue, relatively flat year on year, and US$186.9 million of commission revenue, down 3%. US credit commission revenue fell, while emerging-markets commission revenue and several trading channels supplied counterevidence. A lower offer after due diligence is not proof that the business collapsed.

Signature changes who controls the next move

The deal remains conditional. MarketAxess stockholders must approve it, regulators must clear it and other closing conditions must be met. The initial outside date is 29 July 2027, with up to two automatic six-month extensions when specified regulatory conditions remain open and the other conditions are satisfied or would be satisfied.

The agreement prices different failure paths. MarketAxess can owe US$148.8 million under specified termination circumstances. ICE can owe US$327.4 million if specified antitrust failures prevent closing. The amounts are not probabilities or compensation for every delay. They allocate particular contractual risks.

The advisor's incentives are also disclosed. MarketAxess expects to pay J.P. Morgan about US$65.5 million, of which US$3 million became payable with the opinion and the remainder depends on completion. The filing also describes relationships with both parties. That does not invalidate the opinion; it tells readers where the advisor's economic incentives sit and why the underlying ranges and limitations should remain visible.

Sources