Megan Skiendel: David, hey — I have been staring at this Equifax release since six this morning and I genuinely cannot decide if I'm angry or impressed.
David Sterling: Well, pick one. What's the line that did it?
Megan Skiendel: The stock. Down 11% premarket on a beat. Eleven. They printed $1.700 billion in revenue, up 11% year-over-year, beat their own guidance, and the market just — took a hammer to it.
David Sterling: That's not a mystery. That's the Q3 guide. $1.68 to $1.71 billion — below consensus.
Megan Skiendel: Sure, but listen — they doubled the AI savings target. $75 million to $150 million. On the same call where Mark Begor announced a $750 million acquisition of Círculo de Crédito. And they returned $366 million to shareholders in the quarter. That is a lot of narrative being launched at once.
David Sterling: Which is exactly what EFX2028 is designed to look like. Begor's running a framework. The question is whether the framework survives what mortgage does in Q3.
Megan Skiendel: Honestly, I think the market already decided that question. Seventeen percent down year-to-date — that's not a reaction, that's a verdict being confirmed.
David Sterling: A beat is a beat. I'm not ready to call it a verdict yet — walk me through what you think the market actually saw.
Megan Skiendel: What the market saw is a restaurant that had one great Saturday because a music festival came to town. The food was fine — the crowd was rented. U.S. mortgage revenue up 40% is the festival. The second the acts pack up, USIS gives back most of that 17% growth.
David Sterling: That's the right frame. And management is telling you the same thing — Q3 guide of $1.68 to $1.71 billion. If you believed 40% mortgage growth was structural, you don't guide below LSEG consensus. You just don't.
Megan Skiendel: I hear that — but wait, the Diversified Markets number. Within USIS, that segment accelerated over 300 basis points to 6% growth in Q2. That is not mortgage. That's a different customer set moving.
David Sterling: 6% does not offset a 40% spike carrying the segment.
Megan Skiendel: Mathematically, no — but it's the direction that matters. It's actually moving, which means — I mean, it's not nothing.
David Sterling: Workforce Solutions hit $705.4 million, up 7%. I'll grant that's real — Verification Services, Talent Solutions, Consumer Lending, none of that is a mortgage story. But the headline beat came from USIS, and USIS ran on mortgage fuel.
Megan Skiendel: Which is exactly what Begor is not saying out loud on the call.
David Sterling: He doesn't have to. The Q3 guide says it for him. $1.68 to $1.71 billion is the tell — that's management implicitly admitting the festival crowd is leaving.
Megan Skiendel: Honestly, that's what stands out. Begor beats Q2, doubles the AI target, announces Círculo, returns $366 million to shareholders, and still can't guide above consensus. At some point the EFX2028 framework has to produce a quarter that doesn't need a footnote.
David Sterling: That footnote is the thread I'd pull first. 120 basis points of EBITDA expansion to nearly 35% — but the comparison excludes FICO mortgage royalties. What exactly is in that exclusion? Because if mortgage EBITDA is structurally lower-margin than the rest of the book, then the efficiency story is even more dependent on cost cuts than the headline suggests.
Megan Skiendel: That's the number I keep staring at. $150 million in AI savings over three years — against $6.7 billion in annual revenue. That is 0.7% of revenue. You cannot call 0.7% a structural transformation with a straight face.
David Sterling: I don't disagree on the scale. But 120 basis points is real. You cannot manufacture 120 basis points — not in a single quarter, not against a 50-basis-point annual target. That's 2.4 years of targeted efficiency in six months. The number is too large to be purely narrative.
Megan Skiendel: Or — wait, actually — the number is exactly large enough to do a specific job. Begor doubled the target from $75 million to $150 million on the same call he announced Círculo. Not the quarter before. Not six months later. Same call. That is a prepared cushion, not a discovery.
David Sterling: The timing optics are real. I'll grant that.
Megan Skiendel: I mean, I've seen this exact move. You announce a big acquisition, you give the board a savings number that tells them integration costs won't crater margins — and then if margins do compress, you say mortgage was soft and the savings washed it out. That's not capability acceleration, that's optionality management. Begor is pre-loading the narrative.
David Sterling: Then what's the falsifiable version? If the $150 million is real, where does it show — call center headcount, legal spend, software licensing costs? The release says call centers, document processing, code development, back-office legal and HR. That's a specific list. Which line breaks first if the savings aren't there?
Megan Skiendel: Honestly — the one that breaks first is adjusted EPS. They printed $2.25 in Q2, up from $2.00 a year ago, 13% — that's ahead of guidance. But if mortgage normalizes in Q3 and the AI savings don't compound fast enough, that cushion disappears before Círculo generates a dollar of revenue.
David Sterling: And that's before the debt math on the acquisition even enters the picture — which, frankly, is the part of this conversation that makes the $366 million in shareholder returns look like a very deliberate sequencing decision, or a very concentrated execution risk. We'll get to which one of us concedes that point.
Megan Skiendel: That sequencing question is actually where I want to land, because — look, I've been pushing the engineered-narrative read, and I'm not walking that back. But the $1.49 billion buyback completion? That's pre-committed. That's a contract executing on schedule, not a capital allocation decision Begor made in Q2 2026. I'll give you that narrow point.
David Sterling: The timing optic on doubling the AI target — same call as Círculo — that one I'm not defending. That's real.
Megan Skiendel: Pre-committed or not, the net effect is identical. Equifax exits Q2 having deployed north of $1.1 billion in a single quarter — buyback completion plus $750 million for Círculo — and the only durable revenue floor I can actually see is the Workforce Solutions government contracts.
David Sterling: The $300 million in annual contract value. First half of 2026.
Megan Skiendel: Right — but if mortgage normalizes in Q3 and Círculo integration costs hit inside 18 months, that $300 million is the entire load-bearing wall. That floor has to hold everything.
David Sterling: What's the net debt-to-EBITDA post-Círculo? That's the number. If it crosses 2.5x before Mexico integration compounds — that is a covenant pressure story, not an efficiency story.
Megan Skiendel: And Begor has no answer to that on the call — because the AI savings cushion, I mean, even if the full $150 million materializes, it doesn't compound fast enough to absorb that pressure before the mortgage cycle gives back.
David Sterling: The gap stays open. The pre-committed buyback is technically clean. The concentration of risk in one quarter is structurally real.
Megan Skiendel: The market is already answering the question, though. EFX down 17% year-to-date — on a day they beat. That's not investors waiting to see how Círculo integrates. That's a verdict on whether EFX2028 holds three simultaneous moves at once. Begor can call it a strategic framework all he wants. The mortgage cycle does not read strategy decks.
David Sterling: Neither does a credit bureau in Mexico that just became Equifax's most expensive bet.
Megan Skiendel: No. It doesn't. And I think — I mean, that's where I actually land on this. Not angry, not impressed. Just uneasy. The pieces are all defensible individually. Together they're a lot to hold.
David Sterling: Mm. Q1 2027 will settle it. Either the timeline compresses or the debt math does. Thanks for working through this — you kept me honest on Begor.
