Why Is Globant (GLOB) Stock Down? The Guide Turned Negative

TL;DR

• Globant traded at $35.50 in extended hours on Thursday, down 13.39% from a $40.99 close it had spent the day rallying 5.81% to reach. One wire put the decline at 17.83% earlier in the session, so treat the magnitude as a moving quote until Friday settles it.
• The guidance did the damage. Full-year 2026 revenue guidance moved to $2.428-2.462bn from $2.462-2.508bn. The old range grew at both ends. The new one shrinks 0.4% at its midpoint.
• Adjusted diluted EPS came in at $1.40 against a $1.50 consensus. IFRS diluted EPS for the same quarter was $0.04. The bridge between them is $18.7m of share-based pay, $32.3m of "business optimisation" and $10.0m of acquisition charges.
• Glob.AI annual recurring revenue reached $52.8m, up 61% quarter on quarter, and the year-end target went up to above $110m from $60-100m. Set against a $2.45bn revenue base that is roughly 4.5%.
• At $35.50 the stock trades near 6x the midpoint of its own adjusted EPS guide and near 20x an IFRS run rate. The two numbers describe different companies.

Why Is Globant Stock Down?

Globant cut its full-year revenue guidance, guided the third quarter roughly $15m below where the Street sat, and missed on adjusted earnings. The stock gave back its whole day and more in extended trading.

The quarter itself was defensible. Revenue of $614.4m landed essentially in line with a consensus near $613m, free cash flow of $12.6m compared against negative $2.9m a year ago, and IFRS operating margin more than tripled to 3.2% from 1.0%. None of that is what moved the stock at 4:30pm.

The Board

Two margins, one quarter, and a ten-point gap that a $32.3 million charge helps explain.

A Guide With No Growth Left In It

Here is the change, stated the way the company stated it. Old full-year range: $2.462bn to $2.508bn, or +0.3% to +2.2% year over year. New range: $2.428bn to $2.462bn, or -1.1% to +0.3%. Adjusted EPS guidance went to $5.75-6.15 from $6.10-6.50.

Read the top of the new range against the bottom of the old one and they are the same number. Globant has taken its best case down to what used to be its worst case.

Third quarter guidance is $607-615m, against a consensus near $626.2m. Management named three causes: weakness in new markets, travel and hospitality clients under pressure from oil prices, and longer decision cycles in North America. That last one is the industry's story rather than Globant's, and it has been visible in enterprise software all summer.

The arithmetic underneath is worth doing. First-half revenue was $1.2215bn. The new full-year range leaves $1.207bn to $1.241bn for the second half, so at the low end Globant is guiding to a second half smaller than its first. Revenue has now been flat for four quarters: $607.1m in Q1, $614.4m in Q2, and a Q3 guide whose midpoint is $611m.

$1.40 and $0.04 Came Out of the Same Quarter

The adjusted number the Street quotes is $1.40. The IFRS number in the same release is $0.04, on net income of about $1.8m. Globant discloses the bridge properly, so this is arithmetic rather than an accusation:

• Share-based compensation, equity settled: $18.7m, worth $0.44 a share
• Business optimisation costs: $32.3m, worth $0.75 a share
• Acquisition-related charges: $10.0m at the operating line, worth $0.55 a share on the EPS bridge
• Tax effect of the adjustments: -$0.38 a share

IFRS operating profit of $19.9m plus the first three lines gives adjusted operating profit of $80.9m, a 13.2% margin against 3.2% reported.

The line I keep looking at is business optimisation. Globant's Q1 reconciliation carried share-based pay and acquisition charges and no optimisation line at all, so this $32.3m is new this quarter. Headcount fell to 27,411 from 28,510 on 31 March, which is what the charge bought. Free cash flow for the quarter was $12.6m. The release does not split how much of the $32.3m was cash out the door in Q2, and I would want that number before treating the margin as clean.

My rule on adjusted metrics is boring: an add-back that appears once is an event, and an add-back that appears every quarter is a cost. This one is new as of Q2, so it gets the benefit of the doubt. If it turns up again in the Q3 reconciliation, the 13.2% margin is the fiction and the 3.2% is the business.

The AI Business Is Growing Fast, From Almost Nothing

Glob.AI is the genuinely good news here. ARR reached $52.8m, up 61% quarter on quarter, and management raised the year-end target to above $110m from a prior $60-100m. AI Pods sit at about 2% of revenue today and are guided to 4% by year end.

Now the denominator. $110m of ARR against a $2.45bn revenue base is about 4.5%. For the AI business to offset a flat-to-declining core, it has to grow faster than the core shrinks, and at 2% of revenue it is not close yet.

CEO Martín Migoya said on the call that AI Pods are priced on output and value rather than billed hours, and that customers often keep spending similar amounts while getting more productivity. I think that sentence contains both the bull case and the bear case. Same dollars for more output is a wonderful thing for the client and a flat revenue line for the vendor. It also explains why headcount can fall 1,099 in a quarter without the top line moving. Pegasystems ran into a version of the same question this summer, and IBM's software crash in July was the loudest example of it.

Client concentration is stable and unremarkable: top customer 8.9% of revenue, top five 21.6%. Accounts above $1m of annual revenue slipped to 331 from 339.

Is It Cheap Yet?

At $35.50, on the midpoint of the new adjusted EPS guide of $5.95, Globant trades at 5.97x. That is a distressed multiple for a business generating cash, and it is why the name keeps attracting value buyers on the way down. The 52-week range is $27.56 to $79.67, so the stock has already lost more than half its high.

On the reported numbers it is a different picture. First-half IFRS diluted EPS was $0.89 in total, $0.85 in Q1 and $0.04 in Q2. Annualise that at $1.78 and $35.50 is roughly 20x, which for a company guiding to no revenue growth is ordinary rather than cheap.

I am not buying it here, and my objection is not the multiple. It is that I cannot price a services business whose own guidance permits a revenue decline, when the thing replacing the old revenue is 2% of the total. What would change my mind is narrow and testable: a Q3 print at the top of the $607-615m range with the North America book growing sequentially, or AI Pods above roughly 6% of revenue with the core no longer falling. Either one turns 6x from a warning into a discount. Other names reporting into the same window are on the earnings calendar, and dLocal's Thursday print is a useful contrast: volume growing at 92% while its own unit economics compress.

The One-Line Read

Globant is cheap on a number that adds back a $32 million restructuring, and ordinary on the number that does not. I want one quarter of actual revenue growth before the 6x means anything.