← EPAM deep-dive | SCREEN
Globant S.A. (GLOB) — short-thesis deep-dive
Second-order short candidate flagged by EPAM's 6-peer cross-read. GLOB sits ~2 quarters AHEAD of EPAM on the same AI-disruption curve — already missed organic growth, already cut workforce, already admitted AI shrinks headcount. Luxembourg-incorporated (files 20-F + 6-K). CIK 1557860. Generated 2026-04-18.
Short thesis (bottom line): Globant is the visible proof-of-concept that EPAM's "AI = tailwind" narrative collapses on contact with cash. FY2025 revenue grew only +1.3% — vs. the company's historical +15-25% organic run-rate — and FY2026 guide is +0.2% to +2.2%, essentially flat. Management already executed a "business optimization plan" (headcount reduction) in 2024-2025, and CFO Juan Urthiague stated outright on the Q4 2025 call: "The AI Pod model by definition requires less people… the sector will start to change a little bit that trajectory of headcount and revenue that we have seen in the past twenty years." (peer_GLOB_crossread.md, line 47). That is a materially harder admission of AI-driven labor substitution than anything EPAM, ACN, or INFY management has said on tape. CTO Diego Tartara semi-admits per-profile bill-rate compression ("Yes. There's pressure there" — Q3 2025, line 185), and CFO Urthiague concedes T&M pricing "is not going to be a massive growth this year" (Q4 2025, line 251). The whole AI Pods subscription pivot IS the company's response to hour-based T&M becoming structurally unsustainable — dressed as product innovation.
Edge vs EPAM: The fundamentals have already rolled over. GLOB 2025 +1.3% / 2026 guide +1% << EPAM 2025 +15.4% reported (4.9% organic) / 2026 guide +4.5%. But the stock is -64.4% from 52w high ($140.36 → $49.97) vs EPAM -40.7%. This is the priced-in question: GLOB has been punished almost fully for the deceleration already in the tape. The remaining short thesis on GLOB is narrower than on EPAM and rests on (a) FY26 guide being cut further during the year (both 2024 and 2025 guides were cut mid-year), (b) AI Pods ARR missing the $60-100M 2026 exit-rate target, and (c) a second "business optimization" / headcount cut around Q2-Q3 2026 when utilization fails to inflect.
Signal strength: MEDIUM-WEAK on residual direction, STRONG as read-through to EPAM. GLOB is crowded short (14.77% of float, +6.5% MoM), sell-side still has +40% upside to $70 mean target on 23 analysts (not yet capitulated) but the multiple has already compressed to 7.5x Fwd P/E — the cheapest in the peer set ex-DAVA. Options chain is thin (no Jan 2027 LEAPS; longest expiry is Nov 2026 with OI=1 on OTM puts). Trade is expressible but sizing must be small.
Signal grid
Each row shows an angle of the GLOB thesis and the strength of evidence. Red = strong short signal, orange = medium, yellow = weak, green = counter-signal, gray = neutral. Quotes cite epam_agents/peer_GLOB_crossread.md with transcript line numbers.
| Angle | Key finding | Signal |
|---|
| 1 | Revenue growth collapse | FY25 reported rev +1.3% vs historical +15-25% organic. FY26 guide +0.2% to +2.2% — near-zero, a multi-year trough. Guidance cut multiple times in 2025. | strong |
| 2 | AI-reduces-headcount admission | CFO Urthiague, Q4'25 line 47: "The AI Pod model by definition requires less people… the sector will start to change a little bit that trajectory of headcount and revenue." — directest AI-substitution admission among all peers. | strong |
| 3 | Per-profile bill-rate compression | CTO Tartara, Q3'25 line 185: "On a profile basis … price compare to what it used to be four years back? Yes. There's pressure there." — half-admits per-head rate compression; CEO immediately contradicts. | medium-strong |
| 4 | T&M pricing going flat | CFO Urthiague, Q4'25 line 251: pricing "is not going to be a massive growth this year in terms of pricing for the general business." The AI Pods pivot IS a response to T&M breakage, not an opportunistic overlay. | medium-strong |
| 5 | Sales-cycle elongation | CTO Tartara Q3'25 line 249 ("larger period of times lead times until closure") + CBO Matzkin Q4'25 line 207 ("bigger deals still lagging just a little bit behind, slower than we would like") — acknowledged on both calls and one quarter earlier than EPAM. | medium-strong |
| 6 | Workforce reduction admitted | CFO Urthiague Q3'25 line 172: "Last quarter, we announced a business optimization plan to align our company, our headcount, to the needs of the business." — explicit workforce reduction. EPAM has not made an equivalent aggregate admission. | strong |
| 7 | Utilization sub-target | Q3'25 line 220: utilization is "still below our target of 80%" despite 50bps Q-over-Q improvement. Dynamic similar to EPAM's 75.4% drift, but GLOB acknowledges openly. | medium |
| 8 | AI Pods — small numerator, big claim | AI Pods = $20.6M exit-rate ARR 2025 (<1% of rev); 2026 target $60-100M exit ARR. Same playbook as EPAM's $105M "AI-native" → $600M target — both pitching AI practice as the growth engine while core business stalls. | medium |
| 9 | Mgmt CEO-level denial posture | CEO Migoya Q3'25 line 181: "We don't look measure pressure on the pricing environment… revenue per head is doing okay." Same industry-standard IR denial as EPAM / ACN / INFY — cracks ONE level down the org (CFO, CTO). | medium |
| 10 | Peer read-through to EPAM | GLOB is ~2 quarters ahead of EPAM on the same curve: already printed near-zero growth, already cut workforce, already admitted AI substitution. EPAM still denying at aggregate level — GLOB is the trailer to EPAM's movie. | strong (as read-through) |
| 11 | Priced-in — stock reaction | -64.4% from 52w high, -53% 1y, -21% YTD. Fwd P/E 7.5x is the cheapest in peer set ex-DAVA (3.7x). Mkt cap $2.2B (vs $7.1B EPAM). Most of the fundamentals deterioration is in the tape. | weak (directional) |
| 12 | Sell-side NOT capitulated | 23 analysts, "buy" rec (mean 2.13), target mean $70 (+40.35% upside), median $68, target high $120. Institutional ownership 113% (short rebate lending). If sell-side cuts to $50-55 in a cluster, stock has -10-30% room. | medium (timing edge) |
| 13 | Short crowding | 14.77% of float short, +6.50% MoM (6.20M shares vs 5.83M prior month). Days to cover 5.03. Crowded but NOT as crowded as EPAM's 18.1% / +47.6% spike. | medium |
| 14 | Options liquidity constraint | Only 4 expirations available: May / Aug / Oct / Nov 2026. No Jan 2027 LEAPS. Oct 2026 is the practical long-dated expiry (180 DTE). OI thin beyond ATM. Limits position sizing and strike choice. | strong (mechanical) |
Top 5 red flags (with direct quotes)
1. "The AI Pod model by definition requires less people" — CFO explicitly admits AI labor substitution.
On the Q4 2025 call (Feb 26, 2026), CFO Juan Urthiague said this in response to a question on gross margins. This is more direct than anything EPAM's Jason Peterson or ACN's Julie Sweet has said — Sweet's stance is "AI is expansionary, not deflationary." Urthiague explicitly telescopes industry structure: headcount-leveraged T&M is breaking.
"The AI Pod model by definition requires less people. It is, you know, it is the AI Pods, which are agents supervised by some few people. So there is less need for talent. So I think that not just for Globant S.A., but in general, the sector will start to change a little bit that trajectory of headcount and revenue that we have seen in the past twenty years." — Urthiague, Q4 2025 call (peer_GLOB_crossread.md, line 47)
2. Per-profile bill-rate compression — CTO admits on tape while CEO denies.
Citi's Bryan Bergin asked directly about GenAI pricing pressure through cost-saves passed to clients. CEO Migoya denied. CTO Tartara, in the same call, contradicted.
Migoya: "We don't look measure pressure on the pricing environment… revenue per head is doing okay." (Q3'25, line 181)
Tartara: "If you think about the individuals, like, on a profile basis and you tell me how does probably that price compare to what it used to be four years back? Yes. There's pressure there. But the mindset today is about efficiency." (Q3'25, line 185)
This is the single cleanest peer-level admission of T&M per-head rate compression anywhere in the cross-read — an on-the-record contradiction between CTO and CEO in the SAME earnings call.
3. Bigger deals are lagging — CBO concedes enterprise hesitation on large AI investments.
"What we are seeing is, you know, shorter sales cycles in smaller deals, and the bigger deals still lagging just a little bit behind, slower than we would like to, in terms of closing and ramping up… it is true that the clients are cautious, are taking time to make decisions when it comes to very large investments." — CBO Fernando Matzkin, Q4 2025 call (peer_GLOB_crossread.md, line 63)
"Evaluating a proposal, how to properly use new platforms, the proper model, the implementation which is super complicated, and you have three different offers that might be totally different… tends to take a lot more time." — CTO Tartara, Q3'25 line 249
This is the
same hedging vocab ("sales cycle elongation" / "procurement elongation") that EPAM introduced for the FIRST time in Q4 2025 — GLOB was saying it one quarter earlier AND ties it to AI-platform evaluation maturity.
4. Workforce reduction already executed; 2026 will require a second.
"Last quarter, we announced a business optimization plan to align our company, our headcount, to the needs of the business… You have to keep in mind that we started the year with expectations of much higher growth. Now all that is aligned." — CFO Urthiague, Q3'25 line 172 (peer_GLOB_crossread.md, line 42)
Globant has already done what EPAM has conspicuously avoided doing at the aggregate level — explicitly cut headcount and called it out. If 2026 revenue comes in at the low end of the +0.2% guide (plausible given 2024 and 2025 guides were both cut mid-year), a
second optimization plan is almost certain by Q2-Q3 2026. That is the path to FY27 earnings resets that are
not in the $70 consensus target.
5. "The whole industry growing slower, which is kind of a killer" — CEO candidly on industry deceleration.
"I think 2025 was a year of uncertainty in general. Companies retracted budgets in many cases… we suffered that." — CEO Migoya, Q4'25 line 233
"The whole industry growing slower, which is kind of a killer." — Migoya, Q4'25 (peer_GLOB_crossread.md, line 86)
Migoya is more candid on industry-wide deceleration than any other peer CEO in the cross-read (ACN's Sweet, INFY's Parekh, EPAM's Fejes all frame it as idiosyncratic or macro-transient). Pairs with his cannibalization statement:
"I am not in a position to prevent cannibalization. So I want that transformation to happen." (Q4'25 line 281) — acknowledging the legacy model is being replaced.
GLOB vs EPAM — position on the disruption curve
Where is each company on the same trajectory? Heuristic: GLOB is ~2 quarters ahead.
| Signal | EPAM (denial phase) | GLOB (admission phase) | Who is further along? |
|---|
| AI bill-rate compression | Denied directly by CFO Peterson Q4'25 | CEO denies, CTO admits per-profile rate pressure; CFO admits T&M pricing won't grow | GLOB (one level down the org) |
| Utilization decline | Drifted 78.1% → 75.4% over 3Q; not volunteered | Sub-80% target openly acknowledged; +50bps Q3'25 QoQ inflection | GLOB (disclosure transparency) |
| Workforce reduction | No aggregate admission; buyback-funded retention | Explicit "business optimization plan" / layoffs on tape | GLOB (materially) |
| AI-reduces-headcount | Denied: "not input limited" — Fejes | Admitted: "AI Pod model requires less people" — Urthiague | GLOB (materially) |
| Sales-cycle elongation | First acknowledged Q4'25 ("modest extension") | Acknowledged Q3'25 AND Q4'25 on both calls | GLOB (one quarter earlier) |
| Revenue growth reality | FY25 reported +15.4% (organic +4.9%); FY26 guide +4.5% midpoint | FY25 +1.3%; FY26 guide +0.2% to +2.2% midpoint +1.2% | GLOB (already through the wall) |
| Stock reaction | -40.7% from 52w high; Fwd P/E 9.4x | -64.4% from 52w high; Fwd P/E 7.5x | GLOB (priced more) |
| Short crowding | 18.1% of float, +47.6% MoM — very crowded | 14.77% of float, +6.50% MoM — crowded but less so | EPAM more crowded |
| Mkt cap / liquidity | $7.1B; LEAPS thru Jan 2027 available | $2.2B; only 4 expirations, no LEAPS past Nov 2026 | EPAM tradeable, GLOB options thin |
Reading: GLOB is empirically what EPAM becomes if the denial narrative breaks in 2026. The short thesis on GLOB itself is therefore narrower than on EPAM — GLOB is closer to the fundamental floor and options are less liquid. The primary value of GLOB to this research project is as the falsifiable precedent: GLOB's trajectory (deny → miss → admit → cut → guide flat → stock -65%) is the base-rate template that EPAM is 2 quarters behind on. If GLOB's FY26 guide holds or is cut, EPAM's FY26 re-rate is the higher-conviction trade.
Priced-in check — what the market has already digested
-64.4%from 52w high
($140.36 → $49.97)
-53.1%1-year return
-20.8%YTD return
14.77%short % of float
(crowded, not extreme)
+6.50%MoM Δ shares short
5.83M → 6.20M
5.03days to cover
7.5xFwd P/E
(peer med 11.6x = -35.6%)
6.3xEV/EBITDA
(peer med 8.4x = -24.6%)
$2.2Bmarket cap
EV $2.5B
+40.35%upside to mean target
($50 → $70 mean; $68 median; high $120)
23analyst opinions
rec "buy" (mean 2.13)
-4.7%rev growth YoY (TTM)
113%institutional ownership
(short-lending float)
-7.1%180-DTE put-call ATM skew
calls bid higher — puts cheap on surface
Reading: GLOB has been punished hard — stock -64%, multiple compressed to sub-peer levels, revenue already contracting. But sell-side still has +40% upside and 23 "buy" ratings. The residual edge is in sell-side capitulation timing — when 5-8 of the 23 analysts cut targets toward $50-55 in a cluster (likely triggered by a Q2 2026 guide reduction), the stock can retest $40 (the 52w low). The IV skew at longer tenors is negative (calls bid higher than ATM puts by 7-10 vol points on Oct/Nov 2026) — same structural feature as EPAM, puts are RELATIVELY cheap on the surface.
Options chain snapshot (yfinance pull, 2026-04-18)
Spot $49.97. Only 4 expirations available — no Jan 2027 LEAPS. OI/volume shown for the OTM10 put (10% OTM from spot, ~$45 strike).
| Expiration | DTE | ATM Call IV | ATM Put IV | Put-Call Skew | OTM10 Put ($45) IV | $45P Bid/Ask | $40P Bid/Ask | $45P Vol | $45P OI |
|---|
| 2026-05-15 | 26 | 81.0% | 75.9% | -5.1% | 79.2% | $2.00 / $2.25 | n/a | 35 | 456 |
| 2026-08-21 | 124 | 69.7% | 63.0% | -6.8% | 63.4% | $4.50 / $5.10 | $2.40 / $3.30 | 34 | 139 |
| 2026-10-16 | 180 | 66.8% | 59.6% | -7.1% | 62.2% | $5.80 / $6.20 | $3.80 / $4.30 | 2 | 37 |
| 2026-11-20 | 215 | 69.8% | 60.3% | -9.5% | 62.4% | $5.90 / $7.60 | $4.10 / $5.30 | 0 | 1 |
Trade structure (thesis-driven, sized small due to crowding + thin chain):
Primary: Oct 16, 2026 $40 put (~180 DTE, ~20% OTM), mid $4.05 (bid 3.80 / ask 4.30, spread ~12%). IV 64.5%. Implies ~8% probability-weighted payoff per $1 of premium at $32 spot. Catalyst windows inside expiry: Q1 2026 print (May 2026), Q2 2026 print (Aug 2026 — guide refresh window). Trade thesis: 2026 full-year guide cut mid-year, same pattern as 2024 and 2025, driving a retest of the $41.56 52w low. Position sizing small — OI 37 is thin; expect 10-15% slippage vs mid on exit.
Alternative: Aug 21, 2026 $45 put (~124 DTE, ~10% OTM), mid $4.80 (bid 4.50 / ask 5.10). Denser OI (139), tighter spread (~12%), centered on Q2 2026 print. Higher delta, faster theta decay — use if the catalyst is Q1 2026 print (May 15 call) being the binary event.
Avoid: Nov 2026 puts (OI=1 on $45 strike — effectively no market). Avoid May 2026 front month — 81% ATM call IV is elevated and short-squeeze-on-any-beat risk is high with 14.77% float short.
Kill triggers: (a) Q1 2026 revenue ≥ +3% YoY with 2026 guide RAISED, (b) AI Pods exit ARR run-rate >$30M by mid-year (implies $60-100M target achievable), (c) any announcement of a strategic M&A premium bid — the $2.2B cap makes GLOB a plausible take-out for a tier-1 consulting firm at 40-60% premium.
Counter-arguments (the honest bull case)
1. Priced-in: most of the bear case is already in the tape. Stock -64.4% from 52w high, -53% over 1 year, Fwd P/E 7.5x (cheapest in peer set ex-DAVA), EV/EBITDA 6.3x (-25% vs peer median). Mean analyst target $70 implies +40% upside on 23 analysts still rated "buy". Shorting at these levels is shorting post-capitulation — most of the remaining alpha is timing of the next guide cut, not direction.
2. Guide floor + AI Pods optionality. FY26 guide is already +0.2-2.2% — the bar is so low that a mere flat-rate macro is enough to beat. If AI Pods actually reach $60-100M exit ARR (management's target), that alone contributes ~3-4% growth to 2027. Pricing has room to recover if macro stabilizes — CFO Urthiague said pricing will be "okay," just not growing. Institutional ownership at 113% (float re-counted via short-lending) signals believers are still larger than doubters.
3. Take-out optionality at $2.2B market cap. GLOB is small enough that a strategic acquirer (WIT, TCS, CTSH, ACN) could buy the AI Pods + LATAM talent base at a modest premium. Management holds significant stake (insider ownership 2.7%) and Migoya remains CEO — not the profile of a company at the end of its independent life, but the size profile makes take-out risk real and caps downside for short positions.
4. CEO-level denial is the industry norm — GLOB may simply be the mid-cap beta expression of a cyclical trough. If ACN's record $22.1B bookings and INFY's twice-raised FY26 guide prove correct, the entire AI-disruption thesis for IT services is simply a cyclical demand dip mislabeled. In that scenario, GLOB rerates first (deepest discount) and the short loses money fastest.
Source documents
Materials saved locally and pulled from SEC EDGAR (Globant S.A., CIK 1557860 — Luxembourg-incorporated, files 20-F + 6-K).
Single-synthesis deep-dive (no 12-agent framework) building on EPAM's 6-peer cross-read — GLOB was flagged as the most relevant second-order short in epam_agents/peer_GLOB_crossread.md. All transcript quotes cite line numbers in that file or the raw transcripts in peer_sources/GLOB_transcript_Q3_2025.txt / GLOB_transcript_Q4_2025.txt. Priced-in data pulled 2026-04-18 via yfinance (short interest rate-limited initially, resolved on retry; FMP analyst estimates rate-limited — 429 response, not fatal since yfinance supplies the fields). SEC EDGAR filings via CIK 1557860 (corrected from initial 1571123, which is Science Applications Intl). Options chain is thin: only 4 expirations Nov 2026 and earlier — no Jan 2027 LEAPS exist on GLOB as of pull date. Generated 2026-04-18.