IT Services Rout: EPAM, Cognizant, IBM Slide on AI Fears

Accenture’s (NYSE: ACN) Q3 FY26 earnings release on June 18, 2026 did more than knock its own stock down 17.97%. It triggered the worst single day for the broader IT services group in years. EPAM Systems (NYSE: EPAM) closed at $76.64, down 12.61% and within touching distance of a 52-week low. Cognizant (NASDAQ: CTSH) lost 10.49%. IBM (NYSE: IBM) — a company most investors no longer file under “IT services” — fell 5.05% to $249.10, with its Consulting segment squarely in the line of fire.

The proximate cause was Accenture’s revised full-year outlook. The new FY26 revenue growth range is 3% to 4% in local currency, narrowing from a previously broader band. Strip out the U.S. federal headwind — about 1 percentage point — and the ex-federal range is 4% to 5%. That is the lowest full-year growth Accenture has guided to since the post-pandemic reset, and on a $69.7 billion FY25 revenue base it implies the consulting cycle has flattened, not just slowed.

What the tape actually did

The selloff was not subtle. Volumes were 2-4x recent averages for every name in the group. Investors did not appear to be repricing federal exposure — most of these companies have minimal U.S. federal business. They were repricing the structural margin of the IT services model itself.

Company Ticker Close (Jun 18) 1-Day % Volume
Accenture ACN $249.50* -17.97% 41.7M
EPAM Systems EPAM $76.64 -12.61% 5.6M
Cognizant CTSH ~$66* -10.49% 61.3M
IBM IBM $249.10 -5.05% 16.6M
Source: Investing.com market data for June 18, 2026 close. *ACN and CTSH prices intraday/late-tape; percent moves from same source.
IT services sector one-day decline, June 18, 2026 Bar chart showing the percent decline for Accenture, EPAM, Cognizant, and IBM on June 18, 2026. 0% -5% -10% -15% -20% -17.97% ACN -12.61% EPAM -10.49% CTSH -5.05% IBM One-day percent change, June 18, 2026
Source: Investing.com, intraday and closing data for June 18, 2026.

Why the contagion ran beyond Accenture

If the federal headwind were the whole story, EPAM should not have moved 12.6%. EPAM books almost none of its revenue from U.S. federal contracts; its exposure is global commercial enterprises, particularly financial services and travel. The same is true of Cognizant, where North American banking, healthcare, and retail are the dominant verticals.

What the market punished was a thesis it has been quietly carrying for two years and could no longer deny: enterprise demand for staff-augmentation-style consulting is decelerating, and generative AI is showing up as a deflationary force on billable-hour businesses. Accenture has been the bull-case stalwart for this group precisely because of its scale, GenAI bookings momentum, and ability to pivot. When even Accenture trims, the read-through to smaller, less diversified peers is mechanical.

EPAM offers the cleanest example. The company beat Q1 2026 EPS at $2.86 versus a $2.75 consensus, and management guided to roughly $600 million of AI-related revenue for the full year. The stock still fell on that print because investors looked past the AI line and focused on negative operating and free cash flow trends. Wedbush initiated coverage at Neutral. Jefferies downgraded to Hold. By June 18, 2026, the stock had given back the entire post-earnings bounce and then some.

IBM’s consulting tax

IBM is the awkward member of this group. Investors increasingly value IBM as a software and infrastructure story — Red Hat, watsonx, mainframe — and treat Consulting as a low-multiple drag on the sum-of-the-parts. The 5.05% decline on June 18 was the market reminding everyone that Consulting is still a meaningful percent of revenue and that any softness there compresses the blended multiple.

IBM trades at a roughly 22 P/E with a 2.58% dividend yield and a $234 billion market cap. None of those numbers, in isolation, looked expensive coming into June. But IBM has now fallen 11% over the trailing 12 months, and the gap to the consensus 12-month price target of about $291 has widened. The math works only if Software and Infrastructure carry the load through the back half of 2026.

The structural question the chart is asking

For the better part of a decade, the IT services group earned premium multiples for predictable mid-single-digit growth, stable margins, and reliable buybacks. The bear case for 2026 is that all three legs are wobbling at once:

  • Growth. If Accenture’s ex-federal range is 4-5% and that is the best-positioned name, the peer-set ceiling is lower than it was in 2024.
  • Margins. GenAI productivity gains accrue to the buyer of services before the seller. Fixed-fee work shrinks; time-and-materials work shortens.
  • Capital return. Buybacks remain a support, but multiple compression eats the EPS lift.

The bull rebuttal is real. Accenture’s GenAI bookings have crossed multi-billion-dollar annual run rates. EPAM’s $600 million AI revenue target, if hit, is not a rounding error on a sub-$2 billion quarterly base. Cognizant has been guiding to accelerating top line on the back of its Belcan acquisition and engineering-services build-out. None of that thesis is broken. It is, however, on probation.

What to watch from here

Three near-term tells will determine whether June 18 was a flash bottom or the start of a re-rating:

  • Cognizant’s Q2 print (late July). Look for fixed-fee mix, AI-related revenue disclosure, and any commentary on price compression in financial services.
  • EPAM’s Q2 print (early August). Cash flow conversion is the single number that matters. If FCF turns positive, the multiple re-rates.
  • IBM’s Q2 print (late July). Consulting bookings and Red Hat ARR growth need to move in opposite directions for the sum-of-the-parts thesis to hold.

For now, the tape has spoken: the market is no longer willing to pay for “AI as upside optionality” inside the IT services group. It wants to see AI as cash. Until at least one of these names delivers, the multiple compression that arrived on June 18 is likely to stick.

Sources

Disclosure: This article was produced with AI assistance and reviewed before publication. It is for informational purposes only and is not investment advice.

Leave a Comment