EPAM Systems: AI Is Eating The Old Business, But 8.7x Earnings Is Too Cheap

EPAM Systems (EPAM) customers are cutting back on manual testing, user-experience work and other task-based services because of AI. These are just the areas investors were worried about, and these are the areas it is getting hit. This work can be replaced by
EPAM Systems: AI Is Eating The Old Business, But 8.7x Earnings Is Too Cheap
Summary
- EPAM Systems faces near-term revenue headwinds as AI cannibalizes legacy services, but trades at just 8.7x forward non-GAAP earnings.
- AI-native revenue now exceeds 11% of EPAM, growing double-digit sequentially, while margins and adjusted EPS are improving despite flat topline growth.
- Engine Capital's activist push for a $750M buyback leverages EPAM's strong balance sheet and could materially shrink the share count at current valuations.
- I rate EPAM a Buy: the stock is priced for too much permanence in the current transition, with upside if AI-native work fills the legacy gap by 2027.
This article was written by
Analyst’s Disclosure: I/we have no stock, option or similar derivative position in any of the companies mentioned, and no plans to initiate any such positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

Facts Only

* EPAM customers are cutting back on manual testing, user-experience work, and other task-based services due to AI.
* AI-native revenue now exceeds 11% of EPAM.
* AI-native revenue is growing double-digit sequentially.
* Margins and adjusted EPS are improving despite flat topline growth.
* Engine Capital is pushing for a $750M buyback.
* The stock is currently trading at 8.7x forward non-GAAP earnings.
* A rating of Buy was issued, contingent on AI-native work filling the legacy gap by 2027.

Executive Summary

EPAM Systems is experiencing near-term revenue challenges as artificial intelligence reduces demand for manual testing, user-experience work, and other task-based services. This shift means that the areas investors were concerned about are currently being impacted by AI cannibalization of legacy services. Despite this revenue pressure, AI-native revenue now constitutes over 11% of EPAM's total, growing sequentially at a double-digit rate. Furthermore, margins and adjusted earnings per share are showing improvement even as topline growth remains flat. Engine Capital is pushing for a $750 million buyback to leverage the company's balance sheet. An analyst rates the stock a Buy, suggesting that there is potential upside if AI-native work successfully fills the gap in legacy services by 2027.

Full Take

The narrative presents a tension between immediate revenue headwinds from AI disruption in established service lines and underlying operational improvements driven by newer AI-native segments. The core dynamic involves assessing whether the transition phase—where legacy business declines while new AI-focused work grows—is adequately priced into the current valuation, as suggested by the analyst's view that the stock is valued for too much permanence. The implication centers on the relative velocity of adoption in AI services versus the inertia within legacy contracts. The pattern suggests that investor focus shifts rapidly toward emergent revenue streams (AI-native) while the underlying profitability metrics are demonstrating resilience despite flat overall growth, suggesting a potential decoupling between top-line performance and operational health. A key unanswered question is whether the projected filling rate of the legacy gap by 2027 materializes; if AI adoption accelerates faster than anticipated, or if service transitions stall, the current valuation predicated on a slow transition could prove tenuous. What factors influence the speed at which demand for traditional services erodes compared to the growth trajectory of specialized AI offerings? What is the actual cost structure shift resulting from shifting work to AI-native models versus legacy delivery methods?

Sentinel — Human

Confidence

This text appears to be a personal investment analysis written by an individual analyst, blending factual reporting on company performance with subjective forward-looking commentary.

Signals Detected

Sentence length variance is natural; exhibits variation between complex analytical statements and punchy summary points.

The argument flows logically from problem identification (AI impact) to specific financial metrics, activist strategy, and a forward-looking investment thesis.

Uses clear bullet points for summary mixed with flowing prose; attribution is personal ('I wrote this article myself').

The content relies heavily on specific financial data and clearly stated analyst opinions, typical of investment commentary.

Human Indicators

Explicit author disclosure states the writer personally created the article and expresses personal opinions, suggesting an individual investment perspective.

The tone mixes objective-sounding data points with subjective investment ratings ('I rate EPAM a Buy').