Persistent Systems Is Buying Its Own Margin — And Asking Shareholders Not to Look at the EPS
A 26th straight sequential growth quarter, a $650M reinvention deal that costs 90bps a quarter, and a 94%-subscribed Nagarro bid that still hasn't closed.
PERSISTENT.NS · Earnings Call · 2026-10-15
The quarter that looked great and printed ugly
Persistent Systems did what it has done for six and a half years: it grew. Q2 FY2027 revenue came in at $487.6 million, up 7.8% quarter-on-quarter and 20% year-on-year, and CEO Sandeep Kalra opened the call by reminding everyone this is “our 26th sequential quarter of revenue growth” — Sandeep Kalra, Executive Director and Chief Executive Officer (CEO) · 2026-10-15. In rupee terms the top line grew 30.3% year-on-year. Order book, cash, collections — all healthy on the surface.
Then you get to the bottom of the P&L. Reported PAT fell 25.4% year-on-year to a 7.5% margin, and reported EPS dropped 26%. The cause is a single, self-inflicted line item: roughly INR 1,853 million of one-time transaction costs tied to the Nagarro acquisition, which knocked about 3.1 percentage points off reported PAT. That is not deterioration in the business — but it is deterioration in the *report* the shareholder actually receives, and management knows it.
The real story: Persistent is quietly changing what business it is in
The single most important new keyword this quarter is Software Reinvention Services, a newly packaged offering that underpins the monster large deal signed last quarter — a 6.5-year, $650M-plus engagement with one of the largest technology companies. Read the mechanics carefully, because they describe a different company than the one that sold engineering hours. Persistent is taking over the end-to-end engineering roadmap of a portfolio of 20 software products, including the associated third-party licenses and cloud spend. It becomes, in effect, the P&L owner of someone else's product economics.
That is a genuinely new shape. The commercial construct is not time-and-materials; per Kalra, “It is a revenue share kind of a construct where as the customer sells the product more, we get a certain percentage of revenue with a certain backstop.” — Sandeep Kalra, Executive Director and Chief Executive Officer (CEO) · 2026-10-15 That is a software-adjacent risk profile bolted onto a services firm.
The tell-tale is on the balance sheet and in the margin walk. CFO Vinit Teredesai flagged that “This investment in the third-party tools and licenses led to an increase in the quarterly amortization of approximately $4 million, which had an impact of 90 basis point on our Q2 EBIT margin.” — Vinit Teredesai, Executive Director and Chief Financial Officer (CFO) · 2026-10-15 Intangibles rose roughly INR 900 crore quarter-on-quarter, and the newly-prominent intangible assets under development line is now a visible slice of spending. EBIT margin landed at 15%, down 130bps year-on-year; wage hikes alone cost 180bps, offset partly by lower stock cost.
This is the central tension of the call. Growth is accelerating because Persistent is absorbing cost — licenses, cloud — that used to sit on the client's books. Revenue is real; so is the drag. Analysts pressed hard on whether it's a margin recovery story or a structurally lower-margin business, and Kalra's answer was a plea for portfolio thinking:
Please look at that just as you would look at it as investors at a portfolio of your own investments. Not everything pans out every month at the same clip.
The stated established products thesis is that squeezing cost out of aging software funds the client's AI future — and Persistent takes a cut of the savings.
The things management stopped emphasizing
Two subtler shifts matter. First, new deal win metrics softened markedly: ACV of $481 million and total contract value of $633.5 million, versus $536.8 million and $1,146.2 million in the prior quarter. Management immediately reframed to trailing-twelve-month comparisons (TCV up 36.3%, ACV up 21.9%), but admitted the underlying issue: “We had a few deals that slipped over into this quarter. No excuses... The slowdown of decision-making in the last few weeks was pretty apparent.” — Sandeep Kalra, Executive Director and Chief Executive Officer (CEO) · 2026-10-15 Macro fragility, not productivity, was the culprit — an important distinction for anyone modeling the sector.
Second, the analyst community is not fully buying the "new offering" framing. Nitin Padmanabhan pointed out flatly that “the Software Reinvention Services which we have launched, we had done this prior to 2019 as well.” — Nitin Padmanabhan, Analyst · 2026-10-15 Persistent ran a product business before. Kalra's defense — that AI now lets you map a product's entire release history, re-architect for cheaper cloud consumption, and replace third-party code with custom agentic components — is a technology answer to a business-model question. It may be right. It has not yet been proven at scale.
Where this sits in the wider market
Persistent is riding a wave the market is still pricing as expensive. The token cost theme — enterprise AI consumption becoming a runaway cost line — is now a genuine sales door for services firms, and Persistent leans in with FinOps-style token routing and its SASVA platform, which supports open weight models alongside frontier providers so clients can push cheap tasks off expensive models. That is exactly the same economic logic the tape has been rewarding in the broader AI supply chain — and punishing when it stalls: over the last 30 and 90 days, the market's AI data centers cluster is a broad-based decliner, with dozens of negative tickers across semis and hardware. Persistent's own pitch is essentially that enterprise AI eats the *application* layer first, and that high token cost is what drives customers toward firms that can optimize it.
Management also concedes who is on the bleeding edge. Kalra was explicit that hi-tech, not BFSI, is the true early adopter: “within the hi-tech part as well, 20% of the ISVs, I would say mid-market ISVs, are on the bleeding edge.” — Sandeep Kalra, Executive Director and Chief Executive Officer (CEO) · 2026-10-15 Persistent's concentration in Hi tech — which grew 28.5% year-on-year this quarter — is therefore both an accelerant and a single-point of failure. Europe revenue fell 8.5%, blamed on two client-specific events rather than demand, and healthcare life sciences has decelerated from the 50-70% growth prints of FY2025 to 12%.
The governance footnote nobody asked hard about
Buried in the prepared remarks is the most quietly newsworthy line of the quarter: ESOP vesting is conditional on 12% revenue and 12% EPS growth, and Kalra noted the board *may seek shareholder approval to treat one-time transaction expenses as an exception* when assessing vesting. That is a company asking its owners to ignore a 26% EPS decline for compensation purposes while simultaneously telling them the margin dilution is only temporary and aspirational margins of 16%-17% lie ahead — “We do not give forward-looking guidance. Please take this as an aspirational goal.” — Sandeep Kalra, Executive Director and Chief Executive Officer (CEO) · 2026-10-15
The counterweight is genuine operational quality: billed DSO collapsed from 61 to 50 days, record collections pushed cash to ~$315 million, and trailing OCF-to-PAT rose from 76.2% to 96.6%. Persistent is funding a transformation from its own working capital. Whether the reinvention deal becomes a moat or a margin sink is the whole question — and it won't be settled until Nagarro closes and the first full year of the revenue-share economics prints.