Operating Leverage Holds Through the Float Headwind; PEO Hiring Still Lags

Results vs. Consensus

Q3 FY2026 scorecard

The March quarter beat on both revenue and adjusted earnings. More consequentially, it cleared the specific hurdle in our Q2 case: operating leverage continued through the quarter management had identified as vulnerable to lower short-term interest rates.

Year-over-year financial comparison

Sequential comparison

The large sequential increase reflects the March quarter’s payroll-tax and bonus season. The investment test is year-over-year conversion, rather than extrapolating the 30.2% quarterly margin into the full year.

Nine-month performance

Revenue: Reported growth of 7% included about one point of currency benefit; organic constant-currency growth was 6%. Employer Services remained at 5% organic constant-currency growth, so the faster consolidated print is not evidence of a broad acceleration in underlying ES demand. The favorable change is that retention, employment and price now support the outlook together.

Margins: Employer Services produced $158.0M of the $159.9M increase in consolidated pretax earnings. Its 130-basis-point margin gain absorbed the PEO decline. The ex-float comparison is a useful earnings-quality test, although it still includes other interest and nonoperating items. It supports stronger payroll economics without assigning every dollar of improvement to AI.

EPS: Adjusted net earnings grew 8.7%, while diluted shares fell 1.5%; buybacks lifted per-share growth to 10.1%. Tax worked against the quarter, with the adjusted effective rate rising to 23.7% from 23.0%. GAAP EPS of $3.38 exceeds adjusted EPS by one cent, principally because of a venture-investment gain. The beat is not dependent on large exclusions or a favorable tax rate.

Segment Performance

Employer Services: the margin recovery broadens

Q2 showed that the WorkForce Software anniversary could remove a drag. Q3 is more useful evidence of continuing execution: ES revenue grew 7.1%, earnings rose 10.5%, and margin expanded 130 basis points. Foreign exchange contributed nearly two points of reported growth, while the client-funds interest line increased $48.7M. Both help the result, but neither alone explains the $158.0M earnings increase.

Retention and client satisfaction reached records for a third quarter, and pricing improved without an evident retention penalty. This is the economic connection between the service moat and the profit statement: fewer avoidable client contacts lower delivery costs, while satisfied clients accept more value-based pricing.

Assessment: The retention and satisfaction pillar remains ON TRACK. We raise our ES growth assumption to 6.5%, but keep organic growth expectations in the mid-single digits. Better conversion of that growth supports the earnings revision; a currency-assisted 7% print does not justify assuming a new sales growth regime.

PEO Services: better sales, weaker incremental economics

Revenue excluding zero-margin benefits pass-throughs increased to $736.0M from $698.5M, up 5.4%, versus 3.4% in Q2. Stronger new business bookings helped keep average worksite employees growing about 2%, despite further softening in employment within existing clients. Gross payroll and state unemployment insurance revenue also supported growth. The prior quarter’s concern about another step down in service revenue did not materialize.

Earnings nevertheless fell $6.1M. Benefits pass-through costs rose to $1,170.0M from $1,090.0M; stronger lower-margin state unemployment revenue, selling expense and smaller favorable workers’ compensation reserve releases also weighed. PEO profit as a share of revenue excluding benefits pass-throughs fell to 33.6% from 36.3%. The pressure therefore extends beyond dilution from zero-margin revenue.

Assessment: Bookings are doing the work that existing-client hiring is not. We increase full-year PEO revenue growth excluding pass-throughs to 4.5% from 3.9%, but retain the volume risk at EMERGING. The unchanged 2% worksite-employee guide and declining profit prevent a clean recovery judgment.

Other: smaller loss, limited incremental help

The Other loss narrowed $8.0M to $123.5M. Corporate extended interest income increased $8.9M to $43.2M, while related financing expense rose $0.5M to $46.0M. Unlike Q2, lower financing rates did not translate into lower dollar financing expense, because average financing balances increased. Most of the consolidated profit improvement came through ES.

Assessment: The float structure continues to cushion lower overnight yields, but it is no longer the dominant explanation for earnings growth. The distinction strengthens the case for valuing ADP on recurring service earnings rather than treating the entire growth rate as a rate-cycle windfall.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: Management was more assertive about AI and its competitive position than in January, supported by stronger retention and margin delivery. It remained cautious about the final-quarter bookings range and did not convert early FY27 confidence into numerical guidance.

1. Pricing improves without sacrificing retention

January’s outlook embedded around 100 basis points of price contribution. Management now expects about 130, alongside a 10-basis-point improvement in the retention guide and a pays-per-control outlook raised to about 1%. The combination matters more than any single lever: better retention preserves the recurring base, price increases its yield, and modest employment growth adds volume.

“Both our employer services retention rate and our overall client satisfaction levels reached new record highs for a third quarter.”
— Maria Black, President and CEO

Assessment: The earlier moat hypothesis has produced a stronger financial outcome this quarter. We give pricing more weight than the currency benefit when assessing durability, while recognizing that retention at an already high level has less room to improve indefinitely.

2. AI productivity moves from examples toward service economics

An 8% reduction in small-business client contacts during the busiest quarter is a meaningful advance on January’s unquantified cost-to-serve argument. Management also described an Indian compliance workflow with approximately 35% less labor and a service-platform rollout covering 20% of the service population, targeted above 40% by fiscal year-end. The Indian result is a specific process example, not a group-wide labor saving.

“The margin expansion we achieved reflects disciplined investment. We are funding our AI transformation across our products, internal tools and service delivery while continuing to deliver on our financial commitments.”
— Peter Hadley, CFO

The economic opportunity is lower service effort per client without lowering accuracy or satisfaction. ADP can retain part of those savings in EBIT and reinvest part in product development and distribution. R&D expense alone will not measure the investment: spending also sits in operations and capitalized software, while nine-month additions to intangibles increased 27.3%.

Assessment: We lift the full-year adjusted EBIT margin assumption to 26.75% from 26.55%. That matches the revised company framework and observed productivity direction; it does not require attributing all 130 basis points of ES improvement to AI. The next test is whether broader rollout produces further service efficiency while retention remains strong.

3. Lyric’s opportunity expands, but bookings still have to close

Lyric’s position-based architecture allows flexible payroll and workforce configurations, and management described combining global payroll, time management and service. Internal implementation investment and system-integrator relationships can expand deployment capacity. That addresses the practical constraint behind January’s large enterprise wins: selling the platform is only the first step toward recurring revenue.

“So we are scaling internally. But we also have this ability to go to market with system integrators in a more meaningful way than we have in the past.”
— Maria Black, President and CEO

The broader ES bookings guide remains 4% to 7%. Management indicated roughly half of bookings come from new logos and half from additional business, and that each percentage point of annual growth represents about $20M to $21M. The three-point range is therefore around $60M of sales execution, small enough for late deals to change the outcome and large enough to matter for the following year.

Assessment: The product cycle remains ON TRACK, with a stronger route to implementation. We continue to treat Lyric as a source of future growth within the existing franchise, rather than add a separate revenue stream to the forecast before bookings dollars and conversion timing are disclosed.

4. The expected short-rate headwind arrived; the ladder absorbed it

The client-short yield fell to 3.1% from 4.2%, while client-extended yield rose to 3.5% from 2.7% and client-long yield to 3.3% from 3.1%. Balances grew to $48.3B, and net strategy income increased 16.6%. January correctly identified the seasonal short-rate exposure, but stronger balances and reinvestment at higher yields prevented it from derailing margin expansion.

“Client funds interest revenue increased by more than we anticipated in Q3, driven by 9% growth in our average client funds balances.”
— Peter Hadley, CFO

The full-year interest-revenue and net-strategy midpoints each rise $25M. The forward new-purchase rate used in the outlook increased to 3.9% from January’s 3.8%, reversing the earlier sequence of declines. Management expects portfolio yields to rise again in FY27 as maturing securities are reinvested, but has not supplied a full-year dollar forecast.

Assessment: Our FY26 net-strategy estimate rises to $1,305M. The 36% share of incremental pretax earnings keeps float dependence CONTAINED; the business is earning through the headwind. FY27 reinvestment remains supportive, with wage-driven balances and the rate curve still capable of changing the result.

5. Cash conversion catches up, with tax timing doing part of the work

Nine-month operating cash flow of $4,013.2M rose 14.6%, reversing the first-half decline of 10.1%. Derived Q3 cash flow was $2,238.0M versus $1,525.8M a year ago. That meets the prior recap’s explicit requirement for the March quarter to close the cumulative cash gap.

The improvement is not all a new recurring cash run rate. Nine-month cash taxes paid fell $374.8M, equivalent to 73% of the $512.7M net increase in operating cash flow. After $123.8M of capital spending and $352.4M of intangible additions, cash generation was $3,537.0M, covering $3,406.3M of dividends and repurchases. Cash held for clients is matched against client obligations and is not shareholder liquidity.

“We expect to continue share repurchases at or above these elevated levels across the balance of this year and throughout fiscal 2027, absent major changes in the market backdrop.”
— Peter Hadley, CFO

Assessment: The capital-return pillar stays ON TRACK. Q3 buybacks of $581.4M exceed Q2’s $515.8M, and the business now funds cumulative distributions after both physical and intangible investment. Tax timing limits how much of the cash acceleration we extrapolate; the buyback commitment is explicitly conditional on the market backdrop.

6. AI changes the revenue debate as well as the cost base

Management’s defense of HCM demand is economically plausible: correct payment, regulatory compliance and auditability remain necessary as workforce structures change. Yet a payroll vendor can remain essential while earning less from a client that employs fewer people. The more important question is how ADP gets paid for that continuing value.

“We are not immune to shifting employment trends, but we are built for the world they represent.”
— Maria Black, President and CEO

The call clarified a mixed model: small-business revenue relies heavily on base fees, mid-market and enterprise are more seat-based, and PEO billing is predominantly tied to a percentage of payroll. Wage growth can partly offset headcount pressure in the PEO, while larger-client ES revenue remains exposed to employment. These are buffers, not immunity.

Assessment: The raised pays-per-control guide challenges the immediate labor bear case. It does not settle the longer-term AI question. We retain the prior base valuation multiple, with no separate AI premium, and use a lower downside multiple to allow for a market that discounts a weaker future employment base even before current EPS falls sharply.

Guidance & Outlook

Implied Q4: Off the $20,561M FY25 revenue base, the revised range implies roughly $5,321M to $5,527M in Q4, or 3.8% to 7.8% growth. The $10.01 prior adjusted EPS base implies approximately $11.01 to $11.11 for FY26 and $2.53 to $2.63 for Q4 after $8.48 earned through March. That is about 12% to 16% growth against the prior Q4’s $2.26; quarterly subtraction is approximate because weighted shares differ.

The remaining margin requirement still matters. Pairing the revenue and margin endpoints implies roughly 24.7% to 25.2% adjusted EBIT margin in Q4 versus about 23.7% a year ago, or approximately 109 to 156 basis points of expansion. Our forecast is near 25.0%, about 130 basis points higher. Q3’s 80-basis-point gain strengthens confidence, but does not by itself satisfy that remaining hurdle.

Street comparison: Pre-release FY26 expectations were about $21.80B of revenue and $10.97 of adjusted EPS. The new guide midpoint is about $21.90B and $11.06, a modest estimate lift. The larger stock reaction therefore reflects a change in confidence as well as higher current-year earnings.

Guidance judgment: The raised price and retention assumptions make the revenue outlook broader than another float-only raise. We regard the midpoint as achievable, while keeping the 4% to 7% bookings range and softer PEO hiring as genuine constraints. Early FY27 remarks signal faster productivity gains, higher portfolio yields and continued buybacks; they are not formal numerical FY27 guidance.

Analyst Q&A Highlights

What sits behind the raised outlook

The question tested whether balances and retention explained most of the increase. Management added a previously unquantified improvement in price, which gives the service business a more durable earnings lever.

Q: “Can we go through quickly the attribution of what's driving the change the outlook? It seems like it's higher balances, some improvement in retention is the majority of it? Did I catch that?”
— Tien-Tsin Huang, JPMorgan

A: “So the last couple of quarters, we've been talking about looking our goal to achieve around 100 basis points of contribution from price, I'm pleased that our outlook actually is reflected a little increase in that. So expecting more like 130 basis points from price.”
— Peter Hadley, CFO

Assessment: Price is an additional lever, not proof that the entire outlook raise comes from service execution. The higher balance assumption still contributes $25M at the net-strategy midpoint; the important change is that several drivers now reinforce one another.

Why a wide bookings range remains with one quarter left

The challenge was whether management could narrow the annual range after three quarters. The answer acknowledged remaining execution and explained why relatively few dollars can shift the percentage result.

Q: “You didn't change the forecast range for the year, and it's still relatively wide with 1 quarter to go. Can you just discuss a little bit about what you're seeing with regards to the bookings in the third quarter and year-to-date?”
— Mark Marcon, Baird

A: “I would say, as we head into the fourth quarter, there's always a lot to get done. We left the range relatively wide, as you mentioned. I think all of those options are outcomes for us. The sensitivity of it, if you will, is around $20 million, $21 million per percent.”
— Maria Black, President and CEO

Assessment: This is an explanation of sales timing, not a disclosure of year-to-date growth. The product pillar survives, but confidence in FY27 acceleration should depend on the eventual bookings result and conversion, rather than treating unchanged guidance as evidence that the high end is likely.

Whether higher PEO selling expense signals distress or growth

The question distinguished concessions from sales compensation. Management linked the expense to stronger sales, while also disclosing weaker existing-client hiring and smaller favorable reserve releases elsewhere in the answer.

Q: “The PEO segment margins came in a little bit below our model. And you mentioned a few drivers. I think one of them was higher selling expenses. What does that mean exactly in this context? Is it like concessions to new clients or higher incentive for your sales staff? Just trying to figure out sort of what that is and what it implies.”
— Ramsey El-Assal, Cantor Fitzgerald

A: “But back to the selling expenses, the real reason why the selling expenses were higher was, we had a really strong quarter in terms of sales. So again, that creates -- there's a variable cost model with respect to selling, and we had a strong quarter, as Maria alluded to earlier with respect to PO sales, so that brought additional selling expenses […]”
— Peter Hadley, CFO

Assessment: A sales-driven cost increase is healthier than discounting to defend the base. It still depresses current profit, and the wider answer does not remove the lost contribution from softening employment. That combination supports a higher revenue estimate with continued caution on PEO margin.

How PEO billing responds to headcount and wages

The PEO question tested how closely revenue tracks client headcount. Management described payroll-percentage billing, in contrast with the more seat-based mid-market and enterprise ES models discussed earlier in the same exchange.

Q: “And then just on the PEO side, I think the revenues are more tied to client head count there. But maybe if you can clarify all that with some numbers, that would be really helpful.”
— Jason Kupferberg, Wells Fargo

A: “The predominant billing model we have in the PEO is the percentage of payroll. So obviously, the number of employees can influence the percentage of payroll that so can wage levels, wage inflation, obviously, some of the pass-through revenues, like taxes and workers' comp and things like that.”
— Peter Hadley, CFO

Assessment: The PEO has wage exposure as well as headcount exposure, and management separately described mid-market and enterprise ES as more seat-based. The answer offers real diversification, but no consolidated percentage that would justify calling ADP insulated from AI-related employment change.

How client AI savings become revenue or margin

The question combined adoption and repeat usage with the allocation of value from time savings. Management described embedded value capture rather than a charge for each agent interaction.

Q: “And then the second part of the question is you commented about the payroll agent saving a lot of time, smart actions saving a lot of time. As you roll more of these out, how do you view out sharing some of the value from the time savings that these agents are providing?”
— Daniel Jester, BMO Capital Markets

A: “So we're not really looking at this as a discrete usage type of fee at a piece by piece level. We really look at it as core in the fabric of how we operate and how we deploy our products to our clients. And I think it shows up in things like margin and efficiency. I think it shows up in bookings. I think it shows up in retention, and that's kind of the way we think about sharing this opportunity with our clients.”
— Maria Black, President and CEO

Assessment: The answer supports the existing service and software model rather than a new agent-fee revenue forecast. It does not supply adoption rates or isolate AI-driven pricing. Our forecast captures productivity within the ordinary margin framework and leaves an additional monetization upside unpriced.

What They’re NOT Saying

1. The quarterly ES bookings growth rate and Lyric revenue conversion. The annual range remains intact, but the missing sales dollars and deployment timing limit how precisely current momentum can be carried into FY27.
2. Whether PEO existing-client employment is near a trough. New sales are offsetting softening pays per control. A stronger bookings quarter does not establish when higher-margin hiring revenue resumes.
3. The consolidated financial contribution from AI. Service contact reductions and client time-saving examples are useful; active-user penetration, group savings and incremental price attributable specifically to AI remain undisclosed.
4. A new Next-Gen migration or CashFlow Central monetization milestone. January’s product commitments remain part of the long-term case, but this call provides no material new revenue measurement for either.
5. A numerical FY27 plan or rate-shock sensitivity. Management described favorable reinvestment and productivity, without sizing the net effect against employment, currency and investment needs.

Market Reaction

• Pre-print: ADP closed April 28 at $199.17, down 22.6% year to date, 32.7% over twelve months and 3.1% over thirty days. The pre-print 52-week closing range was $188.79 to $326.81; the S&P 500 was up 4.3% year to date.
• April 29 reaction session: The stock opened at $207.78, traded between $205.06 and $215.17, and closed at $215.06, up $15.89 or 8.0%. The S&P 500 was essentially unchanged.
• Volume: About 4.4 million shares versus a 3.6 million thirty-day average, approximately 1.3 times normal.

The beat, higher outlook and stronger service economics give the positive reaction a company-specific foundation. After the severe pre-print decline, the result also challenged the idea that ADP’s payroll franchise must deteriorate as AI adoption broadens. The data do not identify how much of the move was short covering or a change in long-term investor positioning.

Despite the rally, the close remains 14.2% below the $250.69 reaction price at our January upgrade. The estimate has risen while the price has fallen. That does not prove mispricing, but it creates a substantially larger valuation cushion for the same service-productivity thesis.

Street Perspective

Debate: durable productivity or another float-assisted beat?

Bull view: Rising retention and pricing, lower service contacts and a stronger margin guide suggest that product investment is improving the economics of the installed base.

Bear view: The float still grew faster than processing revenue, currency helped reported growth, and AI savings are not separately measured. Some improvement may be cyclical rather than structural.

Our take: The bulls have the stronger current evidence because profit excluding the client-funds strategy grew 8.0% and the float’s contribution to incremental profit declined again. We credit demonstrated operating leverage, while retaining a conventional earnings multiple that does not require a separately monetized AI business.

Debate: does the PEO rebound settle the employment concern?

Bull view: Stronger bookings and service revenue growth show that the January weakness was manageable; the raised revenue range argues against a broad demand break.

Bear view: Existing-client hiring continued to soften, average worksite-employee guidance did not improve and profit fell. Sales-led expansion can cost more to acquire than growth within existing accounts.

Our take: The revenue recovery is real, but the bears retain a valid incremental-profit argument. We raise PEO sales assumptions and keep the employment risk open; the stronger ES result is carrying the consolidated upgrade.

Debate: can a payroll franchise defend value as AI changes headcount?

Bull view: Compliance, accurate payment and trust become more valuable in a complex workforce, while base fees and payroll-value billing diversify exposure.

Bear view: Essential infrastructure can still face slower revenue if clients reduce employees, and new competitors may force productivity savings to be shared through price.

Our take: The raised price contribution and retention outlook contradict near-term erosion. They do not eliminate longer-term labor or competitive risk, which is why our downside case allows substantial multiple compression even with only a modest current-year EPS miss.

Our Estimates & Valuation

Earnings bridge: Revenue of $21,900M at a 26.75% adjusted EBIT margin gives about $5,858M of EBIT. Deducting an assumed $55M of net interest expense outside the client-funds strategy, applying a 23% tax rate and dividing by 404M diluted shares produces $11.06 of adjusted EPS. The assumptions imply Q4 revenue of $5,426M, margin near 25.0% and EPS around $2.58. This is a midpoint outcome with meaningful remaining margin execution, not an extrapolation of March seasonality.

Valuation continuity: The base target increases to $277 from $274 using the same FY26 earnings period and the same 25-times multiple as January. The approximately 1.1% EPS revision explains the target change; neither a forecast roll nor an AI premium is doing hidden work. Spot is only 19.4 times our estimate. The target multiple assumes renewed confidence in recurring service growth and cash returns, supported by retention near 92% and a second quarter of meaningful margin expansion.

The bull case requires stronger bookings conversion and Q4 earnings above the new guide’s top end, supporting a 28-times multiple. The bear case assumes PEO weakness spreads, ES bookings disappoint and Q4 EPS is roughly flat at $2.27, producing FY26 EPS of $10.75. A 17-times multiple then discounts a weaker subsequent employment and earnings trajectory. We lower that downside multiple from January’s 21 times because the stock now trades below the old downside multiple even on the improved base estimate; the prior bear framework no longer describes a meaningful loss scenario.

The twelve-month dividend assumption is $6.80, holding the current quarterly $1.70 rate unchanged. Base total return is approximately 32%, versus an explicit 8% S&P 500 comparison assumption; bear total return is about -12%. At an unchanged 19.4-times multiple, the valuation re-rating would not occur, so much of the base upside depends on investors accepting that the productivity improvement persists. The combination of operating evidence and a 3.2% dividend yield nevertheless supports Outperform. A severe employment or technology disruption could produce losses beyond the modeled bear case.

Thesis Scorecard Post-Earnings

Overall: The standing thesis strengthens without a status-tag change. Q3 confirms more durable service conversion and resolves the cumulative cash shortfall, while the PEO risk narrows toward existing-client hiring and profitability rather than another revenue deceleration.

Next signposts: The Q4 margin requirement and 4% to 7% annual ES bookings target are the principal execution tests. We also watch the flat-to-down-20-basis-point retention range, about 1% pays per control, 4% to 5% PEO revenue growth excluding pass-throughs, 2% worksite-employee growth, service-platform coverage above 40% by June, and the promised conditional continuation of elevated buybacks into FY27. A PEO hiring recovery would strengthen the case; missed ES bookings together with weaker retention, or failure to reach the raised margin band, would undermine it.

Action: Maintain Outperform. January’s upgrade depended on the operating business sustaining earnings growth as float conditions normalized. It has done so through the identified difficult quarter. The stock offers materially more prospective return than at that upgrade, while the remaining PEO and employment risks fit within a credible downside case rather than invalidate the service-franchise thesis.