Everything Went Right and the Stock Fell: Taking the Gain After a 71% Run

Results vs. Consensus

Q2 2026 Scorecard

Year-Over-Year Comparisons — Continuing Operations

The expense discipline is the standout. Marketing dollars were flat year-over-year while GMS grew 7.5%, which is the definition of earning growth rather than buying it, and management said so explicitly. General and administrative fell 12.0% in absolute dollars. Product development rose only 1.6%. Total operating expenses declined 2.1% against 6.2% revenue growth, which is what produced 379bp of operating margin expansion. The one deleveraging line is cost of revenue, up 9.3%, which management attributed not to AI compute but to trust and safety, refunds and other customer-experience costs.

First-Half 2026

The first-half operating income comparison of +171.7% is distorted by the $101.7M asset impairment charge that sat in Q1 2025; the underlying improvement is real but nowhere near that scale. The genuinely informative first-half figure is operating cash flow from continuing operations of $268.4M against $168.8M, up 59.0%, which contains no such distortion. The diluted share count comparison runs the wrong way at the half-year level because H1 2025's diluted count collapsed to the basic count in a loss-making period, not because Etsy issued shares.

Quarterly figures tie to the half-year statement exactly: Q1 revenue of $631.3M plus Q2 revenue of $668.3M equals the reported $1,299.6M, and Q1 operating income of $119.8M plus Q2's $125.3M equals the reported $245.1M.

Quality of Beat

Revenue: $668.3M against a Street at $649.7M is a 2.9% beat, and against the alternative $646.3M estimate a 3.4% beat, so the direction is unambiguous this time. The composition is better than the headline: on an Etsy standalone basis, revenue grew 9.3%, with marketplace revenue up 8.4% and services revenue up 11.2%. Marketplace revenue is the unit-linked line, and 8.4% growth in it is the strongest signal in the release that this is a volume story rather than a monetization story. A year ago that line was declining.

Margins: a 29.2% adjusted EBITDA margin against a 27% to 29% guide is 20bp above the ceiling, delivered without cutting marketing and while absorbing higher AI compute costs. The CFO's framing is the useful one: nearly half of the year-over-year growth in Etsy marketplace revenue flowed through to adjusted EBITDA. Incremental margins near 50% on a business already running at 29% is the arithmetic that makes modest GMS growth valuable.

EPS: GAAP diluted EPS from continuing operations of $0.98 against $0.39 a year ago is a 151.3% increase, and roughly a third of that comes from below the line rather than from operations. Other income swung $34.9M in Etsy's favor, and the effective tax rate fell from 33.8% to 15.3%, an 1,848bp move that is by far the largest single contributor to the year-over-year comparison. Income from operations grew 33.1%; income from continuing operations grew 150.5%. Readers should anchor on the former. The Street's adjusted EPS figures of $1.31 to $1.34 against a $1.17 to $1.18 consensus are on a different basis again and are not comparable to either.

Segment Performance

Etsy is now a genuinely single-asset company. Depop closed on July 30 and Reverb is fourteen months gone. The analysis below is channel, cohort and geography.

Channel and Cohort Mix

The Most Valuable Cohorts Finally Turn

The single most important disclosure in this release is that habitual and repeat buyers each showed slight sequential growth for the first time since 2023. These are Etsy's highest-frequency, highest-value buyers, the cohort whose 8.6% year-over-year decline we identified in February as the last unimproved metric and the one the entire loyalty and personalization strategy exists to move.

"Active buyers improved, growing by 350,000 sequentially to approximately 87 million, roughly stable on a year-over-year basis. Gross buyer additions accelerated and habitual and repeat buyer cohorts, our most valuable buyers, each showed slight sequential growth for the first time since 2023." — Kruti Patel Goyal, CEO

The CFO's framing of what this means structurally is worth holding onto: gross additions are widening the top of the funnel while the most valuable cohorts grow at the back end, which is the combination that produces durable GMS rather than a one-quarter surge.

Assessment: this is the metric that had to turn for the turnaround to be more than a comparison effect, and it turned. The qualifier is that "slight sequential growth" on a cohort that is still down year-over-year is the earliest possible confirmation, and purchase frequency, which is the underlying driver, remains negative. The right read is that the leading indicators have now propagated to the lagging ones, which materially raises confidence in the 2027 growth rate without yet establishing it.

The App Keeps Compounding

App GMS grew 12.5%, accelerating from 11.2% in Q1 and 6.6% in Q4. More important than the growth rate is what sits underneath it: both visits per monthly active user and orders per visit increased year-over-year. Those are the two components of frequency at the channel level, and both are positive on the app even while blended frequency is negative. Buyer profiles now cover more than 65 million buyers.

Assessment: the app is running the experiment that the rest of the marketplace has not yet completed. If visits per user and orders per visit are both rising on the surface that carries 47% of GMS and 40% higher lifetime value, then blended frequency turning positive is a question of mix and time rather than of whether the mechanism works. The bull case for 2027 rests almost entirely on that proposition.

Geography and the Consumer

Growth was broad. Management reported GMS growth among both US and non-US buyers, with accelerated growth in the US import trade lane specifically, and good numbers across all household income segments, with the high end growing somewhat faster than lower-income households. Asked directly about fuel prices, the CFO said there was no discernible impact and that consumer demand held up well through the quarter. Management also said Etsy grew in every one of its top categories and grew faster than peer benchmarks in those categories in the US.

Assessment: the breadth is the point. A recovery concentrated in one geography, one channel or one income cohort would be fragile. This one is showing up in US and international, in domestic and import trade lanes, across income segments and across every major category. The persistent high-income skew remains the concentration risk we have flagged since initiation and it has not worsened.

Key Topics & Management Commentary

Overall Management Tone: The most confident of the four calls in our coverage, and the first in which management led with conviction rather than with caveats. The prepared remarks opened on the raised outlook and the new authorization rather than on the results, and the restructuring was framed as leaning in during a period of strength rather than as a response to pressure. Where prior calls volunteered the exogenous contributors to growth before claiming credit, this one credited execution first. Candour was reserved for the two places it still applies: purchase frequency remains below prior year, and the paid-search benefit was attributed to a named competitor withdrawing from the auction.

1. A Second Consecutive Raise, on Growth and Margin Together

Full-year GMS growth was raised from low single-digit to mid single-digit, and the adjusted EBITDA margin outlook was tightened upward from 28% to 30% to 29% to 30%. Both moved in the same direction in the same quarter, which is the rarest and most valuable form of guidance revision. Q3 GMS is guided to $2.53–2.58B, or 4% to 6% growth, with take rate approximately 26% and margin 28% to 30%.

Management summarized the arc in terms that are, for once, not overstated.

"We've inflected Etsy's year-over-year growth trajectory from high single-digit GMS declines in early 2025 to mid-single-digit growth anticipated for the full year 2026, a more than 10 percentage point improvement in performance." — Kruti Patel Goyal, CEO

Assessment: a 10-plus point swing in growth trajectory inside eighteen months, achieved while expanding margin, is a genuine corporate turnaround and management deserves the credit it claimed. The forward question is arithmetic rather than qualitative. A mid-single-digit 2026 against a Q2 that printed 7.5% and a Q3 guided to 4% to 6% implies the second half runs slower than the first, consistent with the deceleration bridge management laid out in April. The 2027 rate is what remains unknown, and it is what a 14.6 times multiple is now paying for.

2. The Depop Exit Delivers $1.4 Billion, Above the Headline

Etsy completed the sale of Depop to eBay on July 30 and received $1.4 billion in cash proceeds, above the $1.2 billion headline consideration announced in February, which was stated as subject to adjustments. Depop's results are in discontinued operations, and the quarter carries a $161.0M loss there, which is the disposal being recognized.

Assessment: $200M more than the headline is a genuine positive that received almost no airtime. Against the June 30 balance sheet of $1.28B of cash and investments and $2.99B of debt, the proceeds take net debt from roughly $1.70B to roughly $0.30B. The GAAP loss on disposal is noise for valuation purposes and is a reminder that Etsy paid $1.6B for Depop in 2021; the cash outcome is what matters, and the cash outcome improved.

3. A New $2 Billion Authorization, and an Accelerating Pace

Etsy repurchased approximately $250M in the quarter, about 70% more than the first quarter, retiring roughly 3.9 million shares. It finished with $578M remaining on the prior authorization and announced a new $2 billion program alongside these results.

"Many of you have expressed interest in when or how we would step up our share repurchases given the cash coming in from the sale of Depop. I'm pleased to report we've already done so." — Lanny Baker, CFO

Total authorized capacity is now roughly $2.58B against a fully diluted market capitalization of approximately $9.87B at the reaction-session close, or about 26%. Free cash flow conversion in the quarter was 81% of adjusted EBITDA, implying roughly $158M of quarterly free cash flow, so the program is funded from operations and the Depop windfall rather than from leverage.

Assessment: this is the most shareholder-friendly capital allocation in our coverage of Etsy and it is the strongest remaining argument for owning the shares. It is also, from a rating perspective, now fully disclosed. In April we identified an accelerated buyback as one of two events that would justify staying constructive, and it has happened. Having credited it in our estimates and in our fair value, we cannot also treat it as a reason to ignore the price it is being executed at. The average repurchase price rose from roughly $53.70 in Q1 to roughly $64.10 in Q2, and the current price is 28% above that.

4. The Restructuring: 12% of the Workforce, in Product and Engineering

Etsy is reducing headcount by approximately 220 employees, roughly 12% of the workforce, concentrated in the product and engineering group, taking post-restructuring headcount to approximately 1,600. Charges of about $35M are expected, largely cash severance, with execution substantially complete by the end of Q3 2026.

"This is not a cost-cutting move. It's about leaning in during a period of strong momentum so that we can move faster and execute with even greater focus." — Kruti Patel Goyal, CEO

The CFO nonetheless confirmed that the expected benefit is incorporated into the raised full-year margin outlook, and separately said the reorganization is not intended to structurally alter Etsy's long-term margin profile, with some of the savings reinvested into engineering, product and customer-operations talent plus R&D. The stated intent is to reshape the talent mix toward machine-learning expertise and to reduce overlap by consolidating teams around shared problems.

Assessment: the strategic logic is coherent and the execution risk is real. Removing 12% of a workforce from the exact function that produced the last three quarters of results, in the middle of a product-led turnaround, is a bet that the organization was carrying more coordination cost than capability. That bet may well be right. But two management statements are in tension: this is "not a cost-cutting move" and its savings are in the raised margin guide. Both cannot be fully true. We treat the margin raise as partly structural and partly severance-funded, and we treat the product roadmap as carrying more delivery risk over the next two quarters than it did before.

5. Marketing Leverage Without Marketing Cuts

Marketing spend was flat in dollars year-over-year, at $190.9M against $191.1M, while GMS grew 7.5%. That produced 180bp of leverage as a percentage of revenue, the largest single contributor to the operating margin expansion. The CFO was emphatic that this is efficiency rather than retrenchment.

"we are earning our growth there, not buying it. We have increased the efficiency in our more mature channels, i.e., search and PLAs. We are making ongoing learnings and having some nice wins in our less mature channels like social." — Lanny Baker, CFO

He explicitly rejected the diminishing-returns reading, describing the dynamic as increasing marginal return and saying the company is inclined to lean in on marketing rather than pull back.

Assessment: flat marketing dollars against 7.5% GMS growth is the most impressive operating statistic in the release, and it is the mechanism by which a modest top line produces near-50% incremental EBITDA margins. The caution is that management named increasing marginal returns while also saying it is not tapped out on investment, which means marketing dollars are likely to rise in 2027 if the returns hold. That is the right decision and it caps the margin at roughly current levels rather than extending the expansion.

6. Amazon's Withdrawal From the Paid-Search Auction

For three consecutive quarters Etsy has cited favorable competitive dynamics in the Google product-listing-ad auction as a growth contributor without naming the cause. This quarter the CFO named it.

"the competitive dynamic in PLA has gotten -- has changed over the last year as Amazon has pulled out. But the bigger driver right now are internal wins that we've made in our marketing technology, particularly the way we segment the PLA feeds that we give to Google" — Lanny Baker, CFO

He was careful to rank the internal work above the competitive effect, citing PLA feed segmentation that helps Google place Etsy's listings with better bids across a broader set of auctions.

Assessment: naming it is a service to investors and it also quantifies the risk qualitatively for the first time. A major competitor withdrawing from an auction lowers the clearing price for everyone remaining, and that benefit reverses in full if the competitor returns. Etsy has no control over that variable and it has been a tailwind to marketing efficiency for a year. This does not undermine the internal work, which is clearly real given that owned channels and app engagement are contributing independently. It does mean some portion of the 180bp of marketing leverage is rented.

7. Average Order Value: Durable Prices, Fading Growth Contribution

GMS per active buyer reached $124 on a trailing-twelve-month basis, growing 2.8% year-over-year. The CFO's decomposition separates three effects: tariff and de minimis-driven seller listing-price increases that arrived in waves and were absorbed by buyers, a foreign-currency translation benefit that was smaller this quarter than in prior ones, and a newer effect from surfacing higher-quality inventory in search results.

"the thing that's changing, sort of the new signal that started to emerge across 2026 is that the efforts that we've undertaken on relevance and quality that we are surfacing in the Etsy search results, is starting to present items that are higher quality with a little bit higher listings price that are better matches for what the customers are looking for." — Lanny Baker, CFO

On durability, the CFO expects the listing-price increases themselves to persist while the year-over-year growth contribution slows, absent new stimulants.

Assessment: this is the correct and honest framing. Price levels are sticky; price growth is not. The average-order-value contribution to GMS growth therefore fades through 2027 even if nothing goes wrong, which is precisely why the mix shift toward quality-driven order value matters. Management is right that a relevance-driven increase in order value is a different and better thing than a tariff-driven one. It is also, on its own admission, only just emerging.

8. Reaching Younger Buyers

Etsy's marketing shift toward social, connected TV and creator partnerships produced a fivefold increase in visits from millennial and Gen Z audiences on YouTube and TikTok in the first half of 2026. The company ran a partnership with Olivia Rodrigo combining an in-person activation with an exclusive merchandise collaboration, and is building a Creator Collective to curate content for younger buyers.

The economics were addressed directly: younger buyers carry somewhat lower current lifetime value but a materially higher growth rate in that value as they move through income and household formation.

Assessment: the diagnosis established in February, that Etsy has a presence gap rather than an appeal gap with younger shoppers, is being executed against with measurable results. A fivefold increase in reach among the target demographic in six months is a real number. The lower current lifetime value means this is a 2028 payoff being funded in 2026, which is the right trade for a business with a 29% margin and no growth constraint on capital, and it is not a driver of the next four quarters.

9. Take Rate at 25.9%, Guided to 26%

Take rate was 25.9%, up 130bp year-over-year, of which roughly 80bp is the arithmetic effect of the Reverb divestiture. The underlying Etsy marketplace expansion was driven by Etsy Ads, using machine learning to improve relevance and seller budget pacing, with offsite ads contributing from a shift of paid marketing toward higher-monetizing channels. Q3 is guided to approximately 26% and the full year to roughly the first-half level.

Assessment: the take-rate story is now doing exactly what management said it would do in April, drifting modestly higher on ads execution rather than on new fee structures, while GMS carries the revenue growth. Marketplace revenue growing 8.4% against services revenue at 11.2% is the healthiest split we have seen in our coverage; a year ago marketplace revenue was declining while services grew double digits. The monetization concern that opened our coverage is fully retired.

10. Compute Costs Are Not the Margin Story

Asked whether higher AI compute costs are compressing gross margin, the CFO gave a specific and useful denial. Gross margin compression is coming from trust and safety, refunds and other customer-experience costs that sit in cost of revenue, not from compute. Total spending on hosting, bandwidth and technology is tracking exactly to plan for the year, with offsets found elsewhere and shifting usage patterns absorbing the incremental AI compute.

Assessment: a clean answer to a question that is compressing multiples across consumer internet. Etsy is spending more on AI and absorbing it inside an unchanged infrastructure budget. The 77bp of gross margin compression is a deliberate investment in marketplace trust, which for a marketplace whose differentiation is human sellers is closer to a moat expense than a cost overrun. This removes one of the more plausible 2027 margin risks.

11. What Etsy Is Choosing Not To Do

Asked about live commerce and about newer marketplace models, the CEO called it an interesting space with potential relevance for sellers who have a story to tell, and said plainly that Etsy has no plans at the moment to launch anything in live shopping. The CFO added that live and resale commerce are growing the overall market rather than dividing a fixed one.

Assessment: declining an adjacent format while the core is inflecting is the right discipline, and it is a notable change from the House of Brands posture that produced Depop and Reverb. Etsy has spent eighteen months simplifying to a single asset; adding a new surface now would contradict that. The cost is optionality, which is the concentration risk we introduced in February and which this answer confirms management has accepted deliberately.

Guidance & Outlook

The double raise is the story. Both the growth outlook and the margin floor moved up in the same quarter, for the second consecutive quarter on the growth line. Full-year GMS growth has now gone from "slight growth" in February to low single-digit in April to mid single-digit in August. The margin floor went from 28% to 29%, with management explicitly stating that the restructuring benefit is incorporated.

The implied second half. Q3 is guided to 4% to 6% growth against a Q2 that delivered 7.5%, so deceleration is embedded in the guide as management has consistently signalled. The full-year mid-single-digit framing against a first half that ran above that implies Q4 lands in the low-to-mid single digits. That fourth quarter, when the currency benefit, the easy comparisons and the de minimis price effect have all lapped, remains the single most informative number on the forward calendar. It is the first clean read on the product-driven growth rate.

Guidance style: four quarters of data now, and the pattern is settled. Every guided metric has cleared in every quarter of our coverage, and the GMS range has been beaten outright twice. Management has also pre-flagged each subsequent deceleration before being asked. We now assume a conservative posture as a working model, which means the mid-single-digit full-year framing should be read as a floor rather than a target.

What is still not guided: no 2027 framework of any kind, which matters more than usual because 2027 is the first year without any of the temporary tailwinds; no deployment schedule for the $2.58B of authorization; and no target for purchase frequency or active buyer growth, both of which management has now declined to commit to for two consecutive quarters.

Analyst Q&A Highlights

What Is Actually Driving the Acceleration

The opening question asked for the specific drivers behind the improved GMS growth and where the remaining opportunity lies. The answer was the most complete articulation of the operating model management has given, and it named a new disclosure: buyer profiles now covering more than 65 million buyers.

Q: "Can you dig deeper on what you think are the most effective drivers of strong and improved GMS growth in Q2? What's working? And where do you see the most opportunity for improvement?"
— Rick Patel, Raymond James

A: "as buyers are engaging more, we are developing a richer understanding of what they're looking for and what makes them tick. That's powering richer buyer profiles that we highlighted in our shareholder letter, and they're now covering over 65 million buyers. And that makes -- if you think about it, that means every subsequent visit that these buyers make more relevant and more personalized, and that creates the flywheel that gets stronger over time."
— Kruti Patel Goyal, CEO

Assessment: Buyer-profile coverage of 65 million against an active base of roughly 87 million is a genuinely new metric and the most concrete evidence yet that the personalization work has scale rather than being confined to a test cohort. The sequencing described (marketing efficiency brings buyers in, on-site improvements convert them better, the resulting data improves the next visit) is a compounding loop rather than a one-time gain, which is the case for the 2027 growth rate holding. It is also unfalsifiable from outside until the frequency metric confirms it.

Whether the Average-Order-Value Gains Survive Normalizing Trade Conditions

A question probed how sustainable the order-value gains are if trade conditions normalize, given that sellers have historically been slow to raise prices. The answer separated the durability of price levels from the durability of price growth, and identified what is replacing the fading contributor.

Q: "in terms of AOV gains that you've seen here, historically, we've seen more limited uptake from sellers. How sustainable do you think that can be if trade changes normalize?"
— Nathaniel Feather, Morgan Stanley

A: "the sort of increases in listing prices that have been taken, I think those -- in our experience, those, I think, will turn out to be fairly durable increases. The year-over-year increase will probably slow down unless there are other stimulants that the sellers are reacting to."
— Lanny Baker, CFO

Assessment: Exactly the right distinction, and it should be built into any 2027 model. Sticky price levels mean the base does not give back; a slowing year-over-year increase means the growth contribution decays to zero over roughly four quarters. The offsetting factor management identified, higher-quality inventory being surfaced by better relevance, is real but described as only starting. The net effect is that order value shifts from a tailwind to a neutral through 2027, which is a meaningful part of why we now model deceleration rather than continuation.

How Smaller Product and Engineering Teams Move Faster

The sharpest question of the call asked how cutting 12% of the workforce, concentrated in product and engineering, squares with naming machine learning as the core of the growth strategy, and what specifically the company is choosing to stop doing. The answer addressed structure and talent mix and did not name anything being stopped.

Q: "I just wanted to ask on workforce reduction, which is largely across product and engineering. But then at the same time, you're calling ML as kind of the core of discovery and personalization to sort of accelerate growth further from here. Can you maybe give us a little bit more color on sort of how smaller teams move the road map faster? And sort of what specifically are you choosing to stop doing that you focused on before?"
— Maria Ripps, Canaccord Genuity

A: "when we looked across product and engineering, we saw an opportunity to really simplify the organization, and that's where the streamlining comes in, really this opportunity to reduce overlap by bringing teams together around shared problems or shared capabilities… The other part of that is this change allows us to reshape our talent mix around the capabilities that matter most for the next phase of growth."
— Kruti Patel Goyal, CEO

Assessment: The reduce-overlap and reshape-talent-mix framing is a coherent theory of organizational design and it is not evidence. The second half of the question, what Etsy is choosing to stop doing, went unanswered, and a 12% reduction that stops nothing is a productivity assumption rather than a plan. We are not able to underwrite the claim either way from outside; what we can do is treat the next two quarters of product delivery as carrying more execution risk than the prior two, which is a change from a business that had been executing cleanly.

Why the Restructuring Savings Are Not All Reinvested

A question asked why the reorganization savings flow to the bottom line rather than into more marketing, and whether that signals diminishing returns. The answer rejected the premise on marketing and reframed what kind of margin growth management wants.

Q: "just a quick one for Lanny on some of the cost savings from the reorganization. Sounds like you're letting this flow through to the bottom line. I was wondering why not reinvest even more in marketing or if there's diminishing marginal returns there or anything along those lines would be great."
— Michael Morton, MoffettNathanson

A: "I think the kind of margin growth that we really like is the margin growth that comes from revenue and GMS growth. And in this quarter, 50% of the revenue growth went down to the bottom line… And actually, that doesn't -- that's not declining marginal return. That's increasing marginal return, and that's sort of causing us to lean in on marketing."
— Lanny Baker, CFO

Assessment: The claim of increasing marginal returns on marketing is unusual and, if durable, is the most valuable statement on the call: it means Etsy can grow spend without losing leverage. It also caps the margin story, because management said explicitly it intends to lean in rather than harvest. Investors modelling continued margin expansion off the raised 29% to 30% range should not; management has told them the incremental dollar goes back into growth. That is the correct decision and it means the 2027 case rests on GMS, not on margin.

Whether Compute Costs Are Eating the Gross Margin

A question on the year-over-year gross margin compression asked whether higher AI compute costs are the cause and whether the pressure continues. The answer was a specific denial with an alternative attribution.

Q: "And then just a follow-up on gross margin. You've seen some compression on the higher compute costs. Should we expect that pressure to continue in '26 and just any way we can think about that?"
— Noah Helfstein, Piper Sandler

A: "we've seen a little bit of compression year-to-year, and it's not coming from compute… the overall spending that we're doing on hosting and bandwidth and technology is exactly where we thought it would be this year, and we've been able to find offsets in other places and shifting usage patterns allow us to absorb the cost of the incremental compute without going higher on our overall sort of infrastructure cost."
— Lanny Baker, CFO

Assessment: A clean, checkable answer to the question compressing multiples across the sector. The 77bp of gross margin compression is attributed to trust and safety, refunds and customer-experience costs, which for a marketplace whose entire differentiation is human sellers is closer to a moat investment than a cost problem. Absorbing incremental AI compute inside an unchanged infrastructure budget also removes a plausible 2027 margin risk. This is the most reassuring answer on the call.

What Changed in Paid Search

Asked what specifically has changed in marketing spend and where the returns are best, the CFO gave the disclosure that had been missing for three quarters: the identity of the competitive change in the product-listing-ad auction.

Q: "Could you please talk to, if anything has -- what specifically has changed in your marketing spend where you're seeing spend and strategy, where you're seeing best returns on that spend? And how much of that runway do you have left ahead of -- to drive potentially new and reactivated buyer growth?"
— Shweta Khajuria, Wolfe Research

A: "the competitive dynamic in PLA has gotten -- has changed over the last year as Amazon has pulled out. But the bigger driver right now are internal wins that we've made in our marketing technology, particularly the way we segment the PLA feeds that we give to Google, the way -- the information that we give them around listings is helping them, I think, place our listings with better, more competitive bids in a broader number of auctions."
— Lanny Baker, CFO

Assessment: Naming the competitor withdrawal is a service to investors and it also converts a vague tailwind into a specific, reversible one. A large bidder exiting an auction lowers the clearing price for everyone remaining, and that benefit unwinds completely if they return. Management ranked the internal feed-segmentation work above it, which is credible given that owned channels and app engagement are contributing independently. But some part of the 180bp of marketing leverage that drove this quarter's margin beat is rented from a competitor's decision, and Etsy controls none of it.

Whether Live Commerce Is a Format Etsy Should Enter

A question on emerging commerce formats asked whether live shopping and similar models fit Etsy's sellers. The answer was an unusually direct decline, which is itself a strategic statement from a company that spent the prior five years acquiring adjacent marketplaces.

Q: "Could you talk about the opportunity in live commerce, one? And two, you're seeing players like a Whatnot, et cetera, that have kind of a unique take… as you think about your seller base and maybe they're thinking about innovative ways to sell their products and market their products, how these may or may not fit with Etsy?"
— Ken Gawrelski, Wells Fargo

A: "we think that there's some potential opportunity for sellers like ours who have a really great story to tell, who have a lot to share about how they create, what they create that's relevant to a buying audience. So we're certainly open to exploring this, but we don't have plans at the moment to launch anything in live shopping."
— Kruti Patel Goyal, CEO

Assessment: Declining a fashionable adjacency in the middle of a working turnaround is discipline, and it is a deliberate reversal of the strategy that produced Depop and Reverb. The trade is explicit: Etsy gives up optionality in exchange for focus, and the concentration risk we introduced in February is a risk management has now chosen twice. Given that live commerce would require building a new surface, a new seller toolset and a new discovery model while the existing ones are only just inflecting, we think the trade is correct.

What They're NOT Saying

1. Anything about 2027. This matters more than in prior quarters. 2027 is the first year without the currency benefit, the easy comparisons or the de minimis-driven price increases, and by management's own account those were three of the four contributors to 2026's acceleration. No framework, no directional marker, and no hint of what the product-driven growth rate looks like on its own.
2. What the restructuring stops. Asked directly what the company is choosing to stop doing, management described structural simplification and talent-mix reshaping and named nothing that is being discontinued. A 12% reduction that stops no work is a productivity assumption.
3. How the "not a cost-cutting move" squares with a raised margin guide. The CEO said the restructuring is not cost cutting; the CFO said its benefit is incorporated into the raised full-year margin outlook. Both cannot be fully true, and the split between structural improvement and severance-funded savings was not given.
4. A deployment schedule for $2.58B. The new $2B authorization plus $578M remaining is 26% of the fully diluted market capitalization, and there is no timeframe, no pace commitment, and no indication of whether the $1.4B of Depop cash gets deployed over four quarters or twelve.
5. How much of the marketing leverage is Amazon's absence. Management named the competitor withdrawal, ranked the internal work above it, and did not size either. It has now been a contributor for a year and it is entirely outside Etsy's control.
6. A frequency or active-buyer target, for a second consecutive quarter. The metric management calls the ultimate goal remains below prior year, and there is still no stated objective, timeline or definition of what success looks like.
7. Category detail. Management said Etsy grew in every top category and outgrew peer benchmarks in the US, then explicitly declined to give category-level figures on the grounds that buyers shop by occasion rather than category. That framing is defensible and it also removes a disclosure the company used to provide.
8. Agentic commerce, almost entirely. After three quarters in which it was the dominant Q&A topic, agentic partnerships were mentioned only in passing. Traffic share, fee structure and GMS contribution all remain undisclosed. Silence is better than bad news here, and it is still silence.
9. Total active listings. Raised in all three prior recaps and still absent.

Market Reaction

• Pre-print setup: ETSY closed at $85.86 on August 5, up 54.9% year to date against an S&P 500 up 12.8%, up 53.4% over the trailing twelve months, and up 14.4% over the trailing thirty days. The 52-week closing range entering the print was $44.05 to $86.47, so the stock closed within 1% of its 52-week closing high on the day of the release.
• Reaction session (August 6, the day of the 8:30 a.m. ET call): the shares gapped down to open at $81.99, a 4.5% gap, traded a $77.85 to $84.22 range, and closed at $82.26, down 4.2% or $3.60.
• Volume: 7.1 million shares against a 2.7 million 30-day average, 2.7 times normal.
• Market context: the S&P 500 fell 0.2% on the session, so essentially the entire move was idiosyncratic.

The timing within the reaction is informative. Etsy released results after the close on August 5 and held its call the following morning; press coverage that evening described the shares as roughly flat immediately after the release. The decline came during the August 6 session, after the call. Whatever repriced the stock was said on the call rather than printed in the release, and the two candidates are the 12% restructuring and the deceleration embedded in the Q3 guide.

The intraday pattern shows the market working through it rather than dumping. The shares opened at $81.99, fell to $77.85 (a 9.3% decline from the pre-print close), and then recovered to close at $82.26, above the open and in the upper half of the range. Buyers stepped in near the $77.85 low. That is not a thesis break; it is a valuation adjustment.

The signal we take from this is the important one for the rating. Etsy delivered a GMS beat against its own guide, a margin beat against its own guide, a revenue and EPS beat against the Street, a raise to the full-year growth outlook, a raise to the full-year margin floor, $1.4B of cash received, and a new $2B buyback authorization. The stock fell 4.2%. When a stock cannot rally on that combination, the market is telling you the good news is in the price. Our job is to listen rather than to argue with it.

For coverage continuity: we initiated at Hold on October 30, 2025 at $65.21, upgraded to Outperform on February 20, 2026 at $48.14, and maintained on April 30 at $69.60. From the upgrade the shares have returned 70.9% against an S&P 500 up roughly 12% over the same stretch.

Street Perspective

Debate: Does a 7.5% Quarter Mean the Growth Rate Is Structurally Higher?

Bull view: Three consecutive quarters of accelerating growth, with the acceleration now showing up in the hardest metrics: habitual and repeat buyers growing sequentially for the first time since 2023, active buyers adding 350,000, GMS per buyer up 2.8%, app GMS up 12.5% with visits per user and orders per visit both rising. Buyer profiles cover 65 million people and improve with every visit. Marketing dollars were flat while GMS grew 7.5%. This is a compounding loop, and the guide has been raised twice.

Bear view: Management has told investors three of the four contributors to the acceleration are temporary, and the Q3 guide of 4% to 6% against a 7.5% Q2 shows the fade beginning on schedule. Purchase frequency, the metric the CEO calls the ultimate goal, is still negative. Part of the marketing leverage is a competitor's absence from an ad auction. Average order value gains persist in level but decay in growth. Strip all of that and the durable rate is unknown, and 2027 has no framework at all.

Our take: The bull case has strengthened materially this quarter, and it still does not resolve the question. The cohort inflections are exactly what a genuine structural improvement looks like at this stage, and they are the reason we are downgrading to Hold rather than to Underperform. But the decisive evidence is Q4 2026, when the currency benefit, the easy comparisons and the de minimis price effect have all lapped. Everything before that is a mixture of real progress and fading help in proportions nobody outside the company can measure.

Debate: Is 14.6 Times Forward Free Cash Flow the Right Price?

Bull view: A capital-light marketplace with 72% gross margins, a 29% to 30% adjusted EBITDA margin, mid-single-digit growth that is accelerating, near-zero net debt after the Depop proceeds, and $2.58B of buyback authorization equal to 26% of the market capitalization should trade at 16 to 18 times free cash flow, not 14.6. Nearly half of incremental revenue drops to adjusted EBITDA, so even modest GMS growth compounds. The buyback alone delivers high single-digit annual growth in free cash flow per share.

Bear view: The multiple has expanded from 8.8 times in February to 12.5 in April to 14.6 today, so the entire re-rating has already happened. Free cash flow of roughly $660M to $695M against a fully diluted market capitalization of $9.87B is a 6.9% yield, and absolute free cash flow is roughly flat because management has said it will reinvest marketing savings rather than harvest them. Paying 14.6 times for flat cash flow plus a buyback requires the multiple to hold through a 2027 that has no guidance and no visible growth driver once the tailwinds lapse.

Our take: The bear has it, narrowly, and that is the downgrade. At $48.14 we were buying an 11.4% forward free cash flow yield on a business the market thought was in decline. At $82.26 we are buying a 6.9% yield on a business the market now understands is stable-to-growing. That is the correct repricing and we captured it. What remains is a low-to-mid teens annual return if everything goes right, against a bear case that takes the shares back to the mid-$50s if the growth fades faster than the buyback compounds. That is a Hold, not a sell.

Debate: Is Cutting 12% of Product and Engineering Bold or Reckless?

Bull view: Management is concentrating talent where it matters, reducing handoffs and overlap, and reshaping the skill mix toward machine learning, which is the capability the entire strategy depends on. It is doing this from a position of strength rather than distress, and reinvesting part of the savings into engineering, product and customer-operations talent plus new R&D. A flatter organization executing faster is worth more than a larger one coordinating more.

Bear view: Removing 12% of the workforce from the exact function that produced three quarters of accelerating results, while naming nothing that will stop being done, is a productivity assumption dressed as a strategy. The savings are simultaneously described as not being the point and as funding a raised margin guide. Reorganizations reliably cost two to three quarters of delivery velocity, and this one lands precisely as the comparisons harden.

Our take: The strategic logic is sound and the timing is the risk. We would rather see this executed from strength than from distress, and the machine-learning talent argument is genuine given where the value is being created. But the unanswered question, what stops, is the one that matters, and the delivery risk over the next two quarters is real and new. This is a reason the rating goes to Hold rather than staying at Outperform, though it is not the main one.

Model Update & Valuation Framework

Valuation framework: at the $82.26 close, Etsy carries a fully diluted market capitalization of approximately $9.87B on the 120.0M diluted share count, or approximately $7.76B on the 94.4M basic count. Against our FY2026 free cash flow estimate of $660–695M, that is a 6.7% to 7.0% fully diluted free cash flow yield, or roughly 14.6 times at the midpoint. The June 30 balance sheet carried $1.28B of cash and investments against $2.99B of debt, so net debt of roughly $1.70B, reduced to approximately $0.30B by the $1.4B of Depop proceeds received on July 30.

The multiple progression is the rating case. At our February upgrade the shares traded at roughly 8.8 times forward free cash flow, a multiple that priced structural decline. In April they traded at 12.5 times, a multiple that priced stability. They now trade at 14.6 times, which prices modest growth. Each step was justified by the evidence and each step consumed part of the return. What is left is a multiple appropriate to what the company has actually demonstrated, which means further gains have to come from free cash flow per share rather than from the market's opinion.

The buyback arithmetic, done honestly. Roughly $2.58B of authorization against a $9.87B fully diluted market capitalization is 26%, and management has demonstrated it will spend. But the accretion per dollar has fallen. In Q1 Etsy retired shares at roughly $53.70; in Q2 at roughly $64.10; at $82.26 the same $250M retires 3.0 million shares rather than 4.7 million. Assuming roughly $1.3B deployed over the next twelve months at prevailing prices, the diluted count falls from 120.0M toward roughly 105M, which is a 12% reduction and the largest single component of our base case.

Risk/reward: base case $88–94 implies +7% to +14% against the $82.26 close. Bull case $112–124 implies +36% to +51%. Bear case $54–62 implies −34% to −25%. The base-to-bear ratio is roughly 0.4 to 1, down from 0.6 to 1 in April and 2.2 to 1 at the upgrade. At some point a deteriorating ratio has to change the rating rather than be explained away, and this is that point.

Thesis Scorecard Post-Earnings

Overall: the operating thesis is the strongest it has been and the investment case is the weakest. Five of the original bear points are now resolved, including the one (core contraction) that defined the initiation. What replaces them are three risks that are all consequences of success: the growth was partly rented, the organization is being rebuilt mid-flight, and the price now reflects the outcome. This is the position a successful turnaround call arrives at, and the discipline it requires is to recognize it rather than to keep re-underwriting the same thesis at higher prices.

Action: take the gain. Not because anything is wrong, but because the asymmetry that justified the position has been earned and spent.

Bottom Line: Downgrading to Hold

Rating decision: we downgrade Etsy to Hold from Outperform. This is a valuation call and not a business call. The second quarter was the best in our coverage: a GMS beat against the company's own guide, a margin beat against its own guide, revenue and EPS above the Street, the full-year growth outlook raised for the second consecutive quarter, the margin floor raised with it, the Depop sale closed above headline value, and a new $2 billion buyback authorized. We were right to upgrade at $48.14 and the shares have returned 70.9% since.

Why we are stepping aside anyway. Three numbers. The free cash flow yield has compressed from 11.4% at the upgrade to 6.9%, both measured forward. The multiple has expanded from 8.8 to 14.6 times forward free cash flow. And the base-to-bear risk/reward has fallen from 2.2 to 1 to roughly 0.4 to 1. A fourth is the market's own verdict: this stock fell 4.2% on a quarter that beat and raised twice and announced a buyback worth a quarter of the company. When that happens, the news is priced.

On overriding our own trigger. In April we wrote that we would downgrade if the shares reached the upper half of our base range "without a frequency inflection or a visible acceleration in buyback pace." The shares are there, frequency has not inflected, and the buyback did visibly accelerate, so the mechanical trigger did not fire. We are downgrading anyway, for a reason worth stating plainly: we have already credited the accelerated buyback inside our estimates and inside a fair-value range we raised from $78–86 to $88–94. Counting it a second time as a reason to disregard the price would be double-counting our own good news. The escape clause was designed to test whether management would convert the Depop cash into per-share value. It has, that conversion is in the numerator, and the price has moved to reflect it.

What we would need to go back to Outperform. Not much, and it is specific: evidence that the growth is durable once the help is gone. A purchase-frequency inflection, or Q4 2026 GMS growth holding in the mid single digits after the currency benefit, the easy comparisons and the de minimis price effect have all lapped, would tell us the 2027 rate is higher than we are modelling and would justify a higher multiple on a larger number. Alternatively, a pullback to the low $70s without a change in the fundamentals restores the asymmetry on its own.

What would take us to Underperform:

• Q4 2026 GMS growth falling below 2%, which would confirm that the acceleration was substantially rented rather than earned.
• Visible product-delivery slippage following the restructuring, showing up as app GMS growth decelerating below 8% or the personalization roadmap stalling.
• Amazon returning to the product-listing-ad auction, or any other reversal that takes marketing efficiency back to 2024 levels while GMS growth is still dependent on it.
• The buyback pace slowing materially while the $1.4B of Depop cash sits idle, which would remove the largest component of the base case.

Signposts for Q3 2026 earnings (November 2026):