Compass, Inc. (COMP) Earnings Call Transcript & Summary

Good day, everyone. Welcome to the Compass Inc. 2026 Q2 Earnings Call Conference Call. [Operator Instructions] This call is being recorded. If you have any objections, please disconnect at this time. I would now like to turn our call over to Soham Bhonsle, Head of Investor Relations. Please go ahead.

Thank you very much, operator, and good afternoon, everybody, and thank you for joining the Compass Second Quarter 2026 Earnings Call. Joining us today will be Robert Reffkin, our Founder and CEO; and Scott Wahlers, our Chief Financial Officer. In discussing our company's performance, we will refer to some non-GAAP measures and discuss some metrics on a non-GAAP pro forma basis. You can find the reconciliation of the non-GAAP measures to the most directly comparable GAAP measures and supplemental non-GAAP pro forma information for prior quarters in our second quarter 2026 earnings release posted on our Investor Relations website. We will also make forward-looking statements that are based on our current expectations, forecasts and assumptions and involve risks and uncertainties. These statements include our guidance for the third quarter of 2026 and full year 2026 and comments related to our expectations for realizing cost synergies and operational achievements. Our actual results may differ materially from these statements. For more information, see our most recent annual report on Form 10-K and our recent Form 10-Q. You should not place undue reliance on any forward-looking statements. I will now turn the call over to Robert Reffkin. Robert?

Good afternoon, and thank you for joining us for our Second Quarter Conference Call. On today's call, I will be discussing five topics. First, I will provide a quick recap of our record Q2 results. Second, I will share an update on our cost synergy targets and our technology rollout. Next, I'll touch on our partnership with Rocket-Redfin and how we are infusing competition in residential real estate. Fourth, I'll discuss how we are moving on the offense with AI. And I'll end by revisiting the long-term earnings potential of our business. Starting with our record Q2 results, where all my year-over-year and quarter-over-quarter comments will be against pro forma figures. In Q2, Compass delivered record revenue and record adjusted EBITDA, above the high end of our guide. Revenue of $4.3 billion was up 14% year-over-year compared to pro forma revenue of $3.8 billion a year ago, while adjusted EBITDA was $363 million. Cash on hand increased by $210 million quarter-over-quarter to $694 million, which drove net debt to adjusted EBITDA on a trailing 12-month and a pro forma basis to 3.3x compared to the 4.2x in Q1 of 2026. So this means over the past quarter, we have already lowered our net leverage ratio by roughly one turn. Assuming current 2026 consensus adjusted EBITDA estimates and our Q2 ending cash balance of $694 million, which would be conservative as we expect to generate free cash flow in the second half of the year, our net debt to adjusted EBITDA ratio would move into the 2s by the end of the year, illustrating the progress we continue to make on reducing our net leverage ratio even as we are at the bottom of the cycle. In our brokerage business, which includes the Coldwell Banker, Compass, Corcoran and Sotheby's International Realty brands, transactions were up 7.4% year-over-year compared to the market, which was up 3.5% year-over-year. As a result, for 21 consecutive quarters, spanning our entire history as a public company, our brokerage business has outperformed the market on an organic basis for the Compass stand-alone brand. And for the second consecutive quarter, including the Anywhere transaction, we have now outperformed the market as well. Additionally, brokerage gross transaction volumes, or GTV, was up approximately 16% year-over-year compared to the market that was up 6%. This reflects roughly 1,000 basis points of out-performance compared to the market, which is an acceleration compared to the roughly 600 basis points of out-performance we delivered in Q1 2026. We believe this out-performance is a reflection of the quality of our agents and exposure to the higher end of the market, which tends to be less rate sensitive. We believe the wealth effect created by a record stock market and a growing U.S. economy has been a driver of demand for our business and is helping offset the rise in interest rates. Recently, there has been much discussion regarding the Bay Area real estate boom, driven by the SpaceX IPO, the upcoming SpaceX lockup and the potential IPOs for Anthropic and OpenAI. So we thought it would be helpful to provide some color on what we are seeing there given our presence in that market. On a year-over-year basis, in the Compass stand-alone brand, we are seeing revenue in the Bay Area up 19% year-over-year in both July and August on a business day adjusted basis, which is relatively consistent to the up 20% we observed in June. This suggests the momentum coming out of June in the Bay Area is continuing. In franchise, GTV was up 11.7% year-over-year compared to the housing market volumes that were up 6%, reflecting 570 basis points of out-performance. Our high-end brands, including the Corcoran and Sotheby's International Realty brands, continue to significantly outperform the company average. Integrated services revenue grew 7.7% year-over-year with title and escrow revenue being the primary driver. Total T&E transactions grew by 7.6% year-over-year with purchase transactions growing by 6% and Refi transactions growing by 25% year-over-year. Our focus in 2026 into 2027 is to unify our T&E operations by migrating all T&E operations to the Compass's title production platform. Transitioning to one title production platform is expected to unlock additional efficiencies and expand incremental margins in the future. While we are making this transition, however, we will continue to optimize our attach efforts by scaling best practices from both organizations and expect our attach efforts to accelerate in 2027 as we deploy tools such as one-click title across our footprint. In mortgage, our JVs more than doubled profitability in the Q2 year-over-year period, primarily due to strong volume growth and disciplined expense management. Our focus in 2026 in mortgage is to drive operational efficiencies within our JV entities while also making progress on attach by continuing to attract the best loan officers to the business. Now an update on our cost synergy and integration efforts, starting with our cost synergies. As of July, we actioned our entire year 1 target of $300 million in net cost synergies 5 months ahead of plan. As such, we are now pacing above our $300 million target and expect our actioned net cost synergies to be $330 million by the end of year 1, with our 2026 in-year realized net cost synergies to be $220 million compared to the $200 million that we previously stated. We expect to realize about $150 million of the $220 million through the P&L compared to $130 million previously stated, and we continue to expect the remaining $70 million to be realized as CapEx synergies later this year. Lastly, regarding our total actioned net synergy target of $500 million. While we are not changing our total target at this time, given the accelerated pace at which we have been moving, it would be fair to assume that we will achieve the $500 million in cost synergies in less than 3 years and that we will achieve more than $500 million in cost synergies in 3 years. As a reminder, approximately $420 million of the $500 million is expected to be realized through the P&L and $80 million is expected to be realized as CapEx synergies. Next, on to our technology rollout. In July, we achieved a significant milestone as we made our newly branded home platform technology available to over 4,000 agents at Coldwell Banker, Corcoran and Sotheby's International Realty in the pilot beta. Feedback since the rollout has been positive with agents noting the ease of use on both mobile and desktop, the network effect benefits of the platform, particularly in the private exclusive phase and the highly integrated nature of the platform. So far, the platform has received an 82% CSAT or customer satisfaction score, which is generally considered to be a strong score in the software industry. By the end of September, nearly 50,000 new agents in the owned brokerage brands will have access to the platform, reflecting over 80,000 agents on the platform, including the Compass Brokerage brand. All agents will have access to our private exclusives and coming soon inventory. Our 128,000 domestic franchise agents will begin to onboard in Q1 2027. I want to give a special thanks to our product and engineering team, our marketing team and our coaching teams for the monumental effort in getting us to the point in record time. I cannot overstate the relentless execution and tireless commitment the team put into this to make sure that the agents have access to the platform ahead of the fall market. I am incredibly proud of the team. And in all my years at the company, I have never seen the team come together to realize such a big and audacious goal. Thank you. Now let me provide a few thoughts on our partnership with Rocket-Redfin and how we are infusing competition in residential real estate. Starting with our Rocket-Redfin partnership. Since launching Coming Soons on Redfin in late Q1, our agents have received more than 60,000 leads from Rocket-Redfin, and Compass has delivered more than 20,000 coming soon listings to Redfin. Moreover, we want to share our first data point highlighting the impact to consumer traffic resulting from the coming soon inventory on Compass.com. In Chicago, where we have the most Coming Soons of any market, the number of sessions on Compass.com in Chicago were up 111% year-over-year, outpacing the average sessions growth of 34% year-over-year on Compass.com, or by 77 percentage points. This reflects what happens when MLS rules let clients and their agents choose to market properties how they see fit. Which, in this case, is through the local MLS, MRED. We expect more than 90% of MLSs to have rules that allow sellers to market both private exclusives and Coming Soons by the end of the year. The trend for seller choice is moving quickly. Moving on to how we are infusing competition in real estate. I believe in competition. Not only does the law require companies to compete, but competition is the bedrock of our economy. Competition is the engine that produces the most value and options for consumers. However, today, the most powerful entity in real estate, the Multiple Listing Services, or MLS, they do not compete. Instead, many abuse their power by creating mandatory rules and they enforce with MLS fines up to $5,000 that every real estate professional and their sellers are expected to follow. Real estate professionals have no ability to push back because they need access to MLS to do their job. That is because over the past 100 years, all real estate professionals have been conditioned to use the MLS to market a listing to other real estate professionals. So without access to MLS, you can't access the listing data for your buyers. In almost all markets, agents have only one choice for Multiple Listing Services, making that single MLS in that market a monopoly that agents need to use to do their jobs. It is that market power that has empowered MLSs to fine agents up to $5,000 for marketing outside MLS, even though the MLS is not the government. It's just a private entity. What other private business can fine and punish other private businesses. The untold secret in real estate is that the MLS is controlled by a collection of our competitors that tell us how we can and cannot compete. The MLS is controlled by a collection of our competitors, who are running the board, and they are telling us how we can and cannot compete. What other private business is told how to compete by a collection of their competitors? This is why the MLS system has been investigated or sued by the United States government over 100 times in the last 50 years. The MLS has been weaponized against the very customer that pays to get access to that service utility, namely the real estate agent and the real estate brokerages, those agents associate with. Let me say that again. Real estate agents pay the MLSs money. Real estate agents give the MLS the result of their hard work and their intellectual property in exchange, the MLS tells the real estate agent how to compete and fines and punishes the agent if they compete too hard. This is anticompetitive. This is anti-consumer and it's illegal. Multiple Listing Services should have to compete for our business just as brokerages have to compete for agents and agents have to compete for their clients. Today, our agents compete relentlessly for every client. They compete at every kitchen table across the U.S. every single day. Brokerages also compete fiercely on compensation, on technology, on coaching, on culture to attract and retain the industry's top professionals. Over the years, Compass has invested over $2 billion in technology to compete to provide competitive technology offerings. Because of competition, our real estate professionals cannot fine or punish other agents or tell them how to compete. And because of competition, we cannot fine or punish brokerages and agents or tell them how to compete, but most MLSs can. And most MLSs do because they have no competition. I am working to change that and bring competition to the MLSs so that they can compete for real estate professionals and succeed in those MLSs that do not compete fail. Similarly, the dominant portal once competed for the attention of the consumer, competed for listings. But over the past 20 years, it gained dominant market power and with it, adopted restrictive rules that punish brokers and agents who are competing for those same consumers. Like MLSs, the dominant portal takes the hard work and intellectual property created by brokerages and agents without our permission and [ forge ] free, takes it from the MLS and use it for lead diversion to make money on it time and time again. But if the dominant portal had to actually compete to get our listings, compete on price, compete on features, compete on value, they would have to completely change their business model or fail. If the dominant portal no longer got listings for free from MLS, they would need to compete for those listings resulting in listing agents getting their buyer inquiries and their name and brand back on their listings. So everything Compass has done in the past and everything it is doing today is designed to infuse competition into MLS and portals and real estate. Why? Because if multiple listing services and the dominant portal have to compete like real estate professionals and brokers have to compete every day, consumers and the agents that represent them win. If they have to compete for brokerages, brokerages win. This is why I am supporting Multiple Listing Services and portals that compete for our real estate professionals, and I'm working to inject competition in the MLS and portal ecosystem. A good example of how competition can create change is the Rocket-Redfin partnership that was announced earlier this year. Less than a month after we announced the partnership, the dominant portal discontinued the restrictive ban on Coming Soons marketed outside of their platform and launched a coming soon product for agents and sellers on their own platform. Competition resulted in more choices for agents and more choices for the consumer. Another example is our recent announcement with several multiple listing services across the country that supports those that actually compete for real estate professionals. As a result, we have seen some of the largest MLSs in the country, including those in Chicago, Washington, D.C., Philadelphia, Southern California, Florida and Nashville, they have begun to compete, including by offering more flexible rules that let home sellers and their agents determine where and how to market their own properties. MLSs that are expanding recognize that if they want to get agents in new markets to sign up to their MLS, they can't expand with more restrictive rules and fines. Instead, they need to compete with more marketing options and more marketing flexibility that helps listing agents and their sellers as opposed to helping the dominant portal. If MLSs and portals had to compete just like we compete with over 80,000 brokers in the country or how agents compete with the 1.5 million agents in the country, I am confident that our company will be able to create an incredible amount of value for all stakeholders, including our real estate professionals, their clients, our employees and our shareholders. Every month, we are seeing more and more MLSs competing in ways that help our clients, our agents and our company. Competition will empower the best agents and the best brokerages. Competition will naturally eliminate restrictive rules and empower the 3-phase marketing strategy. Competition will unlock the full potential of the Compass business model and competition is coming. Finally, we are also seeing several states encode into law a seller's right to market property how they see fit through seller opt-outs. Connecticut, New York and Wisconsin now allow sellers to market their home privately with seller opt-outs. Over the past year, our 3-phase marketing strategy has required similar seller disclosure that states are now putting in place, so we only see this as a big positive. Furthermore, Washington State has adopted laws that allow home sellers to publicly market their homes however they want, so long as the home is concurrently marketed to the general public and all other brokers. And our private exclusives and Coming Soons are marketed to the general public through our website. So as you can see, the trend at the state level is also one where the law is emphasizing that homeowners, the people who actually own the property for sale, get to choose how to market their properties, not multiple listing services or the dominant portal. Overall, all of these changes are well aligned with our 3-phase marketing strategy, which continues to see an increase in adoption across the Compass brand with adoption in the most recent week approaching 57% of all new Compass listings. By the end of the third quarter, I would expect 80% of Compass Brokerage listings to launch as a coming soon on Compass.com and Redfin and the total number of Coming Soons to build from there as we expand the offering to all of our brands. Of note, 65% of our franchise network participated in a Redfin coming soon preview session in late June, with 40,000 of those agents already opting in to the Redfin Direct listing tool. And by the end of Q3, we expect over 180,000 agents of our franchise and brokerage agents to have the ability to create Coming Soons. Now shifting to our AI strategy, where we are moving on offense in two ways: one, reducing OpEx; and two, increasing agent productivity. First, we are using AI to reduce OpEx, as you would expect. In Q2, we began deploying FDEs, or forward-deployed engineers within business functions, including transaction management, legal and growth to build automated AI workflows directly into day-to-day operations. To date, the team has identified a mix of roughly $8 million in savings and cost avoidance opportunities. Additionally, across the technology organization, 50% to 60% of all new code is now produced by AI, which is helping us ship code faster and more efficiently, 50% to 60%. Second, on agent productivity, we are rolling out tools to make our real estate professionals more efficient. In July, we demoed our AI assistant more broadly, which is an integrated assistant that helps agents orchestrate more than 90 of our platform tools, simply through natural language prompts. Our earnings deck this quarter includes several direct testimonials from agents. And as you will see, early feedback is incredibly positive. Agents are citing the amount of time the AI assistant is saving them on everyday tasks such as client outreach and importantly, helping them unearth proprietary leads. User data indicates deeper engagement with the platform with a number of tools called per agent and number of conversations per agent up almost 2x since our demo day. I want to end by revisiting the long-term earnings potential of our business as we discussed last quarter. As a reminder, our scenario analysis is not meant to be guidance and assumes no agent adds, no organic share take, no margin improvement, no improvement on T&E or mortgage attach or any contribution from leads or other ancillary revenue. These are all incremental growth levers in our business beyond the housing recovery and levers that we are beginning to pull, as you can see from our Q2 results. Assuming the housing market remains flat at 4.1 million existing home sales, we would generate roughly $1 billion in adjusted EBITDA and $750 million in unlevered free cash flow. In the next scenario, which we've assumed as 4.8 million existing home sales for this analysis, we would generate $1.5 billion in adjusted EBITDA and $1 billion in unlevered free cash flow. At mid-cycle levels of 5.5 million home sales, we would generate $2 billion in adjusted EBITDA and $1.5 billion in unlevered free cash flow. And lastly, we also provided an upside scenario of 6 million home sales. And at that level, we would generate $2.5 billion in adjusted EBITDA and roughly $2 billion in unlevered free cash flow. With that said, I will now hand it over to our CFO, Scott.

Thank you, Robert. Consistent with our prepared comments from last quarter, I'll provide some information about the contribution to our consolidated results from the acquired Anywhere businesses where possible. However, we're integrating the entities quickly and therefore, do not generally expect to break out separate results going forward. You'll find financial information included on our Investor Relations website from last quarter that provides additional supplemental information on a pro forma basis as though the Anywhere business was combined with Compass for the full year of 2025. Where applicable, this prior year pro forma information will be referenced in my prepared comments today. Now moving into the financial details. Revenue in Q2 reached $4.3 billion, which exceeded the high end of our revenue guidance range of $4 billion to $4.2 billion. On a pro forma basis, consolidated revenue was up 14.3% from the year ago period. We saw year-over-year revenue increases in each of our three operating segments during the quarter, but the owned brokerage segment was a standout with $3.96 billion of revenue and a 15% increase on a pro forma basis compared to the year ago period. Wealth being created by the AI boom in the Northern California market supported some of this revenue growth. Our Chief Economist, Mike Simonsen, highlighted that during the first 6 months of the year, 140 transactions in San Francisco closed at least $1 million above the asking price, with 44 of those transactions by 6% compared to just eight homes closing $1 million above asking price in the first half of last year. Gross transaction value for the brokerage segment was $155.2 billion in the second quarter, up 16% year-over-year on a pro forma basis, which reflects a 7% increase in transactions and average selling price appreciation of 8% to an average price of just over $1 million. Our talented real estate professionals and our presence in the luxury markets allowed us to capitalize on this market growth. As Robert commented earlier, the increase in our gross transaction value during the quarter of 16% on a pro forma basis compares very favorably to the overall market that was up 6% on volume. Gross agent adds in the quarter were 2,816, down from 3,503 in Q1. As we mentioned last quarter, we are shifting recruiting practices of the brands acquired in the Anywhere transaction towards more productive agents. And while the total number of agents recruited is lower this quarter, productivity per agent continues to increase meaningfully and is clearly not impacting our ability to grow. Going forward, all the owned brokerage brands acquired in the Anywhere transaction will begin leveraging our enterprise sales team to recruit agents. Once this partnership model is fully set in motion by early 2027, we expect it to lead to a healthy level of agent adds and better productivity per agent. Importantly, total agent retention in our Brokerage business was 95.5%, which was largely flat compared to Q2 of 2025 on a pro forma basis and up 140 basis points compared to Q1 of 2026, where it was 94.1%. Consistent with last quarter, 72% of total agent separations in the quarter had no production or very low production in the trailing 12-month period. As a result, excluding agents with $0 GCI in the last 12 months, agent retention would have been 97.7% in Q2 and excluding agents with $20,000 or less in GCI in the last 12 months, which on average equals less than two transactions at our price points, agent retention would have been 98.7% in Q2. Commissions and other related expenses as a percentage of Brokerage segment revenue increased by 43 basis points to 82.2% from 81.7% on a pro forma basis a year ago. This is largely due to mix, whereby the majority of the revenue growth was in markets with lower overall margins were closed by real estate professionals that are on the higher end of the split schedule. For the Compass stand-alone brand, this commission metric was flat year-over-year. Franchise segment revenue was $135 million in the quarter, up 7.6% on a pro forma basis, and GTV was up 11.7% year-over-year on a pro forma basis compared to housing market volumes that were up 6%, which reflected out-performance led by our luxury brands, Sotheby's International Realty and Corcoran. Integrated services revenue reached $211 million in the second quarter, up 7.7% year-over-year on a pro forma basis, driven by strong performance across title and escrow. Our Cartus relocation business, which is also included in the Integrated Services segment, secured 15 new clients and expanded numerous existing client relationships during the quarter. We believe Cartus is well positioned over other relocation management companies as we integrate into the Brokerage operations and our Brokerage and Franchise operations benefit from the lead flow generated by the relocation side of the business. Our total non-GAAP operating expenses were $699 million in Q2. As a reminder, our Q1 non-GAAP operating expenses of $641 million excluded OpEx from the Anywhere business for the first 8 days before the January 9 acquisition date. So the increase in Q2 reflects the full quarterly run rate of our combined operations and some modest increase in variable compensation accruals as a result of our year-to-date out-performance as well as growth in the Integrated Services segment. Recall that compensation for our title and escrow officers in the Integrated Services segment is included in the operations and support line, and therefore, revenue growth in this business will have some corresponding growth in the OpEx line, but of course, that's coupled with accretive adjusted EBITDA. As Robert touched on earlier, we have continued to make strong early progress on cost synergies. This quarter, we pulled forward some additional synergy actions, and we've now actioned $300 million of our cost synergy target, which was our revised full year target for 2026. As we committed, when we hit our synergy targets, we won't stop looking for opportunities. We now expect to action as much as $330 million during 2026. Last quarter, we updated our expectations to realize about $200 million of the action synergies in our financials in 2026. We said about 2/3 of this amount or $130 million will be reflected as reduced operating expenses in 2026, benefiting adjusted EBITDA and cash flow and the remaining 1/3 or about $70 million will be reflected as lower CapEx, which won't directly benefit adjusted EBITDA, but will benefit free cash flow. As a result of pulling through some additional action synergies this quarter and a higher expected level of action synergies for the full year of 2026, we now expect approximately $150 million to be realized in OpEx, which reflects an increase of about $20 million from our prior expectations, but note that about 50% of this additional $20 million has already been realized in Q2. OpEx was essentially flat year-over-year on a pro forma basis. And one of the drivers of that is our cost to serve per transaction, which hit a record low within the Compass brand in Q2. However, as we consolidate our back-office systems with our other brands over the next year, we see opportunity to continue lowering the cost to serve per transaction through greater utilization of AI and deploying other best practices. Adjusted EBITDA for Q2 was $363 million, a record performance for any second quarter and almost triple the amount of adjusted EBITDA Compass generated a year ago on a stand-alone basis. As an interesting data point, the $363 million of adjusted EBITDA in the quarter reflects more adjusted EBITDA in one quarter than for the full annual period of 2025 or any other annual period in Compass' history. An early proof point of the earnings potential of the combined business. Adjusted EBITDA in the quarter includes segment adjusted EBITDA of $377 million from brokerage, $87 million from franchise and $52 million from integrated services, each reflecting improvements from Q1, both in terms of total dollar value and as a percentage of their respective segment revenues. As a reminder, while certain direct expenses are allocated to each of the three operating segments, there are additional expenses that are not allocated to any of the operating segments because they relate to more of the corporate entity or because they're shared across multiple or all of the operating segments. These include expenses related to our technology, finance, legal, human resources and executive functions. Therefore, the total adjusted EBITDA of the consolidated company will be equal to the total of the segment adjusted EBITDA for our three operating segments, less the unallocated corporate expenses of $153 million to equal the total adjusted EBITDA of $363 million for the second quarter. We achieved GAAP net income of $92 million this quarter, a significant improvement from GAAP net income of $39 million in the prior year despite the increase in depreciation and amortization expense, merger and integration expenses and interest expense resulting from the Anywhere transaction earlier this year. Our basic weighted average share count for the second quarter was 758 million shares, which was within our guidance range of 755 million to 760 million shares. As a reminder, the share count for Q2 reflects an anticipated step-up from Q1 as the shares issued for the Anywhere transaction were only weighted during Q1 for the period post the January 9 closing date, but they have a full quarter weighting in Q2. Free cash flow was very strong at $180 million for the quarter, resulting in $694 million of cash on our balance sheet as of June 30 and no outstanding borrowings on our $500 million revolver. The strong cash flow during the quarter was driven by the over-performance on adjusted EBITDA as well as some favorable timing on working capital items. In particular, our remaining cash liability from Anywhere's NAR class action settlement of $54 million was not paid during the quarter and is now expected to be paid later in 2026. Last quarter, I commented we would redeem the $500 million of our 9.75% notes in Q2 of 2027. The strong cash position as of quarter end, plus additional positive free cash flow anticipated for the second half of the year underscores our ability to achieve this commitment. We also stated that you should not expect to see any changes in our debt levels for the balance of the year or into Q1 of next year. That's because the highest cost tranche of our debt has a no call provision until April of 2027 and our 7% notes have a 1.75% call premium associated with them as well that drops away on the same date in April 2027. So we plan to continue to build our cash position through that date. And in the meantime, we have our cash invested in short-term treasuries with yields in the mid-3% range. It's worth reiterating some of Robert's earlier commentary on our net leverage ratio. Once again, we won't start reducing our gross debt levels until Q2 of next year due to the call provisions of the debt. However, our net debt levels are reducing when you consider our growing cash position. And when you combine this with our growing levels of trailing annual adjusted EBITDA, our net leverage ratio is improving. As Robert just touched on, if you take our net debt as of June 30, which is $2.45 billion and divide that by 2026 consensus adjusted EBITDA estimates of approximately $850 million, that yields a net leverage ratio of just under 3x. But since we expect to generate additional cash flow in the second half of 2026, the actual ratio as of year-end will be even lower. Turning to financial guidance for Q3. For the third quarter of 2026, we expect consolidated revenue in the range of $3.85 billion to $4.05 billion. We expect third quarter consolidated adjusted EBITDA to be in the range of $275 million to $305 million. And for the full year, we expect non-GAAP operating expenses in the range of $2.75 billion to $2.80 billion, which is an increase of $50 million from the guidance provided last quarter. The increase is primarily driven by $35 million of OpEx assumed from a recent brokerage acquisition we completed in early July; and secondly, some additional compensation expenses related to our recent over-performance. We expect basic weighted average share count for the third quarter to be between 767 million to 769 million shares, and we expect to generate positive free cash flow in Q3. Our cash flow in future years will be supported by the use of our net operating losses. It's worth a reminder that we have $1.8 billion of NOLs that will shield us from federal and state taxes as we generate taxable income. At our current combined tax rate of about 26%, which is a federal corporate tax rate of 21% plus about 5% of a blended state tax rate, that's about $470 million of potential cash taxes that will be avoided due to the utilization of the NOLs in the future. As we turn to Q&A, I want to say thank you to our entire team for the exceptional efforts and collaboration. We've all been working very hard for these past 6 months following our transformational merger with Anywhere in January, and it's great to see those efforts embedded in the outstanding results we're sharing today. And finally, we'll be participating in the Barclays Industrials Conference on August 12 in New York City, the Oppenheimer Tech Conference on August 13 and the Zelman Housing Summit on September 17. We hope to see you at one of those events. Operator, you can now begin Q&A.

[Operator Instructions] Our first question will come from Matthew Bouley with Barclays.

You have Elizabeth Langan on for Matt today. I just wanted to start off by asking for an update on your 3-phase marketing strategy. It sounds like you've had a lot of traction on the Redfin partnership. And I know last quarter, you had said that the coming soon had moved up towards the mid-30% range of your listings last quarter. How are you tracking against that? And maybe if you could give a little more detail on how the broader network effects are materializing.

Great. So on the Compass Brokerage side, the Coming Soons in July approached 57%. So for all new listings, 57% of them start off as coming soon, where we believe that we're on a path to have 80% of all new Compass Brokerage listings start off as coming soon in August and September. For the other brands, they are now onboarding to the platform. And so now they're -- I would expect the same trend line over time there as well. Really, there's no downside. There's no reason not to. There's no days of market, no price drop history. The worst thing that can happen is you get an offer. And you can say, no, I don't want the offer. I'm going to go to every site, not just the 60 million people that are buying off of Redfin.com in our brokerage sites. Remember, only 4 million people buy a year. So how many tens of millions of people do you need to see your property? And so that's why we're getting so much traction. And we're seeing a lot of demand for the private exclusives as well as the new brands come on because the same message is there for sellers. There's no downside. The worst thing that happens is you can go coming soon and go to the public markets, but with the benefit of price discovery as a private exclusive. At the core, what that tool is, it has many different benefits -- ways to use it. But at the core, it is a price testing tool in the exact same way that road shows that companies going public test price as the private listing is the exact same thing for people that are taking their property public and selling their property. And so the only people that -- what seller wouldn't want a price testing tool? The only entity that wouldn't want a price testing tool are people whose business models can't make money off of it.

Yes. That makes a lot of sense. And secondly, I wanted to ask if you could touch on cost synergies. You obviously spoke to you're not raising the $500 million target right now. But as you've made progress and you're actioning at a higher level than you were initially with the $330 million. How are you thinking about the upside there? And how should we be thinking about the pace in 2027 versus your prior expectations?

Yes. Thanks for the question. I think we're really happy with the progress made to date. In 6 months, we've made really strong progress on cost synergies. We're really happy with that. And I think as you consider the pace over time, we're getting into some of the deeper operational synergy areas, some of the system integrations. Some of these things are going to take longer time. And so you'll definitely see the pace slow compared to how quickly we came out of the gate for the first 6 months. But I think in terms of what we can achieve overall. I'd put you back to what we did in 2022 when interest rates started to spike with the mortgage rates. We made a commitment to take cost out of the business. We exceeded our initial goals on that. Then we exceeded the goals we put again on top of that. And so we -- the point being is that we're not going to stop looking for opportunities once we get to the $500 million goal. And so I think it's safe to assume or reasonable to assume that based on the pace that we've come out in the first 6 months that we will hit our goals over time and potentially exceed them. But we're not going to go into details as far as 2027 at this point in time.

Your next question will come from Jason Helfstein with Oppenheimer.

Robert, the lead stats in Chicago with Redfin are quite impressive. I think while Chicago has some uniqueness, maybe just talk about your ability to replicate this in other markets. And then just when we think about transaction per agent, obviously, we can do the math on Q1 to Q2. Just how much do we -- upside do we see in transactions per agent if we get back to kind of a normal market versus the current level? So where do you think kind of peak upside is there?

I'll address Chicago and the broad opportunity, and then I'll pass it on to Scott for transactions per agent. So on Chicago, we're really happy to see that. Also, it kind of it shows what happens in a free market without restrictions, without fines, without bans. What it shows you is why wouldn't an agent put it on their sites first? What's the downside? Why wouldn't you want -- what seller doesn't want the agent to get their own buyer -- the buyer inquires and deal with them directly. And to be -- to see our traffic up over 100% year-over-year, 77 basis points over the rest of the regions. It's giving us even more confidence and more conviction. And it's also making our real estate professionals even more excited to continue to make progress here as we will. I don't think Chicago is that unique. To be clear, it is becoming the norm. MLSs have statuses that allow agents to share their listings with all agents across all brokerage firms and decide where it is publicly on their site. And so what I mean by that specifically is now 60% of our markets, but it's now in the 60s, they will allow agents to decide to have it on their site and as long as they're sharing it within the MLS across all brokerage firms. And so we think this is going to become more and more the norm as agents see the value in it for their sellers.

Yes. And Jason, just to touch on the question on transactions per agent. Tough question to answer as far as a specific data point. But as a trend, I think you'll see the number continue to go up just as you did this quarter. And the reason for that is that we're turning out or trading out of the business, some of the unproductive agents. In fact, we cited stats today that are similar to the stats we cited last quarter, where 49% of the agents that left the business in the quarter had 0 GCI in the trailing 12 months. So naturally, it's just going to increase transactions per agent. There's a lot of mix in there, too, because remember, we're now disclosing total agent counts, not just principal agent counts. And the total recruiting number each quarter and the total attrition number is really not 100% in Compass' control. It's also in control of the principal agent, right? The principal agent is who decides to add team members to its team. It's who decides to attrit team members from its team. So that goes into that churn as well. But I think directionally speaking, you'll see that number go up as we, by definition, are only recruiting productive agents to the company. And as a result of the data you saw in the last two quarters, we're trading out on average, the lower-performing agents.

Your next question will come from Kunal Madhukar with Deutsche Bank.

A quick one on the outlook that you provided. Given the K-sided or K-shaped market recovery or the economy that we are seeing right now, as the market recovers, what is your assumption on market share and especially around the recovery part? So if the market recovers from the current $4 million to like, let's say, $5 million or $5.5 million, if you're assuming the same market share going up in that environment, -- is that -- given your bend towards higher-priced properties and more luxury properties, would that assumption be correct?