Redfin: "Still Hot, Just Not As Hot As Before"

(Redfin Q2 FY21 Earnings Review)

Personal Equity Research

Aware that another earnings season is underway, with banks having reported their Q3 earnings last week, I wanted to share my quick thoughts on Redfin’s second quarter before most of my companies begin to report in the coming weeks.

In a US housing report shared by Goldman Sachs on October 11th, GS would claim that “of all the shortages afflicting the US economy, the housing shortage might last the longest”.

Goldman attests there are some concerns around the “sustainability of homebuyer demand, as potential homebuyers seem to have taken note of the sharp rise in home prices: 66% of respondents to the University of Michigan consumer survey now think today is a bad time to buy a home, the highest share in nearly four decades”.

Homebuilders are still hampered by various supply constraints, including but not limited to; lack of construction workers, supply chain disruptions, lumber shortages, and economy-wide labour shortages. These issues are set to “limit the pace of annual homebuilding to around 1.65mn in coming years”. Goldman believes the housing price boom will continue for multiple years and cite government intervention (in the form of loosening barriers for homebuilders) as the most crucial step towards easing the housing problem.

“The market is still hot, just not as hot as before” remarked Redfin’s Glenn Kelman in the second-quarter earnings call, citing that whilst prices still remain elevated, the supply of housing inventory has rebounded modestly to offset the demand in home sales.

This is certainly true. Housing inventory for Redfin homes, up 22% from the March lows, has improved but remain below historic averages for this time of the year. Prices for homes have now levelled off over the last three months with the median sale price of homes declining for two consecutive months from June ($386K) to August ($380K). This is the most up to date data I have.

Minimal, but it has stopped climbing. Data suggests that, over the 4 weeks ending July 18, a higher proportion of active listings are reducing their prices each week (4.3%) compared to the same period in the previous year (3.6%).

Median days on market is still relatively flat over that period, up to 16 from 15, but still historically low.

Overall, Kelman seemed optimistic that the housing market would continue to settle down as Redfin head into the second half of the year, partly due to the hesitation from buyers as prices border on the absurd and the healthier mix of buyers to sellers.

He states that “Listing clients who were once insistent about pricing-in future months of appreciation now acknowledge the market may have peaked”.

For continuity, I must state that Redfin is a position I currently hold in my portfolio (weighted at ~3%).

Section 1: Performance - A glance at the performance of the second quarter, complete with some of the core takeaways including; RedfinNow as a profitable entity, Mortgage pressure, Attrition, Website traffic, and Redfin Premier.

Section 2: RentPath - A discussion on the future of Rentpath, how this will integrate with Redfin’s brokerage business, and my core takeaway for the coming year.

Section 3: Liquidity & Cashflow - A brief discussion of Redfin’s balance sheet as of the second quarter.

Section 4: Guidance - Commentary around Redfin’s third-quarter projections.

Section 5: Concluding Remarks - My final remarks, and some discussion surrounding my own position in Redfin.

Section 1: Performance

Revenues came in at $471M (+121%), above prior guidance of between $446M to $457M. I suspect that both RedfinNow ($172.4M) and Rentals ($42.5M) revenues contributed to the upside given that Redfin previously forecasted ~$154M and ~$41.5M for both segments, respectfully.

Elsewhere, Redfin’s real estate services contributed $252M in revenues and the ‘other’ segment (includes mortgage, title and other services) generated $8.5M.

RE Services: Real estate services benefitted from record levels of transactions both from Redfin (21K) and Partner (4.6K) agents in tandem with strong revenue per transaction which stood at $9.85K on an aggregated basis, up 30% from last year. The average number of lead agents, which sunk to lows of 1.4K in Q2 2020, ended the quarter at an all-time high of 2.46K.

RedfinNow: Redfin’s I-Buying business sold 292 homes during Q2, the strongest quarter to date. That, alongside the 171 homes sold in Q1, puts Redfin’s HY sales at 463, ten more than the 453 they sold for the entirety of 2020. Revenue per home sold also reached a high of $570.9K in Q2.

Rentals: After the acquisition of Renthpath closed in Q2, Redfin reported $42.5M in revenues from this segment. I have carved out a section to discuss RentPath in greater detail.

Other: Growth for Redfin Title and Mortgage was weaker than anticipated after the company paused their title business to realign operations, as well as slower than expected adoption in their mortgage business. Still a tiny portion of the business, for now, so not a huge concern to me personally.

The inclusion of RentPath creates a new dynamic in the product mix for Redfin. Second-quarter revenues are down ~12% on the year, but recovery efforts aside (will discuss later) the rental business contributed 9% towards revenues this quarter and is expected to be ~7% to 8% of the third quarter’s revenue base ($40M to $41M).

With a margin profile greater than any aspect of Redfin’s business (82%), the rental segment accounted for a little over one-fourth of the gross margin in Q2. Of the $126M in gross profit that Redfin generated this quarter (+174%), RentPath accounted for $35M of that.

Despite not yet being a profitable business ($13M in losses in Q2), it does offer up attractive opportunities alongside their core RE services business and the considerably lower margin RedfinNow unit, which accounted for ~37% of the quarter’s revenues but just 4% of the gross margin.

Redfin’s quarterly gross margin has oscillated between 6.7% at the lows of 2020 to as high as 39% back in Q3 2020. This quarter, the gross margin settled at 26.8%, up 530 bps from last year.

Portraying gross margin on a TTM basis, we see the gradual decline that coincides with Redfin’s launch of RedfinNow which was accretive to the top line, whilst being dilutive to gross margin (at least, until 2021).

Kelman’s remark about the “emergence of RedfinNow as a large-scale profitable business” implies that he feels the business will continue to be contributory to the gross line going forward. That, alongside Rentpath’s 80%+ margin profile, provides me with greater confidence that Redfin’s margin will be more sustainable (or perhaps more predictable) in the coming years. At this time, I am still uncertain about where that margin will eventually settle. Given the rate of change amongst the complementary business segments, and the fact that mortgages and titles are yet to become consistently profitable ventures, it’s more or less a ‘wait and see’. Without a mystic ball, I am primarily interested in whether the segments are trending in the right direction.

With respect to operating expenses, Redfin’s $156M expenditure (leading to a $30M loss from ops) accounted for 33% of revenues (up 900bps), primarily as the result of RentPath and timing of marketing.

The Rentpath business contributed ~$48.6M of the $106M in additional expenditure YoY, accounting for 31% of Technology and Dev, 23% of marketing, and 39% of G&A expense in the second quarter. Excluding Rentpath’s gross profit and expense, Redfin would have still been ~$17M in the red for the quarter.

Even without Rentpath’s inclusion, marketing expenses would have been up ~350% on the year, at $42.8M. Two considerations here. One, Redfin postponed the launch of their annual TV commercial from Q1 to Q2. Two, in the second quarter of 2020, Redfin abandoned essentially all non-essentially advertising, including TV. The uplift in marketing is real, but context is needed.

After modest interest expenses, and a $5M tax benefit from the Rentpath acquisition, Redfin’s net losses came in at $27.8M for the quarter and $63.3M for the year thus far.

1.1 High-Level Takeaways

1) RedfinNow Business Shows Profitability

Back in Q4 2020’s earnings call, Kelman would remark that Redfin “remain optimistic that RedfinNow will contribute to 2021 gross profits”.

After two consecutive periods of positive gross margin contribution, he would celebrate "the emergence of RedfinNow as a large-scale profitable business". For the trailing six month period, the RedfinNow unit has established a gross margin of ~2.5%, for a modest $6.6M in gross profit.

Larger scale I can understand, but a large scale profitable business is an overstatement. Nonetheless, Kelman has delivered what he said he would back in February 2021.

More importantly, why is RedfinNow suddenly delivering?

A few reasons.

  • Disciplined Offers: Redfin began to exercise greater pricing discipline in their offers, typically offering less per home whilst still being able to win deals. Around 700bps of the upside from Q2 2020’s -1.6% gross margin came from home cost efficiencies. This was offset by greater personnel (-1.7%) and other expenses (-0.8%).

  • Scale: Through the process of launching the RedfiNow service into new domestic markets, Redfin’s coverage through the I-Buying unit has expanded from 64% in 2020 of listings to 80% at the mid-point of 2021. Kelman cited incremental leverage from scaling as a core reason the RedfinNow business is earning a positive gross margin this year.

  • Market Tailwinds and Service: Kelman would also cite that he thinks “there is some secular demand growth where just more people are interested in liquidating their property that way, less people want to deal with the house without listing a home”. This is, essentially, the RedfinNow bull case so not much further explanation is required. He would also remark that investments in service quality are allowing Redfin to respond faster to inquiries about I-Buying and for those who reject offers, they are generating greater traction in converting those rejections into a home listing on the brokerage site, demonstrating the benefit of the operating a brokerage and I-Buying business.

I would imagine that if (emphasis on if because I am no RE expert) pricing expectations subside, the market becomes a tad more docile, and the number of transactions in the RedfinNow business behind to slow down, causing downward pressure on RedfinNow margins. So, I wonder if Kelman is suggesting that greater responsiveness and service efficiency gains in face of offsetting market conditions are the play here?

Especially considering his commentary surrounding the need to improve technology to determine which houses are more likely to sell faster and the difficulties in attracting labour to renovate the homes they eventually flip. There have been recent signs of fatigue elsewhere in the market.

Redfin’s I-Buying division is tiny in comparison to Zillow and OpenDoor. So far this year, Redfin has sold 463 homes compared to OpenDoor’s 5,943 and Zillow’s 4,051.

This weekend, it was reported that Zillow would cease purchasing new homes for the remainder of the year to focus on burning through the 3,142 homes they have left in their inventory. A spokesperson for the company would cite they are “beyond operational capacity in our Zillow Offers business and are not taking on additional contracts to purchase homes at this time”.

OpenDoor, which has 7,971 homes in their inventory, are yet to pump the brakes.

As far as 2021 is concerned, the Zillow Offers business (their I-Buying unit) still reports a positive gross margin (making an average of $33.8K in gross profit per home in Q2). Furthermore, housing inventory takes up just 18.6% of current assets and 13.3% of total assets, which is not too abstract from Redfin’s exposure.

It is my gut feeling that this may have more to do with Zillow’s poor operational practices than a sign of contagion for OpenDoor and Redfin. After Kelman’s commentary surrounding RedfinNow in the earnings call, there were no signs of operational bandwidth being stretched too thin. In fact, quite the opposite, with greater scrutiny being placed on how they submit offers.

2) Mortgage Pressure

The mortgage industry is currently enduring a phase in which all mortgage suppliers are being hit with downward pressure on margins with loans being priced more aggressively. Kelman candidly admitted that it is “going to be tougher times ahead for the lending industry overall”.

Despite being a smaller aspect of the business, I suspect the Redfin Mortgage will blossom eventually, provided attach rates blossom in tandem. Mortgage revenues would grow 47% in Q2, outpacing the other segment’s 18% growth YoY, and now covers 81% of Redfin’s brokerage purchase transactions with ~94% expected by the end of the year.

Despite strong growth, revenue decelerated from Q1, where it grew 200% on the year. This was largely the result of a pause in the hiring of sales personnel in an effort to improve service quality. Given that NPS scores rose from 58 to 73 from March to May, Kelman has stated that additional “hiring should lift transaction growth in the fourth quarter and beyond”.

This transaction growth is crucial, given that revenue per loan declined 5% YoY whilst the average cost per loan has increased. Despite acknowledging Redfin Mortgage will no longer be likely to turn a profit in FY21 in the face of market pressure, Kelman claimed that they still “expect Redfin Mortgage over time to be a major source of profit”.

3) Attrition

The way that Redfin employs agents will likely be a hot talking point for years to come. Offering a heavier starting salary in place of additional commission-based earnings is enough to turn away seasoned agents, but offers an attractive level of security for a newer entrant to the RE agent world. This has its puts and takes.

Historically, Redfin’s employee churn has not been overly concerning but last quarter, attrition amongst new agents rose to an annualised rate of 53%, up from 26% in the prior year. Second-quarter attrition declined to an annualised rate of 49% but still remains elevated for new hires.

New hires have always been the most likely to churn and whilst Redfin do not disclose the composition in the tenure of their agents, the increased churn in new agents appears to be influencing consolidated attrition, which rose to as high as 37% in Q2.

Without creating excuses for Redfin, there are a few things that spring to mind as I digest this data:

  • Agent Growth: Lead agents have grown 75% YoY, and 8% sequentially, now sitting at 2,456. I suspect a new agent is one with a tenure of <1 year. As such, it would appear that a significant factor in this year’s attrition is the exceptional hiring velocity over the last twelve months.

  • Hiring Mistakes: Without disclosing the composition, Kelman remarked that this is the highest proportion of new agents they have had in years. Citing that Redfin likely made a slew of hiring mistakes during this time, Kelman would suggest “nearly half the people who left in the second quarter are people we wouldn't choose to hire again”. I am not sure how they measure that, but it added some colour for me. Of those who left in Q1, and took an exit survey, 20% claimed they were leaving real estate altogether.

  • Compensation Competition: Earlier this year Glenn humorously suggested that there were more agents than listings in the United States. Given the state of the RE market and increasing competition in the digital brokerage business, well-equipped competitors are simply bidding up compensations for agents, causing churn there.

These factors have led to greater attrition in 2021. Despite the number of lead agents climbing 75% YoY, transactions per lead agent have declined 13% (to 8.6) over that same period, meaning more agents are fighting over fewer listings.

The answer?

Kelman has acknowledged that this will likely persist until the start of the new fiscal year. For now, Redfin has eased performance standards and offered one-time bonuses to new agents based on submitted offers rather than closed sales. Looking ahead, they plan to roll out a comprehensive pay update for buyer's agents in January 2022. Proposed changes to the pay structure are not expected to meaningfully impact brokerage margins.

4) Rising Competition for Visitor Traffic

Taking the average of both Q1 and Q2, Redfin has amassed 47.3M monthly average visitors, up ~10% from the end of last year.

I highlighted some comments concerning the loss of share to the likes of This will apparently persist until Redfin import more local neighbourhood data.

When prompted in the earnings call, Kelman would state that this is mostly due to the way that Google’s algorithm ranks The response can be found below.

Thankfully, it would appear that Redfin’s revenue growth this year has largely been a factor of increasing the supply of available homes, as opposed to search traffic or a greater number of customers. Penetration for the digital brokerage still looks strong, with Redfin now capturing 1.18% of the US market share, up 118bps from 2020.

5) Redfin Premier

Another area I made some notes, was in relation to Redfin’s Premier service for homes valued at $1M or greater. The number of $1M+ listings on Redfin has increased from 5% of the total listings to 13% in the past year. Granted, some of this will be a result of the market environment pushing certain homes over the threshold, but in response to the growth in luxury housing, Redfin decided to pivot slightly towards upper-market RE, after having been almost entirely focussed on mid-market.

Demand for consultations in the Premier price range grew 111% YoY, compared to 33% on the aggregate level. The service now covers 85% of those Premier listings, so this is more of an observation on my end, rather than anything conclusive or actionable.

Section 2: RentPath -

Very quickly, here is a breakdown of Rentpath’s financials in Q2. Revenues of $42.5M (down 12% YoY), gross margin of 82%, with a negative operating and net margin. I have already discussed the impacts Rentpath has had on the consolidated business.

This post is for paid subscribers