Description

Compass, Inc.

The short, short thesis is as followed:

• Compass has long been a VC-fueled revenue-acquisition business, in an extremely competitive industry, masquerading as a “technology platform” 

• COMP’s primary assets (realtors) are not proprietary and probably only loyal so far as they are paid market-leading compensation

• COMP is unprofitable even in the second-hottest real estate market of the past half century, but trades at a big sales premium to peers with better business models (e.g. franchise)

Background

Compass was founded in 2012 by former banker Robert Reffkin and Israeli tech entrepreneur Ori Allon as Urban Compass, a NYC-based real estate brokerage. Initially Compass launched with the idea of using “neighborhood specialists” on a salary-based agent model (akin to Refin) and making urban rental property discovery much easier for renters, powered by data. By 2014 the company pivoted to a more traditional model. Compass soon gained traction in NYC by offering outsized incentives to top agents/teams at competing brokers. This was fueled by over $65M in Series A/B funding within a year of launch. 

Compass also began an M&A-driven expansion model in November 2014 with the purchase of a DC-based agency. This was replicated many times over, with Compass simultaneously claiming that its tech platform gave it the ability to enter new markets quickly. By late 2016 the firm had raised over $200M and was valued above $1B. Softbank led a massive (for the industry) $400M Series E in December 2017 and another $400M Series F less than a year later. 

Compass IPO’d at $18 in April 2021 after repricing down, ultimately raising $440M. 

CEO Reffkin still holds 15M class C shares (20 votes) giving him close to majority voting power. 

While already a broken IPO, the short could have plenty more room to run as the real estate market softens and/or profit-based valuations matter to stocks like COMP again. 

Questioning the Business Model

Compass has long sold the story of building a transformative tech platform for real estate agents. Plenty of buzzwords are bandied about --AI, Machine Learning, cloud, SaaS--but to these eyes Compass looks much more like a classic Softbank-overfunded startup, with huge growth outlays in search of scale, and profitability an oversight. While COMP has reached 5% market share and much higher in key markets, it has also accumulated $1.4B in losses in its relatively short history. 

When it comes to building a tech platform, there is scant evidence behind the hype. Early hype included talk of AI-driven local discovery, and integrated hardware. But even a couple of years ago, it seemed Compass’ biggest digital innovation was the introduction of digital “for sale” yard signs…which cost agents $700 and required manual updating…In the latest quarter, they promote the introduction of a “video messaging” service and AI-driven video marketing features. While a step, it does not sound like a giant leap forward. 

[see  https://archive.curbed.com/2018/7/19/17589110/for-sale-sign-compass-tech]

The core short thesis is that Compass’ key assets are not its tech stack, but its agents. And COMP’s positioning in major metro markets, with high avg. homesale price ( $1.1 million, 3x national average) and productive agents (20 transactions per agent) is actually, in some ways, a risk. Getting these highly productive agents allowed Compass to grow faster, but they are fiercely competed for via incentives and commission splits, and are not proprietary to Compass by any means. Agents are driven by minimizing commission splits and broker costs, and competition is fierce. To win them, Compass has had to offer more favorable compensation schemes. Furthermore, top performers often have the least to gain from brokerage houses, and greatest resources and ability to go independent or limited-service models.

Tech Platform? Little Evidence 

If COMP were really an advantaged tech platform, we would expect to see it show up in higher pricing power i.e. lower agent commission splits, given COMP is providing more value to the agents. But we see the opposite. Despite the technology wrapper, COMP’s real agent value proposition has long been generous payouts. 

Total commission and related expenses were 82% of revenue last year, at the high end compared to peers and leaving little room for profit. This includes commission splits but also signing bonuses and discounted equity programs, which looked particularly attractive given COMP’s steadily inflating venture valuation, and ramped up prior to the IPO. 

For comparison, Realogy’s company-owned brokerage unit (formerly NRT), a close comp, has recently paid mid-70s percentage splits (with this edging up as competition has risen). Douglas Elliman, which has even higher avg. homesale price ($1.5M), paid out <73% of revenue in commissions LTM. 

COMP is still paying out ~82% of revenue in the form of commission and related expenses in 2021, demonstrating no leverage in a hot market. This could come down in the coming quarters as stock comp gets revalued down, but if that plays out it would be small consolation for longs…and a weaker currency makes it harder to use stock comp as a recruitment or retention tool. 

In fact, it actually appears that a substantial amount of agent options will be repriced at market in early 2022 which could mean substantially more dilution given the shares were valued at $6.4B in the 2019 funding round.  

Recent Performance & Outlook

Even with 47% revenue growth, COMP reported a massive $1 million Y/Y jump in Adj. EBITDA in 3Q21 to $12MM. This despite being a seasonal peak, in an on-fire housing market only surpassed by the housing bubble. After raising guidance for the full year, COMP still anticipates $(25) to $(5)M in Adj. EBITDA losses in 2021, translating to $(55)-$(75)M in Adj. EBITDA losses in 4Q21. 

Compass is “confident” in achieving positive Adj. EBITDA in 2022, but this will require either continued housing market buoyancy or cuts to overhead. COMP is in a hard place, as the vast majority of costs are tied to agent retention and the current valuation is based on continued outsized topline growth. 

Valuation - Upside Looks Limited

Even with the broken IPO, COMP still trades at approximately 3.5x gross profit (sales less commissions), and ~100x 2022E Adj. EBITDA. Assuming normalized margins closer to 6%, the Adj. EBITDA multiple would still be low double digits, in what is likely a cyclical peak operating environment. By comparison, Realogy trades at mid-single digit multiple as investors are reading through the current environment. Although Realogy has been bleeding share, its franchise business is more stable and probably deserves a higher multiple. Other traditional brokers like Re/Max and VGR’s Douglas Elliman have similar real or implied valuations. 

COMP could eventually rationalize its overhead, but even here upside looks limited. For margin comparison, Realty’s company owned brokerage division (former NRT segment) is probably the best comp--a fully owned (not franchised) brokerage house, 4% market share, concentrated in the major metro coastal markets with ~2x national avg. transaction size ($660K).  Operating EBITDA margins, pre-corporate (and after backing out inter-company franchise fees) have historically run around 7%. This despite paying 5-7% lower commission splits than COMP, meaning the cost structure would translate to almost zero margins on COMP’s current compensation table. Likewise, Douglas Elliman’s historical Adj. EBITDA margin is LSD (rebounding to 8% LTM), pre-standalone costs, despite even lower commission splits. 

All this is to say that COMP will have to cut back commissions over time to reach profitability, but this will be a tradeoff for lower growth and quite possibly eventual net agent loss. 

While I am not going to speculate exactly where COMP’s margins and multiples ultimately settle, it is clear that anything close to peer levels implies meaningful downside remains.

Cyclical & Secular Factors

As well known, the pandemic has spurred a “great reshuffling” which has powered existing home SAAR to 6.3M from 5.3M pre-pandemic and prices running up mid-teens, further aided by ultra-low mortgage rates. In short, the market is hot, although pricing growth is decelerating and transactions stabilizing...trends that should continue or worse as rates rise. 

Compass is still showing strong market share growth, but this cannot continue forever if profitability becomes a factor. Market share was 5.4% in 3Q21 (up from 4.1% a year earlier) and principal agent count of 11.6k was up 31%.Compass already covers just under 50% of the population with a presence in 67 metro areas, leaving dwindling room for geographic expansion given its target segment. 

Shareholder Turnover + Insider Sales

In October and November, Compass net settled all 6.8 million shares that had vested under RSU plans as of October 31, withholding 4.7 million for tax liabilities. Effectively, insiders cashed in all available options, mostly Rifkin’s, although he still owns a big stake. Selling pressure could be heavy for a while.

Catalyst

Real estate market cooling off

Lack of profitability forcing restructuring, lower commissions, leading to vicious spiral of lower agent retention and recruitment, lower growth and profitability

Continued insider selling, agent selling