Description

3/26/24

eXp World Holdings

Ticker: EXPI

Stock Price: 9.91

S/O: 161.95

Market Cap: $1,562m 

EV: $1,393m

Short Interest 19,018,994 shares - source: NASDAQ

Short Interest Ratio 14.73 Days to Cover

Short Interest % Float 24.42 % - source: NASDAQ (short interest), Capital IQ (float)

Background

eXp Realty is the pioneer of the cloud-based real estate brokerage business. With no physical offices and a cloud-native platform, the business model is far different from that of your traditional real estate brokers, including Keller Williams, HUNT, Coldwell Banker, and Compass, who have physical offices from which their brokers work. 

The brokerage business is simple: eXp and the agent work off a revenue share agreement. In eXp's case, the split is far better for the agent than what is typically offered. At eXp, the agent generates a commission from acting as an agent in a buy/sell transaction. The historical rate for the commission was ~5-6%, split 50/50 with the buyer's agent. 

At eXp, agents retain 80% of commissions, while eXp retains 20%. In addition to the 20%, agents would pay eXp $25/transaction for a transaction review fee and $60/transaction for a risk management fee. Beyond this, a broker's only monthly cost is an $85 cloud brokerage fee to access the eXp platform. Startup costs for new agents are also next to nothing, a benefit for brokers where there could be little revenue for several months during the initial launch. At Keller Williams, the commission split is typically 64/36, with an $18,000 cap.

Delving deeper into the business model, once an agent hits $16,000 in commissions paid to eXp, they no longer operate under the 80/20 model. Instead, they retain 100% of the commission, a great incentive. The only fees paid to eXp are the aforementioned fees and an additional $250 transaction fee. EXp's historical take per transaction is ~$1,700 (excluding fees). Thus, the $250 fee is, on average, a net $1,450 extra for the agent. 

While eXp's model attracted many agents, it's recently hit a stalling point. With significant new headwinds recently transpiring, eXp's business model is no longer as promising as it once was. 

As shown in the chart above, eXp saw significant agent growth through 2022; however, the frozen housing transaction market has led to unproductive agents exiting the business. Q4 2023 marked the first decline in agent count from 89,156 agents to 87,515. Management discussed this decline, stating that having unproductive/inactive agents on their platform costs them money. They offboarded more agents in Q1 and said they should be finished. 

Thesis

eXp's business model was broken from the recent National Association of REALTORs (NAR) announcement on March 15. In the traditional model before the settlement, when a home was listed on the Multiple Listing Service (MLS), the seller's agent typically offered a commission to the buyer's agent. This commission was a part of the overall listing information in the MLS, and it was a standard practice in the industry for the seller to indirectly pay the buyer's agent through this system. For example, the buyer's agent commission rate (2.5-3%) would be stated under the listing on the MLS.

However, with the NAR settlement, this practice will change significantly. The settlement eliminates the requirement for listing brokers to offer upfront compensation to buyer's agents through the MLS. This means the previous standard approach of listing a commission in the MLS for the buyer's agent is no longer mandated. Previously, the seller would pay for the buyer's agent. Once the new regulation goes into effect, the buyer pays, or an agreement must be negotiated outside the MLS. In addition, the agent's commission cannot be rolled into a mortgage; instead, it must be paid out of pocket. 

The recent NAR proposed settlement is fundamentally changing how brokers are paid (primarily buy-side agents(agents representing buyers of a house)). Commissions paid to agents will likely decrease over the next year as these new regulations are fully enacted. The implementation timeline for these new regulations is the summer of 2024.

EXp's buyer agent business model will materially change because of the changes in the commission structure. The seller will no longer pay the buyer's agent commission (unless negotiated), which should result in lower revenue generated from buy-side transactions. New agreements will likely be a flat fee or a lower percentage compared to the historical 2.5-3% fee. Exacerbating this issue, the Q3 2023 earnings call management stated, "So we're a little bit skewed to the sell side. So Jeff can probably give you a little – he might have better data. But you can think about it as 65% or 60% – 55% to 60% being list side and then the balance being buy side." Wait… that should mean eXp is more insulated since they represent more sellers than buyers. On November 28, they released an 8-k stating the following, "During the Q&A portion of eXp World Holdings, Inc.'s (the "Company") earnings call held on November 2, 2023, members of our management indicated that the Company's historical transactions have skewed to the sell-side, and the Company retracts that commentary. The Company's transactions have fluctuated historically and since 2019, the Company's transactions have consisted of slightly more buy-side transactions than sell-side transactions." Therefore, over 50% of eXp's revenue is generated from buy-side transactions. Thus, the new regulations will have an outsized effect on them relative to a company that skews more sell-side transactions. 

How does this all translate through the financials? First, I expect continued exits of lower-producing brokers, resulting in top-producing brokers receiving more transactions. Since eXp caps the commissions at $16,000 again, this has an outsized impact: they generate almost no incremental profit when an Icon or capped agent closes another transaction. 

Revenue is reported on a gross basis (before the 80/20 split). Backing into the COGS figure not associated with agent commissions, where the first 80% of revenue is paid out to brokers, and $16,000 of stock is paid out to Icon agents, historically, 88-89% of COGS is associated with agent commissions.

The remaining 11-12% of costs are likely associated with closing costs and other back-office tasks. These costs are relatively variable and have consistently been between 88-90% of COGS since 2018. We can, therefore, model these costs as a percentage of revenue, which I did at 11%. 

Next, and the crux of the thesis, is the take rate eXp agents received per transaction. Taking the transaction value/revenue gives a ballpark figure of the take rate, around 2.5%. After backing out the per transaction revenue and $85 monthly cloud brokerage fee, the legacy commission rate is relatively the same at 2.4%. This take rate is headed lower due to the new rules; the question is to what degree. Suppose the take rate goes to 2%, implying a 1.5% buyer fee and a 2.5% seller fee. In that case, you'd get a 2% average take rate (assumes 50/50 buy-side/sell-side transactions. eXp historically skews slightly towards the buy-side, making this a conservative estimate). I suspect buyers skew more towards a fixed fee. A 1.5% fee on the average transaction value of $360,000 (eXp's 2023 average transaction value) equates to a $5,400 fee. This 1.5% buyers fee is likely biased on the high side, given that buyers could hire a real estate attorney for $1,500 to complete a transaction while finding the listings themselves. Thinking about the second impact of this, agents will now need to complete more transactions to reach the cap. 

From the Stephens fireside chat, Glen Sanford(CEO) stated, "The top 3% to 4% of our agents are these icon agents and our mega teams. And then, we've got the – the 25% who cap, we've got about 50-ish percent plus maybe 50%, 60% that are producing, but don't cap, and then we've got about 20% who don't do any transactions, which is where the biggest part of our churn is, is in that non-producing category of agents." Icon producers are the broker's top producing agents. To become an Icon, an agent must reach the $16,000 cap and complete 20 transactions after that. In doing so, they receive eXp's 20% split back in stock. 

22% of agents reach the $16,000 cap and, on average, need to complete 8-9 transactions to it (significant home price appreciation lowered this from 13-14 in 2017). To achieve the $16,000 cap at a 2% take rate, agents must close 10-11 transactions. Note that the increase in transactions has almost no revenue impact. In addition, I suspect most capped agents will get to the cap again through lower-producing agents exiting the business and increasing transactions. Lower-producing agents leaving eXp directly lowers revenue. Transactions left by non-capped agents to capped agents result in eXp losing the 20% commission split on those incremental transactions. 

While the market has punished EXPI over the past few years, including lately, analysts and consensus have not worked through the math of a 50bps decline in the take rate to eXp's business. Note, since the new regulations will not come into effect until mid-July 2024(assuming it passes, which is very likely), the actual impact will start showing up in the back half of the year and into 2025. I estimate the culmination of these changes is a $700m revenue headwind in 2025 and a $775m headwind in 2026. The real EPS impact flows through on the gross profit where I have these changes impacting gross profit negatively by $140m and $155m in 2024 and 2025. Unsurprisingly, I am well below the street, currently at $5.13B in revenue in 2025, compared to my estimate of $3.79B in 2025 and $4.17B(no estimates). On adj. EPS, I have them generating ($0.12) in 2025 and $0.20 in 2026 compared to consensus at $0.29 in 2025. 

Catalysts

eXp has been a rapidly growing business until late. The current valuation assumes eXp returns to agent growth shortly. The headwinds of agents leaving the industry result in a muted agent count growth over the next several years, offset by new agents/teams moving from legacy brokers to eXp. Transaction growth will not be enough to generate significant revenue growth due to the capping agreement. In previous earnings calls, management stated that the cap is more likely to decrease than increase. I don't think the business model is currently in a position to do anything like this. Still, if management thinks they need to do this to attract more agents, it's not out of the realm of possibilities. Should they do this, It would be a significant EPS headwind. 

When transactions rebound, but eXp shows little EPS acceleration, the market will dramatically lower the multiple assigned. On a GAAP basis, they were only profitable during 2020-2023(2023 was marginal) when they saw significant agent count growth. I expect them to post GAAP losses (stock comp is a real expense) in 2024-2026. 

While agents loved receiving stock when the chart was up and to the right, the 88% drawdown likely has agents reconsidering the type of compensation they would like. This could result in top agents leaving as their stock is worth significantly less than before.

While eXp was the only cloud-native broker for the last decade, new competitors are entering the game. REAL broker (recently on the planet microcap podcast – ticker REAX) offers agents an even better incentive structure with a $12,000 cap and 85/15 split. This new entrant, while small, could force eXp to change its compensation structure to retain talent. 

On March 21, reports alleged that two eXp agents drugged and sexually assaulted female co-workers. I have no information regarding the merits behind these allegations; however, such a headline could deter agents from moving to eXp due to concerns about the work environment.

Finally, apart of the NAR settlement, within 120 days following preliminary approval of the Settlement Agreement by the Court, eXp must deposit into an Escrow Account an amount equal to 0.0025 multiplied by its average annual Total Transaction Volume over the most recent four calendar years ("Total Monetary Settlement Amount"). This figure is roughly $360m, which eXp does not have and would need to figure out a way to raise. Compass Realty recently settled for ~10% of their eXposure ($517m in eXposure settled for $57.5m). eXp, under a similar settlement, would owe ~$40m. It's not a large sum, but it is likely the minimum they'd owe. 

While the stock is already down significantly, these negative catalysts and the continued negative earnings revision cycle point to further downside. At 15x, my 2026 adj. EPS estimate of $0.20, I arrive at a price target of $3 or ~70% downside. 

Risks

Productive agents leave other brokers for eXp

eXp settles for less than the $40m figure mentioned previously

Agents find ways around the new rules

Catalyst

Agent count decline

Negative earnings revisions

NAR payment comes in higher than expected