Description

I am recommending a short position in Lendingtree Inc. Its business is deteriorating, consumers seem to hate its services and competitors have structural cost advantages. The company is overvalued on the back of promotional management. Insiders recently started selling in large amounts. Lendingtree currently trades at 3.5x revenues with minimal/no growth and breakeven operating cashflow.

Lendingtree (previously Tree.com) has already been covered by Lukai back in June 2012 as a long position. However, share price was $9 at a time compared to $49 currently. I have submitted this as VIC acceptance idea back in November, but as the share price reverted to the levels (with no new information from the company), thesis is still valid.

Business Model

Lendingtree generates potential borrower leads through its website and then sells these to interested lenders. Mortgage leads comprise 83% of revenue. Lendingtree can be thought of as a reversed marketplace, were the lenders are shopping for the borrowers rather than vice versa. Consumer provides his personal and required loan info on Lendingtree and then this information is sold to multiple highest bidding lenders (main revenue of the company). Lenders do not provide quotes instantaneously or in an organized fashion, rather consumers are bombarded by phone calls and emails with various proposals or requests for further information. Thus unlike most insurance price comparison websites, this market place hardly results in prompt, competitive and transparent pricing for the borrowers. Half of the largest US lenders are not even in the Lendingtree’s network, clearly indicating that this sales channel is not considered to be important. This quote from one of the unhappy borrowers nicely summarizes the business model:

“I thought that I was the customer, loans were the product, and Lendingtree would help me find a loan. It turns out that I am the product and banks are the customer”.

Lendingtree should not be confused with innovative and disruptive concepts such as Lending Club or Proster – it’s a simple website with questionnaires and it is 18 year old business model.
   

Decreasing growth and margins. No evidence of operating leverage.

Lendingtree sported very large growth rates in 2012/2013 (some quarters reaching +100% YoY), however this was caused by divesture of captive lender (mid 2012) after which the leads that had been kept internally previously started being sold at market prices. Thus the main growth was not driven by increased popularity of its services or capturing of additional leads but rather by selling the existing inventory externally. Subsequently growth slowed down and amounted to only 10% YoY in the latest quarter (negative sequentially), whereas mortgage segment revenue even declined by -3% YoY.

All of this slowdown in growth is happening in the backdrop of the environment in which Lendingtree is supposed to show the best performance. During the recent year mortgage origination activity in US has slowed down significantly with c. 30-40% drop YoY. This low demand for mortgages is expected to push lenders towards active search for new customers through alternative means, i.e. third party leads such as Lendingtree. Management also confirms this expected inverse industry relation in the annual/quarterly reports.

Margins are also under pressure as the company spends ever increasing % of revenue to acquire visitors to its site (organic traffic is minimal). Gross margin (revenue less COGS and variable marketing expenses) decreased from 45% to 34% in the two years since divesture of the captive lender. Overheads as % of revenue dropped only from 41% to 34% even though the company tripled in size in the same period. There is no pricing power either – in the latest quarter the fee per matched consumer dropped 29% and has been on 10%-20% decline in the last year.

Thus even though these type of online platform businesses are expected to have strong operating leverage, there is no evidence of it within Lendingtree, as increasing visitor acquisition costs eat up any benefits of scale.

Competitors have structural cost advantages

Lendingtree is competing with the likes of Bankrate, Zillow Mortgage and many other unlisted mortgage lead generators (Trulia also has mortgage offerings, but currently uses Bankrate’s platform). Both Zillow and Bankrate have strong organic traffic due to other parts of the business (real estate search for Zillow, and financial content and other financial services for Bankrate) and can direct some of the consumers towards mortgage calculators at almost zero marginal costs. Lendingtree on the other hand does not have that advantage and has to spend 65-70% of revenue just to attract customers to its mortgage lead questionnaires.

Compared to Lendingtree, Zillow Mortgage is still a smaller player in terms of revenue ($7m vs $41m in Q3), but is growing the mortgage business faster (25% vs -3% YoY) and already far surpasses Lendingtree in terms of ‘mortgage requests’ (7m vs c. 0.35m quarterly). Bankrate does not reveal segment info for mortgages.

Given structural cost advantages I would expect both Bankrate and Zillow to continue to expand their lending lead generation businesses, rendering Lendingtree irrelevant in the eyes of consumer (even now Ledingtree organic traffic and popularity are minimal) and too expensive for the lenders.

Revenues are driven by unsustainable business practices

BBB complains reveal aggressive and unsustainable business practices (these are actually quite fun to read):
- Consumers are fooled about the service Lendingtree actually provides – selling leads vs instant quotes on mortgages.

- The same lead seems to be sold to far higher number of lenders than Lendingtree claims;
- Leads are sold even before consumer finishes and confirms the application;
- Consumers are fooled to provide their contact details and lenders start calling even if consumer does not submit the application;
- Erasing account and personal data from Lendingtree is made very difficult.

These kind of aggressive lead selling practices might be profitable in the short term as the higher the number of lenders that the single lead is sold to, the higher the revenues. But eventually the quality (rate of conversion for a specific lender) of leads decreases, creating pressure on the pricing. This is exactly what the recent financial performance seems to indicate. Moreover, these practices alienate the company with potential borrowers and reputation conscious lenders who do not want to be associated with predatory and intrusive (not stop emails and phone calls after giving contact details) lending practices. Such behavior and additional revenue generation is unlikely to be sustained in the longer term.


Valuation

Lendingtree is currently selling at c. 3.5x revenue and is operating around breakeven in terms of operating cashflow and net profit. On adjusted EBITDA basis (metric used by management), the company sells at EV/adj. EBITDA=21. However, the adjusted EBITDA metric excludes $6.2m in non-cash compensation (another excluded item is $12.5m in seemingly non-ending litigation costs). Correcting for non-cash compensation results in EV/EBITDA=30. This is far too optimistic valuation in light of all the headwinds mentioned above. For comparison Bankrate (3 times larger, more diversified and arguably more efficient lead generator) trades at EV/adj. EBITDA=11 after correcting for non-cash compensation. Using the same EBITDA multiple, the fair value of Lendingtree is around $23/share.

A DCF model with a very optimistic growth and margin expectations leads a similar result of $26/share. Assumptions used:
- 15% growth for next 5 years vs 10% currently; 
- 10% net margin in year 5, further efficiencies unlikely as majority of cost base is variable marketing expenses which only seem to be increasing with the expansion of the company;
- Discount rate of 10% and terminal value at PE=15.

These valuation estimates are 45%-50% below the current share price of $49.


Why the opportunity exists?

It all boils down to management's ability to create a hype with regards to the potential of the business and to position Lendingtree as a very popular and revolutionizing technology company in the eyes of investors. There are a number of misconceptions (myths) that should be viewed with high skepticism: 

Myth No.1: Lendingtree competes in the large US lending market. The reality is that Lendingtree operates and competes in lead generation industry. Its total addressable market is limited to the lenders’ marketing spend on customer leads, which is significantly below the trillions of dollars that management keeps on discussing in the presentations.

Myth No.2: Lendingtree is a breakthrough technology company. Actually the business model is 18 years old and Lendingtree is a simple website with questionnaires. The same technology is used for decades and the company hardly invests anything in innovation relative to competitors (R&D spent is 7% vs 20%-30% for competitors).

Myth No.3: Lendingtree is a household name and brand awareness is 5 times larger than that of Bankrate or Zillow (as per chart on Lendingtree.com presentations). Review of website popularity and traffic composition (Alexa, Google Trends, SEMRush, organic search results and etc.) clearly shows that Lendingtree’s brand awareness is at least 10x smaller than Bankrate's and 50x smaller Zillow's. The company hardly gets any organic traffic (it is does not appear in search results with seemingly important keywords such as ‘Mortgage’ or ‘Mortgage calculator’) and is fully relying on paid advertising to attract consumers to its site. Lendingtree’s popularity seems to have declined materially during the last 5 years.

Myth No.4: Non-mortgage product growth will finally bring the company to profitability. Non-mortgage products are likely to cause even higher losses than the mortgage ones as the revenue per lead is materially lower (price of lead is proportional to the loan amount) but admin expenses are quite similar. Thus any growth is value destructive. Management hides segment profitability from investors even when prompted directly in conference calls.

Myth No.5: The new ‘My Lending Tree’ is a game changing platform. ‘My Lending Tree’ is nothing more than a user interface that stores personal details and spams potential borrowers with offers. Consumers seem to hate the idea of it (BBB complains). Opt in is involuntary and opt out is hard. Consumers are unlikely to start using it as subsequent communication with lenders is happening offline.


Risks and other considerations

- Takeout transaction at even higher multiples is always a possibility. However, there is hardly any franchise value as Lendingtree business is only sustained by continuous acquisition of traffic and is very easy copy-able technologically (Lendingtree lost its patent/intellectual property protection in a lawsuit it has filed against Zillow, more info here). Relationships with lenders might take some time to work-out, but still starting a similar business from scratch would be way cheaper. Therefore I view takeout transaction at current prices as very unlikely.
- CEO still holds 24% of the company which is clearly not an ideal situation for the short sellers. But CEO was also an active seller at price above $30/share.
- Liberty Interactive also has 24% stake, but it is a legacy holding from 2008 spin-off, and Liberty has restrictions on disposing it. Liberty bought Lendingtree in 2003 at $700m and then spun it off for only $70m five years later, thus clearly not the best investment for the company.

Catalyst

- Lack of top line growth during 2015 (in line with recent trends) will eventually bring the share price to more reasonbale valuation levels;

- Q4 results are likely to be bad in light of seasonal fluctuations and further mortgage segment deterioration. I would expect non-mortgage revenue growth to be overshadowed be the decline in mortgage revenues, which might completely eliminate any top line growth;

- Continued growth of compitotors, especially Zillow Mortgage.