Excerpted from an email which Whitney Tilson sent to investors this afternoon.
An interesting bull-bear debate on Wayfair going on. Below is the upgrade from yesterday and Xuhua Zhou’s response today.
I’m strongly with the bears (and am short it).
PS–Some of the comments on the message board below Xuhua’s article claim the following:
it’s almost impossible to replicate someone’s search results. That’s why it’s a bad metric to base your opinion on. Search results vary by location and IP Address, past google search click history, etc.
However, I just tried more than a dozen search terms in which the Google rankings Xuhua found for Wayfair were much lower than what the BofA/ML analysts found, and every case was consistent with what Xuhua reported.
Sell Wayfair When Negligent Wall Street Analysts Are Too Lazy To Google
Hedge fund analyst
Disclosure: I am/we are short W. (More…)
Sell-side research analysts misstate Google search results.
Low repeat rate will drive continued advertising spending and destroy EBITDA margins.
Highly frequent insider selling adds pressure.
Inventory-light model raises questions over product quality assurance.
Just when I thought the bear case for Wayfair has become crystal clear that no sell-side research analyst would even entertain defending the stock, BAML analysts Paul Bieber and Justin Post proved me wrong. In a note published on Monday, the pair of research analysts upgraded the stock from neutral to buy, despite slashing their price target by 17%. Furthermore, a substantial portion of their bull case rests on Wayfair’s high placement in organic search results on Google for home goods. But did they properly Google the keywords they claim to have Googled?
Of the 23 items ranked #1 by BAML, only 7 items are actually #1 when searched on Google, with the rest all ranked lower. “Table linen”, for example, is ranked #6 even though BAML claimed it is #1. Of the 11 items ranked #2 by BAML, only 1 item was actually ranked #2 on Google, with the rest all ranked lower. “Dinner plates”, for example, is ranked #5 as opposed to #2 claimed in BAML’s research.
For a complete reference of Google search ranking of key phrases used in BAML research, see the chart below.
Oftentimes, sell-side analysts have different points of view that are subject to debate. As stretched as the merits of their argument might be, the points are, at a minimum, based on facts. Organic Google search results are not subject to debate. Before recommending your clients to get into a money-losing e-commerce business, at a minimum, make sure you have conducted correct Google searches.
To analyze a business model such as Wayfair’s, access to proper cohort data will help investors understand whether the company’s business model and brand are indeed gaining grounds with customers. Analysts have gone to great length to highlight repeat purchase behaviors at Wayfair. No bull thesis stands without assuring investors that the company will be able to retain customers after it turns off the tap on advertising spending. Unfortunately, with the exception of its IPO prospectus, Wayfair seems to have been trying hard not to disclose a proper set of cohort data that will show the percentage of repeat customers over time.
Instead, Wayfair management and sell-side analysts have been actively promoting two sets of seemingly similar data points, hoping investors might simply equate those to proper cohort data.
Almost always, when management actively hides readily available data and direct investors’ attention to data that look alike, there is something under the hood that is worth digging. In the case of Wayfair, it is important to explain why seemingly rapidly improving data points offered by management do not translate to sufficient improvement in customer retention.
Company management likes to point to a PowerPoint slide in its investor presentation deck showing rapid improvement of its gross revenue per customer over time.
On the surface, it looks like an impressive achievement showing customers acquired in more recent period spending more over time than customers acquired in older periods. Management further highlighted that in the 6th month post initial purchase, average 2014 customers spent over twice as much as ones in 2011.
First of all, it is a fairly confusing set of data. To calculate the average spending during the 6th month of initial purchase of customers acquired in a specific year, Wayfair has to track July spending pattern for customers acquired in January, August spending for ones acquired in February, and so on and so forth until it aggregates December spending for customers acquired in June. If you assume a constant order size and a constant percentage of 6th-month repeat customers, the disclosure really provides a snapshot of repeat customer rate in the 6th month post initial purchase relative to prior years.
Going back to the repeat customer definition in Wayfair’s prospectus, the company defines a repeat customer as someone who purchased from the same site at least twice during the quarter and the subsequent 270 days. In another word, a repeat customer is the cumulative probability of someone who bought again in either the 1st month, the 2nd month or 3rd month etc., up until the 9th month after his initial purchase. For the more mathematical minded, think of the repeat customer percentage as a bell-shaped cumulative probability function, with each further month representing a smaller percentage of the repeat customer percentage. This is also easy to understand, because over the span of 9 months, a repeat customer is much more likely to purchase in the first month of their initial purchase rather than wait till the 9th month to make a repeat purchase.
Wayfair disclosed in the third quarter of 2011 and 2013, repeat rates were 15.9% and 24.2% respectively. Due to the way repeat customer is defined, 1st-month and 2nd-month gross revenue per customer over time