Although the accuracy of the data captured by muddy waters is questionable, the shell is not invulnerable
Text | Zhang Guangyu Chen Jiayao
Edit | Wang Bo
In the past three days, Muddy Water, a short-selling agency known as the "Chinese Stock Sniper", has been in a fierce confrontation with Shell House (NYSE: BEKE), China's largest real estate trading platform.
On the evening of December 16, 2021, Muddy Water released a short report saying that by grabbing the transaction data of the shell platform and visiting offline stores, they found that shell exaggerated 126% of GTV (total transaction volume) in the second and third quarters of 2021, falsely reported commission income by 77% to 96%, and inflated the number of stores by at least 23%. Muddy Water concludes that shells are committing systematic fraud.
On the evening of the 17th, Shell issued a statement in response, saying that there were a large number of factual errors in muddy water reports, and their method of grasping transaction volume data and measuring the number of stores and brokers was wrong, resulting in untruths and omissions in the data, "showing that muddy water lacks basic understanding of China's real estate trading industry."
In the interview with caijing reporters, most of the people in the brokerage industry agree with the shell statement. The volume of shells is contributed by multiple channels, which is limited by incomplete information sources, and muddy waters cannot grasp the channel data outside the shell system, and the internal channel data is not allowed. And their logical thinking of inferring GTV and revenue through the number of stores and brokers may not be valid.
But the report of short selling in muddy waters is not worthless. It pointed out the long-standing "ghost shop" and "clone shop" problems of shells, which led to statistical errors in shell stores.
On the 17th, on the day of Shell's response statement, Shell's stock price rose 5.7% to close at $19.36 per share.
Muddy waters shorts and shell counterattacks, whose data is more reasonable, at present, the capital market has chosen the latter.
Where are the statistical loopholes in muddy waters?
In the appendix to the short report, Muddy Waters describes how to scrape trading data.
They enter from the shell platform "find a broker" port, see the broker's store, and then count the transaction records of each store, thus obtaining the transaction volume. Because the shell platform only shows the transaction records of the past 76 days, the statistical range of muddy waters for the second and third quarters is also 76 days, which is then proportionally converted into a complete quarter.
The statistics of the transaction price are divided into new houses and second-hand houses. Muddy Water said that because according to the data of the China Real Estate Association, the unit price of new houses has remained almost unchanged in the past two years, so they used the average transaction price disclosed in Shell's 2020 annual report to estimate the transaction price of new houses.
As for second-hand houses, 62% of cities can obtain the transaction price on the shell platform, and the muddy water is directly grabbed. In the remaining cities, Muddy Waters selects cities with similar characteristics as a reference according to the data of the China Real Estate Association, and estimates the transaction price according to the price ratio relationship between the two cities. It is worth noting that the transaction prices of second-hand houses in Beijing and Shenzhen are estimated.
Multiply the volume by the average transaction price to get GTV.
In response to the statistical method of muddy waters, Shell responded that there are four channels for the company's new housing transactions: chain stores, stores connected by shell platforms, new housing sales teams, and other sales channels (outer channels). Muddy waters only calculate the first two ways.
Within the Shell Group, the above channels are more often divided into "inner channels" (chain homes and shell platforms connecting stores), "new house direct sales" and "outer channels".
A former shell operator told Caijing that Shell set up a special new home direct sales team in the second half of 2020. These brokers are divided into areas according to the business districts in the city, and they are stationed at the sales offices of each property on a daily basis, and do not correspond to a certain store.
In the more mature business district, the team will rent an office in the office building for use in meetings and formalities. In the suburbs and other new developer circles, the conditions are more difficult, there is no office, and all the work is done in the sales hall.
And according to the shell statement, these new house agents can not be found in the "find a broker" function on the app homepage, but must be found in the sub-page of the "new house" channel.
Therefore, muddy waters follow the broker's method of finding stores, and the turnover of the new home direct sales team cannot be counted. Of course, in addition to the direct sales of new houses, the Inner Canal store also sells new houses, and Muddy Water only counts this part.
Let's look at the outer canal. The so-called "outer channel" refers to an external intermediary company that has a cooperative relationship with shells, but is not subordinate to the shell system. These companies can use either shell ports for orders or ports from other platforms.
The above-mentioned former operator explained that Shell will not display the outer channel on the "Shell Find a House" page of the guest platform, and they generally have to find their own way to get customers. The outer canals are all completed through a set of systems called "link" inside the shell to complete the signing report, carry forward commissions and other operational links, and the relevant processes are not displayed to the outside world.
The head of an outer channel agency that cooperates with Shell told reporters that for Shell, the essence of this business is To B, not To C, and "the overall data of the outer channel is only known by Shell."
The "Finance" reporter learned that in addition to the omission of "new house direct sales" and "outer channels", muddy waters also have doubts about the statistical accuracy of the inner canals.
Shell has two statistical paths online and offline for its own data, and the two sets of data are not always consistent. The customer orders traded in the inner canal are not 100% completely recorded online. There are subjective and objective reasons for this.
Subjectively, in addition to a momentary laziness, some stores and brokers will adjust the entry rhythm to ensure the completion of performance appraisals. For example, when the indicators have been completed, first do not rush to put the order on the line, leave some "inventory", and wait until the performance is worried to find a supplement.
There are also various objective reasons that cause orders that have been traded to be unable to be entered. For example, in the new housing business, if the developer and the shell are in the commission payment, or whether the order is sold by the shell or sold by the developer itself, then according to the shell rules, the order cannot be recorded as "underwriting" online, and must wait for the dispute to be resolved.
In addition, the data held by the financial and operational ends may also be inconsistent. For example, the house that has been sold, the buyer is refunded; or in cooperation with the developer, because of the good performance, the commission jump point rises in the middle. For these changes, the statistics of the two ports of financial operations may be one after the other, and there is a time difference.
Shell itself can periodically calibrate each other with two sets of data, online and offline. However, Muddy Water can only capture two online data from each 76-day operation end, and the cycle is too short to ensure accuracy.
Muddy Water's price estimate of second-hand houses used the data disclosed by the China Housing Association, especially the price of Beijing, the base camp of the chain family, was estimated by this method. This may be quite different from the real situation. A chain of home brokers told the "Finance" reporter that since the introduction of the second-hand housing guidance price policy, the listings displayed on the platform have been reserved. Some high-priced listings are hung on the intranet, and buyers can get information about these listings through brokers to complete the transaction.
Fake stores, shells are not invulnerable
Muddy Waters reports that their data collection shows that the total number of stores by the end of the second quarter of 2021 has been exaggerated by 23%. When they visited offline, they found a large number of "ghost stores" and "clone stores", so the actual degree of exaggeration may be higher.
The so-called "ghost store" means that the store is registered on the shell platform, but it has actually closed its doors. "Clone store" means that several stores register the same address, but only one store is actually registered. Muddy Water said that they visited 51 registered chain stores in Langfang and found that 19 were "ghost stores", so the number of stores in the area was exaggerated by 59%.
Muddy Waters also questioned the actual number of brokers for Shell. The report mentioned that Shell claimed to have 21,000 brokers in Shanghai, but when they checked the data of the State Administration of Market Regulation, they found that shell's brokerage subsidiaries in Shanghai had only 9,998 employees.
In response to the above questions, Shell responded that the muddy water information source was incomplete, that is, as mentioned above, there was no statistics on the new housing agent. But there was no response to what the "ghost shop" and "clone shop" were all about.
"Finance" reporter learned in the interview that shell's "ghost shop" and "clone shop" do exist. Shell cleans up such stores from time to time, but it is still difficult to eradicate. Changeability is the characteristic of the intermediary industry, not only the turnover rate of brokers, the turnover rate is high, and it is common for stores to close within a few months of opening. These closed stores, especially the non-chain home brands connected by the platform, generally do not take the initiative to cancel on the shell platform.
Brokers registered online have the same problem. A former shell operator said that his area had done a screening this year and found that 5% of the personnel were not on the job, that is, in the roster status of "on the job", but did not actually punch in to work.
He explained that some employee positions have been transferred, but they have not been modified in the system in time, so they have not been counted. But there are also some people who have just put a name on the shell and have actually left the company and even the industry. As for the situation in Shanghai described by Muddy Water, compared with the statistics of the State Administration for Industry and Commerce, there are indeed more actual brokers of the company, because some of the companies signed by brokers may not be shell subsidiaries, but various external companies.
The operator was working for Shell in a southern city, and party A, whose labor contract was signed, was signed as a human resources company in Beijing.
From this point of view, how many stores and how many brokers there are in Shell are also two sets of online and offline data. Online data is likely to be unreliable. If the shell itself wants to figure it out, it also depends on the effort offline.
One of the core logics of the Muddy Water short report is that there is a positive correlation between the number of stores and brokers of Shell, and its GTV and revenue. If the former has moisture, then the shell's GTV and revenue are bound to have moisture.
Although the number of shell stores and brokers is indeed doubtful, this logic of muddy waters is not agreed by the brokerage industry interviewees.
Because of the direct sales of new houses and external channels, the proportion of shell GTVs has become larger and larger. Former operators said that in a southern provincial capital city where he is located, the proportion of GTVs contributed by new home direct sales and outer channels in 2020 is about 50%.
However, the direct sales team does not increase the number of stores, and the stores and brokers of The Outer Channel are not reflected in the statistics of the Shell housing platform.
Under the business model of declining the proportion of inner canals, Muddy Water only relies on statistics on the number of inner canal stores and people, and it is concluded that shell's GTVs and revenues are untenable.
The author is a reporter for Caijing
There's so much more to see