Muddy Water Research, a big short market known for its air share concept, has released an 89 page short report on Luckin Coffee.

The report author sent 92 full-time and 1400 part-time investigators, collected more than 25000 small tickets, carried out 10,000 hours of store videos, and collected a large number of internal WeChat chat history. It is believed that the number of "items per store per day was inflated by at least 69% in 2019 3Q and 88% in 2019 4Q". At the same time, it also pointed out that Luckin overstated its 2019 3Q advertising expenses by over 150%, especially its spending on focus media.

Luckin Coffee Inc. sank as much as 20% when the news was disclosed.

At the same time, Muddy Waters Capital tweeted that it received an unattributed 89-page report claiming that Luckin Coffee was a hoax :" the number of items per store per day inflated by 69% in 2019 3Q and 88% in 2019 4Q, supported by 11,260 hours of store traffic video. We think the work is credible."

At present, Luckin Coffee has not responded to the short report.

                                                                        The short report collected 25,843 tickets from 10,000 users

The following is the core content of the short report extracted from Deep Insights (ID: deep_insights) translation:

Executive Summary

When Luckin Coffee went public in May 2019, it was basically a failed business to instill a coffee drinking culture into Chinese users through high discounts and free gifts.

After its $645 million IPO, the company started fabricating financial and operating data in the third quarter of 2019, which has become a fraud. A series of results released by the company showed a dramatic turning point in its business, with its share price rising by more than 160% in more than two months. Not surprisingly, it successfully raised $1.1 billion (including a secondary placement) in January 2020.

Luckin Coffee knows exactly what investors are looking for, how to position itself as a growth stock with a wonderful story, and what key indicators to manipulate to maximize investor confidence.

This report is composed of two parts: fraud and basically collapsed business. We show how Luckin Coffee forges data and why there are inherent defects in its business model.

Part One: The Fraud

Smoking Gun Evidence 1

The number of items per store per day inflated by 69% in 2019 3Q and 88% in 2019 4Q, supported by 11,260 hours of store traffic video.

We have tracked 981 offline stores since Q4 2019, with only 263 items sold per store per day. (Luckin Coffee's document shows that in 2019, each Q3 store sold 444 commodities per day, and it is estimated that there will be 483-506 products in Q4 )

We sent 92 full-time and 1418 part-time personnel to monitor on site, and successfully recorded 981 store day traffic, covering 620 stores with 100% business hours. The store selection is based on the distribution of city and location type, which is the same as that of 4507 direct stores of Luckin Coffee opening at the end of 2019. Luckin Coffee's 4507 stores are located in 53 cities, and we cover 38 cities, 96% of which are located in these cities. By analyzing the detailed address of Luckin store to determine the location type of the store: we divide the stores into office, shopping mall, school, residence, transportation, hotel, etc.

We counted the traffic of each store and recorded the video from opening to closing, with an average of 11.5 hours per day. When we check the traffic and recorded video again, if the video monitoring loses more than 10 minutes of clips, we will discard the data for a whole day. The success rate is only 54%, so all successful data are 100% complete.
... (discussion on data rigor)

This is evidence that on November 23, 2019, the store manager was informed to pay attention to the sharp increase in the number of pickup.

However, here's a smart part: company management may think that more and more investors and data companies are tracking their order numbers as part of the due diligence process, so "jump orders" are a simple way to mislead investors. In order to understand the scale of online order expansion, we randomly selected 151 offline to track their online orders. We placed an order at the beginning and end of store hours to get the number of online orders for the day. We found that the number of online orders in the same store on the same day expanded from 34 to 232, with an average of 106 orders per day or 72% of offline orders.

Smoking Gun Evidence 2

Luckin's "Items per order" has declined from 1.38 in 2019 2Q to 1.14 in 2019 4Q.

Starting in the fourth quarter of 2019, we collected 25,843 receipts from 10,119 customers from 2,213 stores in 45 cities. 25,843 receipts showed 1.08 pickup and 1.75 delivery orders per order, or a mix of 1.14 (99% confidence level). It marks a continuous decline in the number of items per order, from 1.74 in the first quarter of 2018 to 1.14 in the first quarter of 2019.

This trend can be attributed to the decline in the contribution of delivery notes, as people naturally tend to buy more items to meet the requirements of free delivery. Luckin said:" By visiting our stores, we found that most door-to-door customers only buy one kind of instant beverage, because in most cases, coupons can only be used for one kind of goods in the order." According to the company's introduction and management communication, the proportion of orders delivered actually decreased from 61.7% in the first quarter of 2018 to 12.8% in the third quarter of 2019, and further decreased to about 10% in early January.

Smoking Gun Evidence 3

We gathered 25,843 customer receipts and found that Luckin inflated its net selling price per item by at least RMB 1.23 or 12.3% to artificially sustain the business model. In the real case, the store level loss is high at 24.7%-28%. Excluding free products, actual selling price was 46% of listed price, instead of 55% claimed by management.

Luckin reported that the net selling price of each product in the third quarter of 2019 was 11.2 yuan. At the revenue conference call on November 13, 2019, Reinout Schakel, Luckin's CFO and COO, guided higher prices for the fourth quarter of 2019. However, our 25,843 receipts show a net selling price of only RMB 9.97, that is to say, compared with the reported situation, the inflation rate is 12.3% (99% confidence level and 1% statistical error mean that the price is between RMB 9.87-10.07, with 1% error).

Excluding free products, the selling prices of instant drinks and other products are 10.94 yuan and 9.16 yuan, respectively. Compared with the reported situation, the inflation rates are 12.3% and 32% respectively. Excluding free products, the actual selling price is 46% of the listing price, rather than 55% claimed by management.

At the revenue conference call, Reinout Schakel sidestepped questions about increasing promotions. However, our receipts show that they even provide free beverage coupons to existing users from the fourth quarter of 2019, while previously only new users and users who invited them were provided with free beverage coupons. It is speculated that the percentage of free projects submitted by each company is declining.

Ruinout schakel also said they continued to increase the number of people who were already paying the price they wanted to pay. Our receipt shows that on the contrary, even in mature markets, with more mature customers, the effective price stagnates at 10 yuan, excluding free products. There is no positive correlation between the net selling price of each commodity and the operating month.

At a recent Citibank meeting in January, he said that more than 63% of customers pay 15-16 yuan per cup of coffee. In the third quarter of 2019, the company reported that 63% of its products sold for more than 50% of the retail price. However, these are too good to be true, which are contradict our receipt findings.

Our receipt shows that only 28.7% of the goods have been sold at a price over 50% of the list price. In fact, most of the goods are priced between 28% and 38% of the price. Luckin's core customers are still very sensitive for the price. Only 39.2% of customers pay more than 12 yuan, and 18.9% of customers pay more than 15 yuan per cup of coffee.

Why is ASP important? If investors remember Luckin's sensitivity analysis of store profitability in his speech, they will find that ASP is the key factor of store profitability. They pointed out that in the case of 400 products per day in each store, the price of each product is 16 yuan, and the store level profit margin can be as high as 28.4%. In the analysis of the management, the earnings per share close to the actual situation is lower than the lower limit of 12 yuan, which is ignored somehow. It represents a more difficult earnings prospect.

In the actual situation, 263 pieces per day for each store, the net selling price is 9.97 yuan, the store level loss is 28.0%, according to the management report. Note that all figures are provided by management. To put it another way, we give the company some credit, achieve economies of scale by sending free coffee, and reduce costs in the figures reported in the second quarter of 2019. The loss at the store level is still as high as 24.7%. At the current price level, they can only make profits at the store level by selling 800 products per store every day, otherwise they must raise the effective price to a minimum of 13 yuan. That's why they need to invent ASP numbers to maintain their business model.

Smoking Gun Evidence 4

Third party media tracking showed that Luckin overstated its 2019 3Q advertising expenses by over 150%, especially its spending on focus media.

Luckin disclosed the quarterly advertising expenses before March 31, 2019 in the prospectus. After IPO, its advertising expenses can be calculated by the sub item of new customer acquisition cost in its quarterly profit report.

In the second quarter of 2019 financial report conference call, the company first disclosed that focus media accounted for 140 million yuan of 240 million yuan + total advertising expenditure in the second quarter of 2019 (they only explained 154.5 million yuan, accounting for 64% of the total advertising expenditure of 242 million yuan).

According to the data tracked by CTR market research, Luckin reported more than 150% of the advertising expenditure in the third quarter of 2019: in the third quarter of 2019, CTR indicated that focus media spent 46 million yuan, accounting for only 12% of Luckin's advertising expenditure, far less than the previous quarters. Assuming that Luckin's advertising expenditure of non focus media in the third quarter of 2019 is equivalent to this, Luckin overstates its advertising expenditure by 158%.

CTR market research tracks the actual advertising broadcast of different brands in various media channels, including the focus media channels of all three media: LCD network (office building elevator), poster / digital framework network (residential elevator), and cinema network - accounting for 82%, 17% and 1% of the total revenue of focus media in the first half of 2019, respectively based on the interim report of focus media 2019.

Below is CTR's monthly tracking result of Luckin's advertising expenditure in focus media channel. Luckin's spending fell to its lowest level between September and November 2019, but rebounded in December 2019.

The dollar amount in the original CTR data is the media list price, which may be much higher than the actual advertising expenditure.

In order to calculate the conversion rate between the list price and the advertising expenditure, we calculated the conversion rate of the total revenue reported by CTR tracking focus media. According to the data from the first quarter to the third quarter of 2019, the actual revenue of focus media is about 8% of the CTR tracking media list price.

According to the accounting policies listed in the financial report of focus media, the advertising revenue of focus media is recognized at the time of "advertising broadcast", which is the same as the advertising broadcast time tracked by CTR. When should Luckin book the advertising expenses.

CTR also publishes monthly, quarterly and annual largest advertiser reports on its website based on tracking results. For example, in May 2019 (link), CTR pointed out that Luckin is the largest advertiser of all the media channels it tracks (including the three channels most used by traditional outdoor, TV, radio and focus media). It is worth noting that 83% of Luckin's tracking advertising budget in the month is for LCD display network, 12% is for poster / digital frame network and 5% is for cinema network.

However, Luckin's ranking in LCD display and poster / digital box advertising fell rapidly in June and July 2019, even falling out of the top 10 (links) since August 2019.

So where did the money go?

Similar clues can be found in inflated store profits and advertising costs.

Luckin claims to have achieved "store level profitability" in the third quarter of 2019. Combined with solid evidence 1 to 3, it actually hides the loss at the store level below the store level, rather than really exceeding the break even point at the store level.

The real case of Luckin store level results is the sales volume of 263 items per store every day, ASP is 9.97 yuan. Compared with real cases and reported cases, Luckin group exaggerated the operating profit of its stores by 397 million yuan in the third quarter of 2019. Coincidentally, the difference between the advertising expenditure reported by Luckin and the actual expenditure of focus media tracked by CCTV is 336 million yuan, just like the exaggerated operating profit of the store. In addition, from the third quarter of 2019 onwards, these two misstatements became apparent. Luckin is likely to reuse its inflated advertising costs for fraudulent revenue and store profits.

Smoking Gun Evidence 5

Luckin's revenue contribution from "other products "was only about 6% in 2019 3Q, representing a nearly 400% inflation, as shown by 25,843 customer receipts and its reported VAT numbers.

Luckin's ambition is not to start a coffee company. Its mission is "everyone's daily life, starting with coffee", which makes "other products", i.e. non-instant drinks, such as light meals, juice, nuts, mugs, etc., an important product. Its revenue contribution is reported to have increased from 7% in the second quarter of 2018 to 23% in the third quarter of 2019, and the project contribution correspondingly increased from 6% to 22%.

However, in the 981 working days we tracked, only 2% of pickup orders found non-instant products. 25,843 receipts further show that 4.9% and 17.5% of the delivery orders are "other products", accounting for 6.2%, that is, the expansion is nearly 400%. Once again, people naturally tend to buy more "other products" to meet the requirements of free delivery. However, if the order rate in the second quarter of 2018 dropped significantly from 62% to nearly 10% now, why did the "revenue percentage from other products" rise from 7% to 23% in the same period?

The VAT rate in Luckin's latest form F-1 also supports our findings: according to the State Administration of Taxation of the People's Republic of China, the VAT rate for sales of goods and services is different. For the provision of services, such as the sale of existing products or delivery, the VAT rate is 6%. In the case of sales, such as packaged food and beverages, i.e. "other products", in Luckin's case, the VAT rate since April 2019 is 13% (or 16% before that). This is further confirmed by the VAT invoice we received after Luckin's purchase (see sample below). According to Luckin's income classification, we can calculate a mixed VAT rate and compare it with the company's report.

Weighted average net income contribution% by product category, we find that the calculated VAT rate is fully consistent with the pre IPO reporting for Q4 2018, full year 2018 and Q1 2019 (see figure below).

However, in the second to third quarter of 2019, the gap suddenly widened, with the reported VAT rate of 6.5% compared to the actual calculation of 7.6%. From another perspective, in order to be consistent with the reported value-added tax of 6.5%, the revenue contribution of other products will actually be 7%, which is very close to 6.2% of 25,843 invoices, while the report of the company is 22% - 23%.

In this case, the actual revenue contribution of "other products" in the third quarter of 2019 is 6% - 7%, or Luckin has tax evasion.

In order to confirm that the VAT rate of other products is 13%, we purchased some products in Luckin and asked for VAT records. It clearly shows that 13% VAT is for nuts, muffins, juice, etc., and 6% for fresh brewed drinks and delivery. Anyone who wants this information can ask for VAT records through Luckin application after purchase.

Red Flag 1: Luckin's management has cashed out on 49% of their stock holdings (or 24% of total shares outstanding) through stock pledges, exposing investors to the risk of margin call induced price plunges.

Luckin's management stressed that they had never sold any of the company's shares; however, they had cashed out through stock pledge financing. The number of shares pledged is almost half of their total, worth $2.5 billion at current prices.

Red Flag 2: CAR (699 HK) déjà vu: Charles Zhengyao Lu and the same group of closely-connected private equity investors walked away with USD 1.6 billion from CAR (699 HK) while minority shareholders took heavy losses.

Red Flag 3: Through acquisition of Borgward, Luckin's Chairman Charles Zhengyao Lu transferred RMB 137 million from UCAR (838006 CH) to his related party, Baiyin Wang. UCAR, Borgward, and Baiyin Wang are on the hook to pay BAIC-Foton Motors RMB 5.95 billion over the next 12 months. Now Baiyin Wang owns a recently founded coffee machine vendor located next door to Luckin's Headquarter.

Red Flag 4: Luckin recently raised USD 865 million through a follow-on offering and a convertible bond offering to develop its "unmanned retail" strategy, which is more likely a convenient way for management to siphon large amount of cash from the company.

Red Flag 5: Luckin's independent board member, Sean Shao, is/was on the board of some very questionable Chinese companies listed in the US that have incurred significant losses on their public investors.

According to Luckin's prospectus, Sean Shao, an independent director, has served as a director of a number of Chinese companies listed in the US after 10 years at Deloitte. We studied these companies in detail and found that 18 companies, Sean Shao had been accused of fraud (CHME, ADI, GRO and Yong) on the board of directors 4 and 5 were reverse acquisitions - the infamous generation of a large number of Chinese fraud companies as early as 2011-2012.

Red Flag 6: Luckin co-founder & Chief Marketing Officer, Yang Fei, was once sentenced to 18 months' imprisonment for crime of illegal business operations when he was the co-founder and general manager of Beijing Koubei Interactive Marketing & Planning Co.,Ltd. ("iWOM"). Afterwards, iWOM became a related party with Beijing QWOM Technology Co., Ltd. ("QWOM"), which is now an affiliate of CAR and is doing related party transactions with Luckin

Part Two: The Fundamentally Broken Business Before 3rd Quarter, 2019

Business Model Flaw 1: Luckin's proposition to target core functional coffee demand is wrong: China's caffeine intake level of 86mg/day per capita is comparable to other Asian countries already, with 95% of the intake from tea. The market of core functional coffee product in China is small and moderately growing in China.

Business Model Flaw 2: Luckin's customers are highly price sensitive and retention is driven by generous price promotion; Luckin's attempt to decrease discount level (i.e. raise effective price) and increase same store sales at the same time is mission impossible

Business Model Flaw 3: Flawed unit economics that has no chance to see profit: Luckin's broken business model is bound to collapse

Business Model Flaw 4: Luckin's dream "to be part of everyone's everyday life, starting with coffee" is unlikely to come true, as it lacks core competence in non-coffee products as well. Its "platform" is full of opportunist customers without brand loyalty. Its labor-light store model is only suitable for making "Generation 1.0" tea drinks that have been in the market for more than a decade, while leading fresh tea players have pioneered "Generation 3.0" products five years ago.

Business Model Flaw 5: The franchise business of Luckin Tea is subject to high compliance risk as it's not registered with relevant authority as required by law, because Luckin Tea launched its franchise business in September 2019 without having at least two directly-operated stores fully operational for at least 1 year.

This is an article from WeChat official accounts Deep Insights (ID: deep_insights), written by Shen Tan, translated by Chris Yuan.