Luckin Coffee Fraud Scandal Short Seller Report Stock Price Crash

For professional coffee knowledge exchange and more coffee bean information, please follow Coffee Workshop (WeChat official account: cafe_style).

FrontStreet Coffee - Luckin Coffee Fraud Incident

In recent days, a short-selling report on Luckin Coffee issued by Muddy Waters has caused a tremendous uproar in the capital market. The company that once aimed to take down Starbucks instantly fell from its pedestal. We've seen it rise, we've seen it flourish, and we've seen it collapse... In WeChat groups, I've seen many people rush to use their coupons, as if it's no longer their concern and they can safely watch the drama unfold.

As auditors, this mindset is wrong. As the saying goes, amateurs watch the excitement, while professionals see the substance. We cannot just focus on the spectacle. You must understand that behind this short-selling report lies a multi-trillion-dollar bet, with impeccable logic and overwhelming evidence.

After enjoying the drama, we should also learn from it. If we can grasp the logic and expression of this report and apply it to our future audit explanations and management recommendations, that would be excellent. Many young professionals say they don't know how to write audit explanations—just look at how these experts do it.

The logic of the short-selling report is very clear: first, it briefly presents opinions and evidence, then elaborates point by point. This is typical of Western expression style—conclusion first, then evidence. Unlike some articles that leave you wondering what they're trying to express after reading for ages, resembling love letters that hesitate to speak their mind.

The first part of the short-selling report proves Luckin Coffee's fraud, while the second part proves that Luckin's business model is fundamentally unprofitable. If it only proved fraud, the company could issue an apology, garner sympathy, and the stock wouldn't fall so dramatically. Proving its business model is unprofitable strikes at the root—equivalent to telling investors that the stocks in their hands are worthless, causing the stock price to plummet.

The author's admiration for this short-selling report flows like an endless river.

1. Sales Volume Fraud

Luckin Coffee has over 4,000 stores across 53 cities. Muddy Waters adopted a sampling approach, deploying 92 full-time and 1,418 part-time personnel to collect valid data from 620 stores across 38 cities, recording 981 store-days of operational data. Additionally, 851 store-days of records were invalidated—if store surveillance videos were missing more than 10 minutes, that day's video data was discarded. Such evidence is highly convincing.

They compiled sales orders from these 981 store-days, calculating that stores average 230 sales orders per day.

Some orders might contain multiple items. They collected over 25,000 receipts, calculating an average of 1.14 items per order.

Therefore, average daily sales volume = 230 × 1.14 = 263 items.

After calculating average sales volume, they compared it with the recorded sales volume. By subtracting recorded sales from average sales, they could calculate the inflated sales volume. How was this recorded sales volume calculated?

If orders are consecutive, they could purchase one order at opening time and another at closing time. The difference between these two numbers would reveal the daily recorded sales volume.

In their investigation, they discovered order jumping—for example, jumping from order #271 to #273, with #272 missing. This missing #272 represents a fictitious order—a space for fraud. Why choose order jumping rather than creating more consecutive orders? Because investigative agencies monitor them constantly. Creating false orders would provide concrete evidence of fraud if obtained by investigators. The advantage of order jumping is that the orders never actually existed, so investigators cannot obtain false order evidence!

I remember when queueing at banks, receipt numbers were categorized into several types, such as A0025, B001, C004, etc.

If Luckin Coffee also used several different types of sequential numbering, investigative agencies couldn't calculate store sales volume by purchasing at opening and closing times.

From Muddy Waters' report, we can learn that when auditing large chain stores, we can adopt sampling methods, with samples covering sufficient regions to infer the overall population—this falls within the realm of statistics. Audit procedures should be designed according to the company's specific circumstances. Understanding that all store sales were processed through online ordering with offline pickup, they calculated average daily order documents, achieving 100% completeness. Considering each order might contain multiple items, they collected over 20,000 receipts to calculate the average number of items per order.

Using samples to infer the overall population—this is quite an excellent case study for annual audits of chain stores.

2. New Type of Fraud

They discovered that Luckin inflated advertising expenses by $336 million and revenue by $397 million.

As we all know: Assets = Liabilities + Owner's Equity.

To increase profit, one must increase assets. However, by inflating both expenses and revenue, with a small difference between the two figures, the final net profit wouldn't increase significantly.

There's considerable complexity here. Revenue is at the store level, while advertising expenses are at the group level. To determine if a store is profitable, we mainly look at store revenue minus rent, employee wages, utilities, etc. If revenue is inflated, stores can turn losses into profits—the stores are making money! Advertising expenses are at the group level, and group-level expenses don't affect store-level net profit. This conveys false information to investors: even if the group level is losing money, once advertising is well-established, won't overall profitability be achieved as advertising expenses decrease? As long as stores are profitable, there's hope! Then the stock price surges upward.

Traditional fraud methods often involve increasing both assets and profits simultaneously. Because increasing assets usually involves tangible assets, auditors can easily uncover them. For example, if placed in fixed assets, auditors conduct physical inventory counts; if placed in receivables, auditors send confirmation requests; if placed in bank deposits, auditors send confirmation requests.

The coffee industry is asset-light. Inflating so many assets would be easily detectable. Therefore, they could only make financial statements look better by inflating group-level expenses while increasing store-level profits.

This represents a new type of fraud—inflating group-level expenses while simultaneously inflating store-level profits. Although overall profit doesn't increase, store-level data appears very attractive! Today's chain enterprises and internet companies typically burn money in their early stages—losses in the first few years are normal. As long as market and sales data look good, they can secure funding, stock prices rise, and then founders and investment institutions cash out, leaving retail investors to bear the burden.

A new type of financial fraud has emerged! It doesn't pursue final net profit—it only wants market presence! On NASDAQ, even loss-making companies can go public!

3. Inflated "Other Product Revenue"

The proportion of "other product revenue" increased from 6% to 22%, an inflation of nearly 400%. Coffee sales constitute primary product revenue, while selling nuts, cups, and similar items belongs to other product revenue. There are two procedures to prove enterprise inflation: one is inspection, the other is analysis.

Among the 20,000+ receipts they collected, "other products" accounted for only 6.2%, while the financial statements showed 22%—clearly inflated. These 20,000+ receipts are sufficiently representative of the overall average. They also analyzed the value-added tax rate. Other products have a VAT rate of 13%, while coffee has a tax rate of 6%. Therefore, average tax rate = coffee revenue proportion × coffee tax rate + other product revenue proportion × other product tax rate.

Substituting the values: 22% × 13% + 6% × 78% = 7.54%. However, the average tax rate disclosed in the report was 6.5%. This creates a contradiction with only two possible explanations: either tax evasion or fraud. In either case, the company is doomed.

Uh... The company might argue that the 6.5% was mistakenly posted by an intern who entered the wrong data. Blaming temporary workers or interns has become commonplace. However, they cannot deny that "other products" accounted for only 6.2% in over 20,000 receipts by claiming that Muddy Waters selectively chose receipts with relatively low "other income" during sampling. This is called overwhelming evidence—when multiple pieces of evidence point to the same result, the company finds it difficult to deny.

Some young auditors may not understand the purpose of calculating average tax rates. After seeing this case, you should understand! Analytical procedures are based on mutual verification between indicators with logical relationships—what we typically call reconciliation. If reconciliation fails, there's likely a problem. In this example, sales revenue and output tax have a reconcilable relationship.

Being adept at analyzing the reconcilable relationships between various data is one of the essential skills for excellent auditors. If a company commits fraud, it must do everything comprehensively, ensuring all data reconciles. This requires significant human and financial resources. For example, for manufacturing companies to inflate revenue, they'd need to keep water and electricity meters running 24 hours continuously, fabricate transportation costs,虚构 taxes, and a series of other indicators. Any slight carelessness might lead to omissions that could be detected once discovered. Why can senior partners identify fraud just by looking at company financial statements? Because they understand their industries too deeply, are thoroughly familiar with the reconcilable relationships between various data, and can spot incorrect data at a glance.

Important Notice :

前街咖啡 FrontStreet Coffee has moved to new addredd:

FrontStreet Coffee Address: 315,Donghua East Road,GuangZhou

Tel:020 38364473