Luckin Coffee 2020: When the Growth Story Was a Fiction

Situation

It is January 31, 2020. Luckin Coffee's stock price hit $50.02 two weeks ago, a 149% premium over its May 2019 IPO price of $20.38. The Company's market capitalization stands at $12.7 billion. In just 2.5 years since opening its first store in Beijing, Luckin has surpassed Starbucks to become China's largest coffee chain by store count, with 4, 500+ locations.

The market narrative is intoxicating: a tech-forward, app-only coffee brand riding China's secular rise in coffee consumption (from 6 cups per capita per year versus 388 in the US). Luckin commands a 3× price-to-sales premium over Starbucks. Pre-IPO investors include BlackRock and Singapore's sovereign wealth fund (GIC). The company has just completed an $865 million equity and debt raise.

Today, an anonymous research report drops. A short seller, backed by 11, 260 hours of store surveillance video, 92 full-time and 1, 418 part-time staff monitoring store traffic, alleges that Luckin is systematically fabricating core operating metrics:

• Items sold per store per day, inflated
• Net selling price per item, inflated
• Advertising expense, misclassified to obscure true operating margins

Management issues a same-day press release categorically denying all allegations. Research analysts publicly support the company. By mid-February the stock has fully recovered.

Sixty days later, Luckin's internal investigation confirms the fraud.

The decision moment

You are a buy-side analyst at an institutional fund that owns Luckin shares. It is February 3, 2020, two days after the short seller report, one day after management's categorical denial.

You are weighing three questions:

1. How do you evaluate the short seller's evidence? The report is anonymous. But it describes surveillance data that is verifiable in principle. Management's rebuttal contains no specific data, it is a categorical denial. What is the correct epistemological weight to give each?

2. What financial red flags were already in the public filings, pre-report? The company was losing RMB9.2 million ($1.37 million) per day as of Q1 2019. ASP per cup (RMB9.02) was 44% below store-level cost per cup (RMB13.26). The Q1 2019 store-level volume (244 items/day) was 32% below Q4 2018 (356/day). At the same time, the stock was up 237% since September. What does this spread between operational reality and market pricing tell you?

3. Does the governance structure of the company make fraud more or less likely? Luckin's board: six insiders (all with UCAR ties) and two independents. Dual-class shares: pre-IPO insiders held 98.2% of voting power. Management self-disclosed two material internal control weaknesses at IPO. PCAOB could not inspect the auditor's China working papers. The CEO and Chairman had each pledged 47% and 30% of their shares as loan collateral.

You must decide: hold, reduce, or exit your position, and explain your reasoning to your investment committee.

Key financial datapoints

The fraud mechanics

Luckin's fraud was not accounting journal manipulation, it was operational theater. A network of related parties (Charles Lu's relatives, former UCAR associates, Luckin employees) purchased bulk coffee vouchers from Luckin. This inflated reported revenue. Luckin then reimbursed the purchasers through raw material purchases and marketing expenses, disguising the cash outflow. The result: revenues appeared high, the related-party expense lines absorbed the cash refund, and gross margins looked healthy.

The scheme exploited three structural features:

1. App-only transactions, all sales were digital, making fabricated transactions harder to distinguish from real ones
2. Related-party opacity, Luckin disclosed related-party transactions but the scope was under-reported
3. PCAOB blind spot, the Chinese government blocked PCAOB from inspecting EY's audit working papers, reducing external verification

Frameworks invoked

• Earnings Quality Analysis. The gap between reported store-level profit (Q3 2019: RMB 64m) and restated reality (RMB 357m loss) illustrates how accrual manipulation can diverge from cash. Always reconcile operating cash flow to EBITDA.
• Principal-Agent Problem. Insiders controlled 98.2% of votes but only 84.3% of economic interest. The wedge between control and economics creates conditions where insiders can extract value at minority shareholder expense.
• Short Seller Research as Market Signal. The categorical denial without specific data is itself a signal. When management refutes specific, verifiable claims with general language ("we stand by our business model"), this asymmetry warrants escalated scrutiny.
• VIE Structure Risk. Variable Interest Entity structures allow Chinese companies to list abroad but introduce opacity: intellectual property sits in an entity controlled by Chinese nationals, reducing investor recourse.
• Related Party Transaction Risk. The bulk of the fraud was routed through related parties. The disclosure existed, but qualitative language masked quantitative scale. Always model related-party transactions separately.

Discussion questions

1. The short seller's evidence was specific and verifiable: hours of video, field researchers counting customers. Management's denial was categorical but data-free. How should investors weight these two signals, and how does the adversarial structure of short selling affect your confidence in the research?
2. At the IPO, Luckin disclosed two material internal control weaknesses and used the "Emerging Growth Company" exemption to avoid auditor attestation on internal controls. How should an investor adjust their valuation or position size in response to these disclosures?
3. Luckin was losing 44% on each cup of coffee in Q1 2019. The market still valued it at 3× Starbucks' revenue multiple. What narrative-to-numbers failure does this represent, and at what point does optimism become willful suspension of disbelief?
4. The dual-class share structure gave insiders 98.2% of votes. Several insiders had pledged significant shares as loan collateral. How do share pledges by controlling shareholders change your view of their alignment with minority investors?
5. EY was the auditor but PCAOB could not inspect their China working papers. Does the presence of a Big Four auditor on a foreign-listed company carry the same evidentiary weight when regulatory inspection is blocked?

The real outcome (revealed at session end)

April 2, 2020: Luckin discloses that its COO fabricated RMB 2.2 billion in revenue from Q2–Q4 2019. The stock falls 76% and trading is halted.

April 29, 2020: SEC opens investigation.

May 12, 2020: Jenny Qian (CEO), the COO, and six others are dismissed. Charles Lu remains briefly, then is terminated July 5, 2020.

February 5, 2021: Luckin files for Chapter 15 bankruptcy in the US.

June 2021: Restated financials show Q3 2019 was not profitable, the reported RMB 64m profit was actually a RMB 357m loss, a swing of RMB 421m. Revenue inflation in Q3: 83%. Marketing expense inflation: 43%.

December 2021: Luckin settles with the SEC for $180 million.

The lesson: Growth metrics without cash generation are a hypothesis, not a fact. When operating cash flow cannot fund store-level economics, every quarter of reported improvement deserves forensic-level scrutiny, not narrative trust.

Sources

• Wharton Forensic Analytics Lab, "Luckin Coffee" case study (2021). Authored by Erik Vagle, Martin Stapleton, and Daniel Taylor.
• SEC v. Luckin Coffee Inc. (2020 investigation).
• Anonymous short seller report (published January 31, 2020).
• Luckin Coffee F-1 prospectus (May 2019).
• Luckin Coffee restated financials for Q2–Q3 2019 (released June 2021).