Lesson 1 of 9

Confirmation bias and overconfidence

Step 3 of 6

Worked example

Luckin Coffee, January 2020, from Module 6 lesson 09. Picture an investor who bought after the third-quarter 2019 results and loves the story.

What the results said (6-K, 13 Nov 2019):

Q3 2019 Figure
Total net revenues vs a year earlier +540.2%
Stores 3,680 (+209.5%)
"Store level operating profit margin" 12.5%
GAAP operating loss RMB590.9M
Operating cash flow −RMB122.8M

How confirmation bias could bite. The investor reads the headline margin and the revenue growth as proof. The GAAP loss and the negative cash flow are in the same document, but they get explained away as "a growth company investing". On 31 Jan 2020 an anonymous short-seller report alleges inflated sales; on 3 Feb Luckin says it "categorically denies all allegations". The investor reads the denial, feels relieved, and doesn't read the report.

How overconfidence could bite. The investor has one number in mind for what the shares are worth and sizes the position as if it were certain.

What the bear-case-first habit would have produced, using only Module 6's red flags: revenue per store up 80% in the same year the store count tripled; a positive own-measure profit beside a GAAP loss; cash flow negative. None of that proved fraud. It would have lowered confidence and argued for a smaller position.

What happened: on 2 Apr 2020 Luckin said an internal investigation found about RMB2.2 billion of fabricated sales in Q2–Q4 2019. Hindsight makes this look obvious; it wasn't. The point is not "you should have known". It's that a bear case written in advance gives the uncomfortable facts a place to land.