In one sentence: After Facebook's one-day drop, Phil argues that your investment story should already assume management will make mistakes and that a moat must carry the company through them. Danielle then walks through Blue Apron step by step and drops it within minutes: weak moat, negative returns and growing debt.
Key ideas
- Your story must survive mistakes. A good story looks ten years out and says the company will be more productive despite management errors. Buffett's idea: buy a business so strong that an idiot could run it, because someday one will. [04:00–06:30]
- Expect the bad years. Like a Nebraska farm with a bad manager or bad weather, plan for bumps, and buy cheaply for a margin of safety in case the future is worse. [07:00–09:30]
- Facebook as a test, not a verdict. Phil says he can't be confident about a tech network ten years out, since technology must keep reinventing itself (Instagram, Snapchat, Friendster) and Buffett and Munger don't own it. He compares it with a railroad that nobody can tweet down. He also notes that owner earnings times ten would put Facebook's price at far more than a conservative value (his figure: about 80 times). Treat that as his rough claim. [04:00–15:00]
- Danielle's research template. She keeps a note template (date, industry, what got me interested, moat, management, numbers) and pastes it for each new company, so she can later check her predictions. [17:00–18:30]
- Look up the industry. Blue Apron's industry on Bloomberg is "food and drug stores", which shows how new businesses fit badly into old boxes. [18:30–20:30]
- A brand moat is the only one it can get. Blue Apron's 10-K talks about brand constantly. Switching is easy (they tried Sun Basket and others in parallel), and the company seems to have introduced the idea and then lost customers to rivals with better retention. [20:30–27:30]
- Start-ups are hard to call. Buffett said he could not have known which early Amazon-type company would win. Some meal kit will win, but picking it now is not an easy decision. [27:30–29:00]
- Understanding and management. She understands the idea but not the logistics (a failed warehouse led a founder to step down). A founder-controlled share structure means outside owners can't change the board. The mission looked like brand-building rather than better food. [29:00–32:00]
- Numbers: return on equity first. About negative 464%, which means "my money is being burned". [32:00–34:00]
- Debt kills. Long-term debt went from zero to $94M to $195M. With no profit, a lender can force bankruptcy without changing a single employee. The house-with-a-mortgage analogy leads to the idea of runway: how long until the money runs out. [34:00–39:00]
- Owner earnings were about negative $300M. Net loss about $210M, adjusted for depreciation, working capital and the full capital spending, because all of it was growth. That means $300M burned in a year. With no sense of when it turns positive, there is no way to put a value on it, so Danielle puts it in the "too hard" box. [39:00–44:00]
How it maps to RuleOne
- The screen filters out exactly the profile Danielle meets here: negative ROE, rising debt and no ten-year record. Seeing it fail the screen is the same quick "no".
- The /stock/TICKER/ page shows ROE and long-term debt first, matching the order of Danielle's checks.
- Her research template is what a notes or Story agent would fill in; the ideas of dating an analysis and re-checking predictions are worth copying.
Buffett, Munger and Graham links
- "A business so wonderful that an idiot can run it": a line Buffett has used in talks. I can't pin a source or year, so don't quote it.
- Buffett on debt and survival: the 2010 letter's section on the risk of leverage is the closest match.
- Graham's margin of safety: The Intelligent Investor, chapter 20.
Words to know
- Return on equity (ROE): net earnings divided by shareholders' equity.
- Runway: how long cash and borrowing can fund losses.
- Dual-class shares: share structure where founders keep control with extra voting rights.
Try this
Open /stock/TICKER/ for any recent IPO you've heard of. Write down ROE and long-term debt for each year shown. If either is negative or climbing and there are fewer than five years of data, write "too hard" and stop, as Danielle did.
Check yourself
- Why does Phil insist the moat should survive "idiots"?
Answer
Management will eventually make mistakes, so the business must stay more productive in ten years regardless. - Why is high debt on a money-losing company so dangerous?
Answer
Lenders can force bankruptcy, wiping out owners even if operations are unchanged. - Why did Danielle stop before finishing the valuation?
Answer
ROE and owner earnings were deeply negative, so no price could be put on it. It went into the "too hard" box.
Short quotes
"Debt kills." (Phil, ~35:00, auto-transcribed)