In one sentence: Phil walks through a real loss, Horsehead Holding (ZINC), a zinc recycler he cloned from two great investors and then lost to a bankruptcy over a small loan payment, and draws three lessons: management is untested until it is under pressure, debt kills, and diversification protects you from the thing you didn't see coming.
Key ideas
- How the idea arrived (cloning). Phil saw that two respected hedge fund managers, Mohnish Pabrai and Guy Spier, had bought sizeable stakes via their 13F filings, around 2012 near $11 a share. He treated that as a reason to research, not to buy. [01:00–05:00]
- The business was simple. Horsehead collected toxic steel-mill dust, which steelmakers pay to get rid of, melted it into a zinc-rich product and refined it into zinc. That meant a cheap raw material and a business he judged easy to understand. [02:00–07:00]
- The research path. He read the latest 10-K, then the 10-Ks of the previous five years in order to see the flow of information to shareholders, then analyst write-ups (long and short cases) and earnings-call transcripts on Seeking Alpha. [05:00–09:00]
- The moat was cost. A new plant would make it the lowest-cost producer of high-grade zinc in North America. In a commodity business, where nobody cares who made the product, the low-cost producer survives the price troughs when others shut down. Phil calls this a price moat. [09:00–13:00]
- Valuation by cash-flow multiple. The plant would add roughly $90–110 million of EBITDA. Mining-type companies tend to sell for about 8–10 times that, which is an 8–10 year payback, so a billion-dollar company, or about $20 a share against roughly $11, with upside toward $40 if zinc prices recovered. Phil checked what comparable companies sold for rather than assuming. [13:00–19:00]
- Red flags he explained away. The company had gone bankrupt in the early 2000s on too much debt and low zinc prices. The current management team had less than a 10-year record. Rule #1 asks for ten years of solid history, and he didn't check that box closely enough. Management integrity, he says, only shows under pressure, and he hasn't solved the problem of researching it. [19:00–24:00]
- The path of the trade. He bought around $11, sold near $19 on plant problems, then bought again from about $12 down to about $5, averaging near $7 and eventually about $3 as management said the problems were under control. Management and Pabrai were also buying in 2015. [24:00–27:00]
- What went wrong. The company filed for Chapter 11 rather than fix a missed $1.5 million interest payment on a roughly $30 million loan from Macquarie. A sudden $1.5 billion zinc sale by Glencore had crashed zinc prices from about $1 to $0.60. Phil argues that a management team with no single owner can look after itself instead of the shareholders. (He is careful to say "allegedly", and a lawsuit was expected.) [27:00–29:00, 36:00–38:00]
- Equity comes last in bankruptcy. Debt rose from about $440 million to about $650 million in seven or eight months, mostly legal fees. The judge valued the business at about the debt, so equity was wiped out, even though the court's valuation roughly agreed with the investors' view that the business was worth much more. Phil and Guy Spier won an equity committee after arguing on their own in court. [29:00–35:00]
- Three lessons. (1) You can't count on management to behave well under pressure. (2) Debt kills, even a small amount, because management can walk away from it while keeping their jobs. (3) Diversify, because "buy it as if it's your only company" still isn't a guarantee. The thesis also needed two things to go right (the plant finished, zinc not collapsing). Danielle notes that this was the second time in about 30 years that he had lost this much. He recommends Taleb's The Black Swan. [35:00–40:00]
How it maps to RuleOne
- The stock pages show debt and interest coverage, which is where this case starts. Check long-term debt against a few years of free cash flow, and read the 10-K's debt maturity and covenant notes.
- The ten-year Big Five history on /stock/TICKER/ is the "10 years of solid performance" check that Phil admits he skipped here.
- Cloning is a tip source, as in 001. The planned Radar agent would surface 13F buys, and this episode shows why a 13F buy is a prompt to research, not a conclusion.
Buffett, Munger and Graham links
- Graham, The Intelligent Investor (ch. 14, defensive investor's stock selection, and the margin-of-safety chapter 20): financial strength and a long record of earnings are basic screens, which Horsehead failed on both counts.
- Buffett's Berkshire letters repeatedly warn about leverage and about managers whose incentives differ from owners'. His 1990s letters on "Mr. Market" make the point that you can be right on value and still be hurt by borrowed money.
- Munger's "avoid stupidity" idea and the checklist approach: one missed box (history, debt) can matter more than all the boxes ticked.
Words to know
- Price moat: being the lowest-cost producer, so you stay profitable when the price drops.
- EBITDA: earnings before interest, taxes, depreciation and amortisation, a rough cash-flow measure.
- Equity committee: a group appointed in bankruptcy to represent shareholders, which is rare.
- Chapter 11: US bankruptcy reorganisation. Creditors get paid ahead of shareholders.
Try this
Pick a company you own or watch on /stocks/. Open its stock page and find the long-term debt. Divide it by one year of free cash flow to get "years to pay it off" (Phil's rule of thumb is a few years). Then check whether the company has ten years of solid Big Five history, and write down the one thing that would have to go right for your thesis to work.
Check yourself
- Why did Phil treat Pabrai and Spier's purchases as only a start?
Answer
13F filings lag and say nothing about the business. He still had to read the 10-Ks, the analysis and the industry himself. - What was the main business reason he liked Horsehead?
Answer
A new plant would make it the low-cost producer of high-grade zinc, a price moat in a commodity business, bought at a price below the value of the finished plant. - Which two Rule #1 warning signs did he later point to?
Answer
A previous bankruptcy tied to debt and zinc prices, and no ten-year record of solid performance for the current team. - Why does he say "debt kills" even for a small loan?
Answer
A panicked or self-interested management can choose bankruptcy over a $30 million fix, and equity is wiped out while managers keep running the company.
Short quotes
"Debt kills." (Phil, ~35:30, auto-transcribed)