Misunderstood Cyclicals - What Buffett and Lynch Say Will Make You a Better Investor

ELI5/TLDR

Some businesses move in long, slow waves that last years, not seasons — carmakers, oil, housing, commodities. These cyclical stocks fool people because a single year’s numbers lie about the real business underneath. Peter Lynch says buy them when they look terrible (huge P/E, no earnings) and sell when they look great (tiny P/E, record earnings) — the opposite of normal stock instincts. Buffett adds: judge the business over an entire up-and-down cycle, and only buy the genuinely good ones.

The Full Story

Christian Ryther of Curreen Capital owns a few cyclical stocks and has been chewing on how to invest in them without, in his words, totally losing his shirt. The stakes are high: Lynch reckoned good cyclical investors find their ten-baggers here, while bad ones lose 80 to 90 percent.

What counts as a cyclical

For me, a cyclical is a business or a security that depends on some underlying factor or force which lasts longer than one year.

That one-year line is the whole point. Seasonal stuff — turkeys at Thanksgiving, flowers on Valentine’s Day — washes out inside a normal annual accounting statement. Cyclicals don’t. Their driving force runs longer than the calendar, so any single year’s financials show you a slice, not the shape.

Ryther makes a nice observation about why humans miss this: some things are invisible not because they happen too fast but because they happen too slowly. Watching grass grow shows you nothing minute to minute, yet skip mowing for a year and the lawn has clearly changed. Cyclicals operate on that kind of clock.

The Peter Lynch dream play

Lynch always wants growing earnings working in his favour. With a cyclical, that means buying near the bottom — when the cycle hasn’t turned yet or has just barely started. He looks for a business that can survive hard times (no crushing debt), plus early signs the tide is shifting: shrinking inventories, rising used-car or commodity prices, pent-up demand finally getting tapped.

Crucially, you buy when the P/E looks insane. Earnings are near zero, so the price-to-earnings ratio is sky-high and everything looks awful — and has looked awful for years. Then you hold. You don’t let a double or a triple shake you out; you ride the multi-year absorption of demand. You sell when it all looks wonderful — years of record earnings, the P/E shrunk to single digits.

Buying a cyclical at a single digit P/E after a period on record earnings is a proven way of losing half your money in a very short period of time.

So the normal value-investor reflex — low P/E good, high P/E bad — gets flipped. For a cyclical, a low P/E on record earnings is a value trap, and a high P/E on miserable results can be the buy signal.

What Buffett adds

Buffett buys whole companies, not just shares, and that changes his comfort level. He’ll happily buy as the sun sets on a cycle, because to him the intrinsic value is unaffected by cyclicality — a dollar is a dollar however much it bounces around in between. You add up all the cash over the full cycle, discount it, and that’s the value.

Two things Buffett contributes. First: evaluate the business across the entire rise-and-fall-and-rise, not the last five or seven years, which are probably misleading either way. Ask whether it’s truly a good business over the whole span, or whether you’re just looking at a lucky boom (or an unlucky trough that’s about to end). Second: actually buy the good businesses, not the ones that merely look good right now. Here he parts ways with Lynch, who’d cheerfully buy a mediocre company whose earnings are about to quintuple and hand him a four-bagger.

One caveat from Buffett’s 1979 letter: with securities rather than whole companies, he’s more Lynch-like — he didn’t want to hold medium-term bonds into fierce interest-rate headwinds. Owning the whole business makes riding a cycle down more bearable; holding a security through the same storm doesn’t.

Ryther’s synthesis

Pick good businesses, judged over the full cycle (Buffett). Then buy, hold, and sell their stocks the way Lynch would (cheap-and-ugly in, expensive-and-pretty out). Applied to mid-2017: autos look like a sell, oil he’d avoid entirely, and housing he’d hold through the dips despite weak housing starts.

Key Takeaways

A cyclical is any business or security driven by a force lasting longer than one year — so its single-year financials are structurally misleading.
Seasonal swings (holidays, weather) net out inside annual accounting; cyclical swings don’t, because they straddle multiple years.
Lynch’s play: buy at a high P/E when earnings and sentiment are at rock bottom; sell at a low P/E after record earnings. The opposite of the usual P/E reflex.
A single-digit P/E on a cyclical after record earnings is a classic value trap — “a proven way of losing half your money.”
Entry signals for the turn: low debt, shrinking inventories, rising prices (used cars, commodities) showing pent-up demand starting to clear.
Don’t get shaken out: the payoff is the multi-year ride, so a quick double or triple isn’t the exit.
Buffett: intrinsic value is unaffected by cyclicality — total the cash across the full cycle and discount it; the wobble in between doesn’t change the sum.
Judge a cyclical over a complete up-and-down cycle (ideally 10+ years of history), not the last five to seven, which flatter or punish unfairly.
Buffett buys good businesses across the whole cycle and is willing to buy whole companies even as the cycle peaks; Lynch will buy a mediocre business if its earnings are about to surge.
Owning a whole business makes riding a cycle down tolerable; holding a security into long headwinds (e.g. Buffett’s medium-term bonds, 1979) does not.

Claude’s Take

This is a clear, honest little video from a small-fund manager thinking out loud — no hype, no course to sell, just a guy reconciling two famous investors and admitting where they disagree. That candour is its strength. The core insight is genuinely useful and counterintuitive: with cyclicals the P/E signal inverts, and the most dangerous-looking moment (high P/E, years of losses) can be the best entry.

The Buffett-Lynch tension he surfaces is real and worth keeping: Lynch is a momentum-of-earnings trader who’ll rent a bad business for a good move; Buffett wants to own a good business through the whole cycle. Ryther’s synthesis — Buffett’s business-quality filter plus Lynch’s timing — is sensible, though he’s honest that it’s a personal stitch-up rather than a proven system. He also waves at the hard part without solving it: knowing when a beaten-down cyclical is genuinely turning versus just cheap-and-still-falling. “You have evidence or a strong reason to believe that is turning” is doing a lot of work there.

Docking points because it’s thin on mechanics (how to actually source full-cycle history, how to size positions against the 80-90% downside risk he opens with) and the 2017 calls are now a dated time capsule rather than a method. But as a mental-model primer it punches above its nine minutes. A 7.