Wisdom

You'll never manage a portfolio for every possible tail event... what is the worst-case scenario? Can I tolerate that loss? It may be an extreme loss. But it's still tolerable. Definable. Tolerable.

Never short the worst 3-5% of companies — the obvious frauds and zombies. Everyone sees them, borrow is expensive, and they are the exact names that squeeze 300% on nothing because path dependency kills you before bankruptcy pays you. Short the next 25% instead: bad businesses with real floats that grind down without the squeeze risk.

Measure a maturity wall in days of market depth, not dollars of balance sheet — refinancing risk is a flow question. A liability many times the cash balance can be trivially refinanceable, and a small one lethal, depending entirely on whether the funding market's daily depth is there when the bill comes due.

Rule No. 1: Never lose money. Rule No. 2: Never forget Rule No. 1.

Don't focus on making money; focus on protecting what you have.

The elements of good trading are: one, cutting losses. Two, cutting losses. Three, cutting losses. If you can follow these three rules, you may have a chance.

After LTCM collapsed in 1998, Ken Griffin studied why: brilliant trades funded by overnight financing that vanished the moment markets panicked. He moved Citadel to locked, term financing that lenders could not pull. In 2008 Citadel's flagship funds fell roughly 55%, but because no lender could yank the lines, there was no forced liquidation. The funds recovered to high-water mark by early 2012. Firms with better positions but worse funding died that year.

The best vehicle for a sector melt-up is cheap BEFORE the melt-up, not expensive. When flows arrive, a name already trading below its fundamental value has two tailwinds pushing the same way — the re-rate and the flow — while the typical bubble darling is already expensive, so its flows fight gravity. Prefer the coiled spring (cheap + marquee story + thin float + unpriced optionality) over the crowded favorite. The re-rate is the thesis; the bubble is a second, optional leg on top — never the reason to own it.

The market doesn't care what price you paid. When a short is working and the fundamentals keep deteriorating, re-up rather than cover — the position at today's price is a fresh decision, and anchoring to your entry is how winners get cut early and losers get held. The same logic runs both directions.

Bet the 57%. Edge in markets rarely feels like certainty; it feels like a coin weighted slightly in your favor. The job is to recognize the 57% situations, bet them at a size that survives the 43%, and take enough of them that the law of large numbers can do its work. Waiting for 90% confidence means never betting, and betting 57s at 90-confidence size means dying.

It's not whether you're right or wrong that's important, but how much money you make when you're right and how much you lose when you're wrong.

The way to build long-term returns is through preservation of capital and home runs.

Bet size should scale with edge and shrink with uncertainty. An oversized bet on a positive edge still guarantees ruin. Ed Thorp ran half-Kelly at his funds: the cost of betting too big is catastrophic and irreversible, while the cost of betting too small is only slower compounding.

Losers average losers.

It was never my thinking that made the big money for me. It always was my sitting.

The first principle is that you must not fool yourself — and you are the easiest person to fool.

Judge decisions by the process that produced them, not by the outcome. Poker players call the error 'resulting.' A win from a bad process is the most expensive lesson the market sells, because it teaches you to repeat the mistake with more size.

Rule number one: most things will prove to be cyclical. Rule number two: some of the greatest opportunities for gain and loss come when other people forget rule number one.

The four most dangerous words in investing are: 'this time it's different.'

The big money is not in the buying and the selling, but in the waiting.

Markets can remain irrational longer than you can remain solvent.

If you want to own duration, own gold. Gold is just a 0% yielding bond of infinite duration, finite issuance, and infinite face value. Why would you own a 10-year Treasury — 4.6% yielding, infinite supply, finite face value, finite yield?

Find the single dominant variable in a thesis — and notice when it is not the fundamentals. Often the biggest force working for or against a position is one person, one regulation, or one funding window, not the science or the numbers. When that variable flips, the whole complex re-rates at once, faster than any DCF updates. The corollary: a rejection on procedure or venue (a closed-door reviewer's objection) carries a different base rate than a rejection on safety or efficacy — change the venue to a public panel and the odds change with it.

Measure yourself against the factor benchmark that actually matches what you own, not the index that flatters you. One period of alpha against one benchmark can be luck; simultaneous alpha against several honest factor benchmarks at once is credible evidence of edge even on a short track record, because the ways to be accidentally right shrink with each independent test you pass.

Company analysis is language learning: thousands of short reps beat dozens of deep dives. Reviewing a company in one or two minutes — filings, chart, capital structure, red flags — thousands of times builds the pattern recognition that makes the rare deep dive fast and the screens trustworthy. Fluency comes from volume of exposure, not intensity of single sessions.

Push versus pull: when counterparties start calling YOU at the lows, that is structural demand arriving. Inbound interest during a drawdown is accumulation-grade evidence for an adoption thesis; interest that only arrives at the highs is momentum-grade.

Renaissance's rule was to never override the models. The moment you override, you no longer have a system, you have moods with a spreadsheet. Jim Simons said the firm's discipline was to trade what the models said even when his own instincts disagreed.

Far more money has been lost by investors preparing for corrections, or trying to anticipate corrections, than has been lost in corrections themselves.

The single greatest edge an investor can have is a long-term orientation.

2008: after $13B of profits in 24 months bred overconfidence, Citadel lost half its capital. Survival came from daily, decisive amputations — sell assets, close lines, suspend redemptions, absorb $500M of costs. 'Often the choice was between painful and more painful, but the one thing we didn't do was put things off. And day by day, we bought ourselves a future.'

Early 2000, Druckenmiller was short tech and right. The positions bled as the bubble made its final run; he capitulated, bought roughly $6 billion of tech stocks near the absolute top, and lost about $3 billion in six weeks. His post-mortem: he knew better, but 'emotional' FOMO took control of his sizing. One of the greatest macro traders ever, undone not by analysis but by watching others get rich.

LTCM, 1998: two Nobel laureates on the board, the best fixed-income arbitrage team ever assembled, roughly 25-to-1 leverage. Nearly every trade they held eventually converged exactly as their models predicted. The firm still died, because the path to convergence ran through a spread widening that leverage could not survive. Being right eventually is worthless if the path bankrupts you first.

Sizing is a solved quant problem that fundamental investors keep doing by feel. Size every position from four quantifiable inputs: volatility, conviction, crowding, and upside asymmetry. Emotion in sizing is the leak: it neither improves your understanding of the company nor your luck.