Actionable insights — Morgan Housel on Getting Rich

1. Judge a track record by its length before its size

1. Before comparing returns, write down the number of years each record covers.
2. Discount short records heavily regardless of magnitude — a spectacular three years carries almost no information.
3. Ask what the manager did in the drawdowns inside the record, not just the compound number at the end.
4. Apply the same test to yourself after a good run, and ask which decisions were process and which were luck.

• A long record that spans only one regime — forty years of falling interest rates is not forty independent tests.
• Your own recency bias after a strong year. The post's warning is directed at the reader first.

2. Treat leverage as the one mistake you cannot recover from

1. Separate mistakes that cost you a position from mistakes that cost you the game. Only the second category matters.
2. Refuse borrowed money on the portfolio regardless of how attractive the opportunity looks, because the cost is forced selling at the worst moment.
3. Check indirectly too — margin loans, leveraged funds, and companies whose own balance sheets carry the leverage for you.
4. Accept a slower path as the price of never being removed from the table.

• Leverage entering through the back door — a holding company or serial acquirer whose debt is the source of its returns.
• The rationalisation that a particular opportunity is safe enough to borrow against. Guerin's opportunities were the same ones Buffett was buying.

3. Screen for survivability first: cash on hand, no debt

1. Before assessing growth or moat, ask a single question: can this business survive a decade in which nothing goes right?
2. Require a net cash position or debt that can be serviced out of a trough year's cash flow.
3. Prefer a business that can fund its own downturn over one that must refinance in it.
4. Treat conservative balance sheets as the enabling condition for compounding, not as a drag on it.

• Cash-rich businesses that are cash-rich because they have nowhere good to reinvest. Survivability without growth is a different investment.
• Debt reported net of cash the company cannot actually access (trapped overseas, or pledged).

4. Optimise for never having a bad year rather than for having great ones

1. Invert the search: instead of hunting for the best investment, build a list of the ways this investment could go badly wrong and eliminate candidates on that list.
2. Accept unremarkable results in good years as the price of avoiding disastrous ones.
3. Measure yourself on the distribution of yearly outcomes, not the average — the worst year is the number that decides where you finish.
4. Recognise that arithmetic does the rest: consistently average results compound past volatile ones.

• Avoiding mistakes turning into avoiding decisions. The record above still compounded; it was not cash.
• Risk that has been hidden rather than removed — steady returns produced by illiquidity or by selling volatility look identical to skill until they don't.

5. Extend the holding period before you try to raise the return

1. When looking to improve an outcome, treat duration as the first lever and rate of return as the second.
2. Compute what the same rate produces over 30, 40 and 50 years before reaching for a higher-risk strategy.
3. Structure decisions so that the holding period is protected — no leverage, no forced selling, no strategy you would abandon in a drawdown.
4. Choose the approach you can sustain over the one that scores highest on paper.

• A duration argument used to excuse a poor holding. Time multiplies losses too.
• The gap between an assumed 50-year horizon and an actual investing life. The 50-year figure needs you to start young and never interrupt.

6. Build the process around behaviour, because that is the variable you control

1. Write down the three or four rules you will follow mechanically — savings rate, quality bar, holding period — and treat them as the strategy.
2. Keep the rules simple enough that they survive stress, since a rule abandoned in a drawdown was never a rule.
3. Prefer doing less: fewer trades, fewer decisions, fewer opportunities to act on emotion.
4. Assume no informational edge, and let the behavioural edge be the whole thesis.

• A behavioural argument used to justify never re-examining a holding. "Sit and wait" is not the same as "never reassess" — the same author sells a position nineteen days later on exactly such a reassessment.
• Survivorship in these stories. For every Ronald Read there are savers who held the wrong three companies.

7. Do not build a plan that requires a forecast to work

1. Identify the forecast each holding secretly depends on, and prefer holdings that depend on fewer of them.
2. Assume the biggest determinant of your outcome will be something you could not have anticipated.
3. Design for robustness under a wide range of futures rather than for accuracy about one.
4. Keep participating long enough for the favourable random events to find you.

• The same logic applied selectively. A newsletter that says prediction is pointless still publishes ten-year EPS growth assumptions in its valuation models.
• Confusing "don't forecast macro" with "don't underwrite the business". The second is still required.