Actionable insights — Quality vs Momentum

1. Diagnose an underperforming style before you change it

1. Separate the two possible causes of a drawdown: the businesses are doing worse, or other buyers are paying more for something else.
2. Test the first directly — check revenue, earnings and cash flow of the holdings themselves, not the share prices.
3. If the fundamentals are intact, the drawdown is a change in what the market will pay, i.e. a multiple, and multiples are the part of a return that mean-reverts.
4. Convert the finding into one of exactly two actions: hold, or add. Selling belongs to the fundamentals branch of the diagnosis, not this one.
5. Write down what would move you to the other branch, before the next fall.

• The diagnosis becoming unfalsifiable — "it's behavioural" is the same sentence a wrong thesis produces on the way down.
• Fundamentals that are intact only because the estimates have not been updated yet; check reported figures, not forecasts.

2. Use a historical analogue with the pain attached

1. Pick a precedent where the same style was declared dead, and name the dates.
2. Report the drawdown as well as the recovery — the size of the loss endured is the part that makes the analogue useful.
3. State the return from buying at the moment of maximum ridicule, not from the top or from the bottom.
4. Then list the differences between then and now, so the analogy can be argued with.

• The analogue chosen because it resolved well. For every 1999 there is a style that never came back.
• Survivorship in the "quality" label itself: the index of quality stocks is reconstituted, the portfolio is not.

3. Judge a defensive strategy on the recovery, not just the fall

1. For any strategy sold as defensive, measure both halves: how much less it falls, and how quickly it regains the prior high.
2. Prefer the second measure. A smaller fall with no recovery is a slower loss; a fast recovery compounds from a higher base.
3. Check across more than one episode — dotcom, 2008, 2020 behave differently.
4. Demand the source of any recovery statistic before repeating it.

• Unattributed multiples ("9x faster") that get repeated until they are treated as established.
• Definitions: "quality" as an index factor is a screen on profitability and leverage, which is not the same as this portfolio's seven criteria.

4. Let the contribution schedule replace the timing decision

1. Fix a monthly amount and a date, and make the purchase regardless of the market's level.
2. Accept the trade explicitly: if prices rise you gain on what you hold; if they fall you buy the next tranche cheaper. Both branches are acceptable, which is what removes the decision.
3. Run the arithmetic once, so you can see how much of the terminal figure comes from contributions rather than from the assumed rate of return.
4. Direct each month's tranche by a rule (e.g. the worst performers of your own pre-vetted list) rather than by conviction on the day.

• Sixty-six-year projections at 12%: the point is the mechanism, not the number. Small changes in the rate or in longevity dominate the result.
• A schedule that is only followed while it feels good — the tranche bought in the worst month is the one that earns the average.

5. State the opposing case with its evidence, then answer it

1. Write the strongest version of the argument against your own positioning, including the data that supports it.
2. Identify precisely which part you dispute — usually not the observation, but the inference drawn from it.
3. Answer with a mechanism, not a preference: why does the observed gap close, and what closes it?
4. Keep the record, so the answer can be scored later.

• "Sooner or later" as a risk-management plan: a mechanism with no horizon cannot be falsified within an investor's patience.
• The dual role of a free post like this one — it is also a marketing document for the subscription, so the confidence in it is not costless.