Actionable insights — Has Terry Smith lost his mind?!

1. When a disciplined manager changes method, look for the constraint before the argument

1. Write down the manager's stated principles as they were before the change, so the size of the deviation is measurable rather than impressionistic.
2. Read the change announcement for the reason given. Separate "I now believe X" from "I can no longer do Y."
3. Identify the binding constraint: redemptions, mandate, fee structure, career risk, fund vehicle. An open-ended fund's constraint is almost always flows.
4. Ask whether that constraint applies to you. If it does not, the manager's conclusion does not transfer even when the analysis does.
5. Only then judge the timing — a forced change made after the losing period is nearly always made late.

• A constraint dressed as an insight — the change is easier to sell to investors as a new idea than as a capitulation.
• Your own version of the same constraint: a self-imposed benchmark, or a spouse or partner whose patience is the real redemption risk.

2. Decompose relative performance into flows and skill before concluding anything about the manager

1. Compare the index against the average manager, not against the best or worst. If both hold broadly similar businesses, a large gap needs an explanation beyond fees.
2. Quantify the fee difference. Anything the fee cannot explain is a candidate for a flow effect.
3. Chart cumulative flows into and out of each vehicle type and look for the date the performance lines separate.
4. If the separation coincides with the flow reversal rather than with any change in the underlying businesses, treat part of the gap as mechanical.
5. Conclude carefully: mechanical gaps can persist for years and can also reverse, so this is a reason to expect volatility in relative performance, not a prediction.

• Survivorship and composition differences in the "average fund" figure.
• Using the flow explanation to excuse genuinely poor stock selection — it is an additional force, not an alibi.

3. Measure who sets the marginal price — share of volume, not share of assets

1. Distinguish two different statistics: who owns the shares, and who is doing today's trading. Ownership is a stock; price-setting is a flow.
2. Find the share-of-volume number (exchange operators publish it) and compare it with the historical figure to establish direction.
3. Classify the dominant traders by price sensitivity. Index funds, ETFs and quant momentum strategies transact on flows and rules, not on valuation.
4. Conclude what that implies for the price discovery of names you own: less anchoring to fundamentals, larger moves on flow events, longer divergences.
5. Convert it into an expectation rather than a trade — expect valuation gaps to close slowly and to widen further before they do.

• Volume statistics that include high-frequency market makers, which net to no directional view.
• Assuming price-insensitive flows can only push one way; the same mechanism runs in reverse on outflows.

4. Reason about a price move through the order book, not through the narrative

1. Remember that a quoted price is only the last point at which a buyer and a seller both gave in — not a valuation.
2. For any large move, ask which side was impatient and why. Forced buyers and forced sellers move prices far more than opinions do.
3. Ask how thin the other side is. A modest flow into a market with few willing sellers produces a large price move, and vice versa.
4. Apply this to your own orders: when you are the impatient side, you are paying for the immediacy.

• Depth that disappears exactly when it is needed; quoted liquidity is not committed liquidity.
• Reading a big single-day move as new information when it may be a single flow.

5. Find the founding assumption of a popular strategy and check whether it still holds

1. Trace a widely accepted strategy back to the argument its originator actually made, including the conditions they assumed.
2. Write those conditions down as testable statements.
3. Test each against current data. A strategy can remain sensible while the reason it worked has quietly been replaced.
4. If a founding condition has inverted, say so plainly and consider what second-order effects follow — without assuming the strategy immediately stops working.

• The temptation to convert "the assumption broke" into "the strategy will now fail" — the first is a fact, the second is a forecast.
• Thresholds asserted without evidence; nobody knows the passive share at which price discovery genuinely degrades.

6. Treat a respected manager's style capitulation as a sentiment reading on that style

1. Keep a short list of managers who genuinely embody the style you follow, and note when any of them publicly abandons a core rule.
2. Date the capitulation against the style's cumulative underperformance — capitulation clusters near the end of a drawdown because that is when the pressure peaks.
3. Do not act on it as a timing signal on its own; combine it with independent evidence such as relative valuations and factor spreads.
4. Record your own reasoning at the same moment so you can audit later whether you were steady or merely stubborn.

• Confirmation bias — one manager's capitulation is an anecdote, and quality can stay out of favour long after it.
• The difference between abandoning a style and adapting a rule; not every change is a capitulation.