None of the index-default arithmetic says never do your own work. It says the burden of proof sits on the sleeve, and it prices that burden exactly.
An index core at 10 basis points and an active sleeve at 150 basis points differ by 140 basis points. So the sleeve must earn 1.40 percentage points of gross alpha a year merely to break even against the capital it displaced, at any sleeve weight, because the break-even is a property of the cost difference and not of the size.
Producing an actual improvement is much harder, because the sleeve's contribution is diluted by its weight:
| Sleeve weight | Blended cost | Gross alpha needed to lift the whole portfolio by 0.50 pp |
|---|---|---|
| 10% | 0.24% | 6.40 pp a year |
| 15% | 0.31% | 4.73 pp |
| 30% | 0.52% | 3.07 pp |
| 50% | 0.80% | 2.40 pp |
| 100% | 1.50% | 1.90 pp |
Read the first row carefully. A 10 percent sleeve that beats the index by 6.40 percentage points a year, every year, lifts the whole portfolio by half a point. That is a spectacular record producing a modest result, and it is the arithmetic reason a small satellite is mostly a device for keeping you interested rather than a driver of wealth.
Read the last row too, because it cuts the other way. Concentrated capital is how alpha actually reaches a portfolio, which is the concentration argument from the portfolio-construction phase arriving from a different direction. If you genuinely have an edge, a 10 percent sleeve wastes it. If you do not, a 100 percent sleeve is catastrophic. The sizing decision is therefore identical to the edge decision, and it cannot be hedged by splitting the difference.
Advertisement
How do you decide what goes in the active sleeve?
So the sentence that decides which capital enters the sleeve has to be strict, and it has three clauses. Name the edge, meaning the specific reason your estimate of value is better than the price. Name who is on the other side and why they are selling to you, because if you cannot, the mispricing you think you found is probably a fact you have not learned yet. Name the evidence that you have the edge, which means a written record of prior decisions and outcomes, not a feeling about your own judgement.
Capital that cannot answer all three goes to the index. That is not modesty; it is pricing. An 85/15 book costs 0.31 percent against 1.50 percent for an all-active one, a saving of 1.19 percentage points a year, which over ten years at 8 percent gross is 10.22 percentage points of terminal wealth. You are being paid that much to admit what you do not know.
Which kinds of edge survive scrutiny?
The categories of edge that survive scrutiny are narrow and mostly structural rather than analytical. Time-horizon arbitrage, where you can hold through a period of poor reported results that a mandated manager cannot. Size, where you can own a position too small to matter to an institution. Forced-selling situations, where the seller is acting under a rule rather than a view, which is the whole subject of the event-driven elective. Notice that none of those is "I read the annual report carefully". Careful reading is the entry ticket, not the edge.