ASSET MANAGER DIGEST

Twenty-one firms run $42 trillion of US-registered funds between them. Every one of them sells at many prices — the fee line inside a single manager can span a hundred basis points — but asset-weighting collapses each shelf to one rate each firm realizes. The rates diverge profoundly: per $100 billion managed, $565 million a year to First Trust but only $73 million to Vanguard. Exhibit 1A is the industry's strategy map in one column while Exhibit 1B shows what scale achieves (Fidelity's mid-panel rate out-earns every firm in the market and Vanguard's floor price still ranks third by sheer weight of assets).

Price is also where the pressure has begun to show. Eight of the 19 firms with a complete flow record are in outflow over the ten months compared and the list reads as the roll-call of traditional active management: Capital Group (−$94B, the panel's largest), T. Rowe Price, MFS, Nuveen, Columbia, Franklin, PGIM, Jackson. Every firm still growing at a fee above 50bps is a specialist.

In the following pages, we explore these and other market dynamics further. The data has always been available — but assembling and digesting, well, that's our job.

The industry's favorite defensive statistic — asset-weighted fees fell 12.5 basis points over four years — is true and mostly beside the point. On the fixed cohort of funds present in both years, only 1.5 points is funds cutting fees. The rest is reallocation: money walking from expensive funds to cheap ones, and expensive funds closing while cheap ones open. Anyone quoting the headline as evidence of pricing pressure on managers is quoting a number that is 88 percent flows. The pressure is real, but it arrives as redemptions, not as repricing — and on the shelf itself, more than a third of classes raised their price.

Where the price range does exist, almost nobody pays the top of it. The same portfolio sells at prices a hundred basis points apart, and a median 0.9 percent of fund assets sit in the dearest class against 31 percent in the cheapest. The expensive classes are not extraction; they are channel architecture — compensation for a distributor on the flows that need one. This was unmeasurable until class-level assets were parsed out of shareholder-report financial statements: nothing in the SEC's structured data joins a class's fee to its assets.

Strip out the management fee and the distribution fee, and what remains is what it simply costs to operate a fund — custody, administration, audit, transfer agency. That line falls 11.6-fold as funds grow, and stops falling at $2–10 billion. The same threshold appears independently in a nine-firm panel one-fifth the size, and again in the structure data: large complexes take administration and transfer agency in-house, small ones buy them.

Below $100 million, the operating line alone runs 51 basis points — more than the entire price of a competitive index fund, before the manager earns anything. Anyone can launch a fund. Almost nobody can run a small one economically, and the panel's own smallest funds escape the curve only by amortising against a large platform. This is the difference between a barrier to entry and a barrier to scale, measured.

Take every fund with a primary custodian and ask one question twice: of those that keep a committed credit line, does the lending group include the custodian? Under third-party custody the answer is 95.8 percent yes — the bank that holds your assets is the natural lender against them. Under affiliated self-custody it is 0.8 percent: exactly two funds of 237. The split is the law — §17(a)/(d) and Regulation W bar a bank from lending to its own funds — visible in the data as a cliff. Blend the populations and the figure reads 91 percent, which describes neither. Read the split, never the blend.

The same either/or shows up in what fund complexes keep in-house. No service layer sits between 6 and 26 percent affiliated: custody and valuation have completed the move to utility; administration and transfer agency never left home. And in every layer the asset-weighted share runs far above the fund-count share — large complexes keep what small ones must buy, the Scale Line's threshold showing up as an organizational decision instead of a cost curve.