The most expensive sentence in asset management is two words long.
It scales.
It's the assumption underneath almost every fund's growth story - that if the strategy works at ten million, it works at a billion, just with more zeros. The belief this piece is here to break is exactly that: most things in trading do not scale. Edges are finite, and past a certain size more money doesn't earn more - it earns less, for everyone.
Bigger feels safer. It usually isn't.
Investors read fund size as a signal. A bigger fund seems more proven, more legitimate, safer. Growth gets reported as success, by the manager and the press alike.
But size and skill are different axes, and in trading they often point in opposite directions. The question is never how much has this fund gathered. It's how much can this strategy actually deploy before it stops working. Those are not the same number, and the gap between them is where investor returns quietly go to die.
Edges are finite
Every real edge is small. It's a specific inefficiency - someone paying you to provide liquidity, to be faster, to absorb a structural flow. And every one of them has a capacity: a size beyond which putting in more capital actively destroys the inefficiency it's chasing.
Why size kills the edge
The mechanism is not mysterious. Push more size through a small edge and you start working against yourself:
- you move the price you're trading against - your own buying lifts your entry;
- you eat your own spread as you scale out of a position that no longer fits the liquidity;
- you crowd the inefficiency you were harvesting, until it's arbitraged away;
- you leave footprints big enough for faster players to trade ahead of.
A small edge is a puddle. Pour in a lake and it doesn't get deeper - it just floods the field flat.
The self-dilution math
Here's the part the growth story leaves out. Total return is two things multiplied: return per dollar, times dollars deployed.
That downturn is not a tail risk. It's arithmetic. Past the frontier, growth isn't neutral for existing investors - it's a cost to them. This is the same idea as investor dilution, one level up: there, bad accounting moves value between investors; here, oversized AUM quietly lowers it for all of them at once.
Whose interest is "growth," exactly?
Now the uncomfortable part, because it explains the behaviour.
A management fee scales with AUM. Performance is capped by capacity. So a fund paid mostly on assets is paid to keep growing past the point that's good for you - the fee keeps rising while your return rolls over the top of that curve. The incentives don't just permit over-gathering; they reward it.
A fund that takes every dollar offered is optimising for its management fee, not your return.
That sentence is the whole article. Unlimited fundraising and investor returns are, past capacity, in direct conflict - and most of the industry has quietly chosen the side that pays it.
So we size to the edge
The discipline is the opposite of the default: we size to the measured edge, not to the inflows. When a strategy approaches its capacity, the honest move isn't to let returns sag while the AUM headline keeps climbing - it's to slow intake or close it. We would rather turn money away than dilute the people already in.
The honest caveat: capacity isn't a fixed number carved in stone. It moves with market depth and volatility, so we measure it continuously and pull back early rather than discover the ceiling by crashing into it. We'd rather under-deploy a real edge than over-deploy it into the ground.
The one line to keep
In most businesses, more customers is good news.
In this one, past a point, every new dollar is a tax on the ones already there - and you have to trust that the manager would rather close the door than collect it.