On this page
The software that harvests losses across 500 stocks has been rebuilt many times, and it is now sold to a retail investor with $20,000 for 9 basis points a year, which is $18 on that account. The same logic, reached through a financial adviser using a platform, arrives at about 115 basis points once the engine, the platform, and the adviser have each taken a layer.
The arithmetic that decides the sign
The long-run benefit of harvesting, for an investor with steady contributions and no recurring short-term gains, is roughly 0 to 30 basis points a year after the first few years. Put the fee next to that number.
| Route | Fee per year | Long-run benefit minus fee |
|---|---|---|
| Retail direct indexing | 9 to 10 bps | about 0 to +20 bps |
| Retail managed account at a large broker | 35 to 70 bps | about −10 to −70 bps |
| Through an adviser and a platform | about 115 bps | about −85 to −115 bps |
The first years are richer, 155 basis points in year one in the standard simulation, and a fresh account earns its fee for a while in every channel. After that, only the cheapest channel keeps a positive sign.
The adviser fee is not the harvesting fee
It does not say advisers are overpaid. The 65 basis points in the adviser layer buys planning, location decisions, giving strategy, and the wash-sale review that no engine performs, and those are the levers worth the most on this site. It says something narrower: the harvesting alone does not pay for the adviser, so the harvesting should not be the reason for paying one.
Minimums, not fees, separate the channels
Retail accounts start at $5,000 to $20,000 at the cheapest providers and $100,000 at the large brokers. Adviser-channel accounts usually start at $100,000 to $250,000. Long/short versions start at $100,000 at one entrant and $1 million at most.