Got more details from my fairly large pure play RIA’s investment team. They’ve been using this for quite a few clients. Sharing the detail mechanics here.
To answer my own orignal question : I think its not possible to DIY. AQR is the leader (biggest and longest) in this leveraged long short scheme to generate losses pioneered out of UChicago research awhile back (AQR principles have lots of UChicago alums)
Example
Goal : Have long term $10M nVidia stock with $1M basis. Want to diversify but minimize tax impact.
Assumptions : Look at just federal for now ignoring state tax differences.
- AQR flex 250 (cost 150 basis points/year. borrow 3?x margin for long short hedge to generate losses) Historically generates 70% loss year1, 110% cumulative year 2.
- Year 1 start near Jan1 to generate full year 70% loss, sell ~70% of nVidia early in the year. End of year tax neutral (CG gain cancelled by the loss)
- Year 2 generate additional 40% loss. Sell remainder 30% of nVidia early in the year. Do similar as year 1.
- After about 18 months. nVidia is fully diversified into few hundred individual positions tracking an index such as SP500. Then exit AQR flex 250 and setup DI (Direct Index) account to reduce cost to manage hundreds of positions with long strategy (~20 basis points)
Note the diversified portfolio still have huge 90% unrealized gains. Only thing we’ve acheived is derisk a concentrated position to diversified one with tax neutrality.
Federal Tax Comparision
24% (20% LTCG cap + 4% NIIT) vs 1.5% AQR Flex 250 fees/year x 2 years + ( < 1% RIA fees/year x 2 years ) (typically lower with larger portfolio > $2M+)
RIA fees unavoidable as AQR don’t deal with individuals (unless you are probably ultra rich haha)
If portfolio are smaller, calculus gets trickier with lower tax brackets etc.
Misc
- This method can also be used to offset future anticipated large CGs, can plan early to generate offsets. Need to do the calculus on value of course.
- Starting asset with higher basis % generate losses faster
- Higher leveraged method like AQR flex 250 generate losses faster.
- Huge number of positions is a natural result. You need a lot of positions to do long/short hedge and harvest the losers. This results in eventually a DI portflio which might be 10 basis points/year higher a cheap index fund cost to hold the resulting positions.
- Tax forms becomes huge with high number of positions. DIY with turbotax and efile is not a problem but amended returns will be printing a book.
Summary
In the above very low basis % example, its basically ~2 tax year sprint to fully diversify with a high leveraged mechanism. Result afterwards is a DI portfolio with hundreds of positions.
Probably people here don’t like fees But the fees can generate real tax benefits if calculus works for your situation.
Side Note
I asked the RIA why they didn’t just use DI (accumulat some losses as an added benefit) instead of index funds when I started with 100% basis assets 8 years ago. Their response was quite interesting. Trade fees disappeared after online competition arrived a few years back which drove the popularity of DIs. Such numerous trades would add up costs quickly if fees persisted. Anyway, with $0 trades, DIs are becoming more popular.
