An investor who sold a business stake this year, or received a first carry distribution, is now facing a $2M capital gain. At the roughly 37% combined federal and state rate that applies to most high earners in high-tax states, the tax comes to about $740,000.1 For a long time, the honest answer for most investors was that not much could be done about the tax burden.
That answer is starting to change. A class of quantitative equity strategies called tax-aware long/short, until recently available mainly to institutions and their largest clients, is built for this situation. This piece explains how the strategies work and the investors they tend to fit.
The size of the problem
Taxes generally receive far less attention than fees, despite being the larger cost. Fees are measured in basis points and negotiated hard, while taxes are measured in percentage points and often go unmanaged. For a long-horizon taxable investor, the tax paid on realized gains will generally cost more over a lifetime than every fee and trading cost combined.
A $10M portfolio compounding at 8% for thirty years finishes near $101M. Taxing the gains each year at the same 37% combined rate lowers the after-tax compounding rate to roughly 5%, and the portfolio finishes near $44M.1
The $57M delta was not all tax burden. Most of it is growth that never occurred, because every dollar paid in tax stops compounding the day it leaves the account. This is a dire case, where all realized gains are taxed annually. The same dynamic, however, exists when realizing any amount of gain.
The standard toolkit, and its ceiling
The familiar tool is tax-loss harvesting. When a position trades below cost, it is sold, a close substitute is held so the market exposure stays in place, and the realized loss is applied against gains elsewhere.
Direct indexing applies the same idea at scale. Rather than holding an S&P 500 fund, the investor holds the five hundred stocks individually. An index that finished the year up 8% still contains dozens of names that finished down, and each is a harvesting opportunity. In the early years, the technique generally adds after-tax value on the order of 0.5% to 1.5% of portfolio value annually.2
The limitation shows up over time, because harvesting requires positions trading below cost and an appreciating portfolio gradually runs out of them. Cumulative losses from direct indexing tend to flatten near 30% of starting capital,3 because markets rise over the long run and rising markets leave fewer positions to harvest.
What adding a short book changes
Tax-aware long/short may remove that ceiling. Alongside its long positions, the portfolio shorts stocks a quantitative model scores poorly, borrowing shares and selling them so the position profits if the price declines. Holding positions on both sides means some part of the portfolio is moving against it in any market environment, which is what allows the structure to generate realized losses across both rising and falling markets.
We have simulated these structures over the last decade of market data. A 150/50 tax-aware long/short portfolio's cumulative net losses pass 100% of starting capital within about three years, several times the direct indexing peak on the same capital base, and the loss generation continues rather than flattening.3
For the investor with the $2M gain, the difference is material. A direct indexing program a few years in might have accumulated enough losses to offset a fraction of the bill. An established tax-aware long/short program of comparable size could plausibly offset all of it, deferring the full $740,000 and leaving it invested.
None of this makes the portfolio a vehicle for generating losses alone. These are investment strategies run for expected return. Carrying a short book costs roughly 50 to 80 basis points a year in financing, and that cost must be earned back through pre-tax alpha before the tax benefits add anything on top.4 Manager quality matters significantly for the successful implementation of these strategies.
Why deferral pays
The most common objection is that deferred taxes eventually come due, so deferral merely postpones the bill. The arithmetic does not support the objection. Tax that has not yet been paid stays invested and earns returns of its own, and over decades those returns compound into significant amounts. George Constantinides formalized the point in 1983, and forty years of subsequent research has refined the result without overturning it.
AQR tested the full lifecycle, thirty years of the strategy followed by a complete liquidation with every deferred gain taxed at exit. Scaled to a $10M start, their modeling has a passive index fund finishing near $78M after the final tax, direct indexing near $81M, and tax-aware long/short near $129M.5
Investors who never liquidate do better still. Under current law, assets passed at death receive a stepped-up basis, and the deferred tax is never paid at all.
Fit, sizing, and costs
These strategies generally fit investors with a taxable portfolio in the millions, a long horizon, and a recurring supply of gains to offset. The natural candidates are founders selling down concentrated stock, PE and VC professionals with carry distributions ahead, and families with appreciated real estate or business sales on the horizon.
Sizing follows from expected gains. Estimate the capital gains realistically expected over the next one to three years, then size the allocation so its loss generation roughly matches. Sensible allocations usually land between 5% and 30% of a taxable equity portfolio, smaller for investors with occasional gains, larger ahead of a major liquidity event.
Several factors are worth understanding before considering one of these strategies. They need years to work rather than quarters, and unwinding a program is a multi-year process of its own. The active book produces intentional tracking error against the benchmark, which some investors find genuinely difficult to live with when the strategy is trailing the index they would otherwise hold. The operational requirements demand institutional infrastructure, with leverage, margin, and continuous risk monitoring. And the pre-tax investment case has to stand on its own. A strategy run purely for its tax treatment is what the economic substance doctrine in tax law exists to disallow.
Where this is heading
Direct indexing went from a specialized institutional service to a standard offering at every major custodian in about a decade. We expect tax-aware long/short to follow a similar path, because the underlying logic is stronger and the after-tax stakes are larger. For now the strategies remain expensive, operationally demanding, and widely misunderstood, which is roughly where direct indexing stood fifteen years ago.