The tax-aware framework
Tax-aware long/short investing is a specific implementation of quantitative equity strategies. It uses the same framework quant firms apply to tax-agnostic portfolio construction. Namely, constructing a portfolio of long and short positions based on expected return models calibrated on a set of proprietary signals, research on equity factors (like value, momentum, quality), and large data sets. These firms then typically use an optimizer to evaluate the characteristics of the portfolio and how individual positions interact with one another (estimated correlations, market impact, volatility, etc.).
The main difference between tax-aware and tax-agnostic versions of these strategies is that the tax-aware optimizer's objective function carries an additional variable. Instead of solving for the highest expected return at a given level of risk, the tax-aware optimizer solves for the highest expected after-tax return (subject to financing and transaction costs).
Over the last several years, AQR Capital Management has published fantastic, in-depth research and articles on tax-aware investing. Their work is essential reading, and largely written for the institutional and advisory channels that allocate to these strategies. This piece is for the investor on the other side of that conversation: the individual or family weighing whether such a strategy belongs in their own portfolio.
Why taxes matter, probably more than you think.
Imagine a high-net-worth investor whose equity portfolio earns 8% annually before taxes. How much of that survives to terminal wealth depends on how often she realizes gains. At the roughly 37% combined federal and state rate that applies to many HNW investors (23.8% federal LTCG plus ~13% in high-tax states), the outcome spans a wide range.1
In the worst case, realizing all gains annually, she compounds at roughly 5% after tax and a $10M portfolio grows to ~$44M over 30 years. A more typical managed account, realizing roughly half of gains annually, implies an after-tax return near 6.5% and a terminal value of ~$67M. A well-run tax-aware strategy, where annual realization falls to around 10% of gains, brings that return to roughly 7.7% and the same $10M to ~$93M, within striking distance of the ~$101M pre-tax outcome. The gap between what most investors typically experience and what tax-aware investing may deliver is roughly $26M on a $10M starting base.2
Inflation makes the picture worse. At 3% inflation the effective tax rate on realized gains rises from a nominal 23.8% to closer to 38%,3 because the tax code taxes nominal gains rather than real ones. Investors who think they are paying long-term capital gains rates are, in real terms, paying something much closer to ordinary income tax rates.
Tax-loss harvesting & direct indexing.
Conventional tax-loss harvesting was the genesis of tax-aware investing. The concept is simple: sell positions that are down, replace them with similar exposures (while observing the wash sale rules) and use or carry the losses forward to defer present or future gains.
This can be applied at the exchange traded fund (ETF) wrapper level, where an investor swaps one ETF for a similar one when the fund itself is down, or at the individual stock level.
These days, this is well-trodden ground, and most investment advisers offer it at little to no cost. However, one issue with tax-loss harvesting is that it requires your position to be down to trigger. After a long bull market, it likely produces little to no advantage.
Direct indexing is the same harvesting technique but seeks to track an index of your choice (the S&P 500, Russell 2000, etc.). The investor holds the underlying constituents of a target index and harvests the losers one by one. Even when the total index is up 8% on the year, individual names may be down, and that dispersion creates harvesting opportunities while maintaining exposure to a diversified index.
In the early years of a direct indexing program, research shows realized losses typically in the range of 0.5% to 1.5%4 of portfolio value annually, with cumulative losses peaking around 30%5 of initial capital before the loss curve flattens. As the portfolio appreciates (markets tend to go up over time), fewer positions are at a loss and harvesting yield falls.
Importantly, direct indexing also produces uncompensated tracking error (deviations from the benchmark that carry no expected return premium).
Direct indexing is now a common offering at almost every major investment advisory firm and has served as an important stepping stone in investing with tax efficiency in mind.
Tax-aware long/short.
For suitable investors, tax-aware long/short equity strategies may present a significant improvement in tax efficient investing. By adding new long and short positions (sometimes called “extensions”), the strategy offers two benefits: 1) the potential for pre-tax alpha (compensated deviation from a benchmark vs. uncompensated in direct indexing) and 2) the potential to generate substantial tax losses.
1) Pre-tax alpha
Overlaying a well-designed long/short strategy on top of your long-only portfolio adds an uncorrelated return stream. A long/short overlay should be market-neutral (not adding additional market risk to the total portfolio) and is expected to produce positive returns. By allowing the portfolio to short securities the model dislikes and to go long those it favors, the long/short structure enables a cleaner, higher-conviction expression of the manager's investment views than a long-only implementation allows.
Pre-tax alpha is a crucial component of tax-aware long/short strategies for two reasons. First, the financing costs of running a long/short book, roughly 50 to 80 basis points6 annually, need to be earned back through excess return before tax advantages can add anything on top.
Second, pre-tax alpha and tax alpha compound together on the same capital base. An investor who earns both ends up with a substantially larger after-tax advantage than either source alone would produce. Said another way, if the manager's long/short strategy produces strong returns, your portfolio grows faster pre-tax (outpacing the benchmark) and allows for the strategy to generate losses on a larger base (more loss generation potential). The opposite is also true. Negative alpha decays the portfolio's growth and loss generation potential over time.
This is why manager quality, investment approach, and track record should be primary considerations when evaluating these strategies, not just the tax benefits.
2) Bi-directional loss generation
In a rising market, short positions in stocks that rise generate harvestable losses for the portfolio. In a falling market, some of the long positions do the same. Because the long/short overlay is an active strategy, the portfolio is continually entering new positions. This implementation helps lessen the “decay curve” that conventional direct indexing runs into.
An AQR paper modeled tax-aware long/short structures (130/30 and 150/50 notional ratios) on the same capital base used for direct indexing.5 Within the first three years of inception, those structures realized cumulative net capital losses exceeding 100% of initially invested capital, while still generating meaningful pre-tax alpha. This means a $5M starting portfolio running the tax-aware strategy could deliver on the order of $5M in tax losses over a 3-year period. That is several times the direct indexing peak on the same dollar base, and the long/short curve does not flatten the same way the long-only one does.
The natural question is what happens at the end, when the strategy is liquidated and all the deferred gains come due. “Isn't this just delaying taxes I'll have to pay eventually?” is a common and fair question. Consistently across literature and empirical results, the deferral advantage survives full liquidation (the next section covers this in greater detail).
Another counterintuitive result in the research concerns the actual source of the net losses. The gross losses realized by a tax-aware version of a long/short strategy are not meaningfully different from the gross losses realized by a tax-agnostic version. What differs sharply is the gross realized gains.7 The tax-agnostic version realizes material gross gains over a multi-year window, while the tax-aware version realizes a small fraction of that. The strategy is not intended to work solely as a loss generator. It seeks pre-tax alpha and aggressively defers gains while taking the losses as they come.
Why deferral is the goal.
A dollar of deferred tax is not a dollar saved. It is a dollar still compounding inside the portfolio. Over a long time horizon, that compounding produces material differences in terminal after-tax wealth.
The most common pushback against deferral is that it is a bet against future tax rates, which may rise. While that is true, practitioners have tested scenarios across a range of future tax rate assumptions. At a 28.8% federal LTCG rate, deferral still wins for almost any reasonable horizon. At 43.4%,8 the rate that would apply if long-term gains were taxed as ordinary income at the top federal bracket, deferral still wins for 10-plus year horizons. Capital gains tax rates would have to reach ~70%9 before crystallizing today's gains beats deferring them.
For a long-term investor with legacy planning in mind, there is another lever as well. Unrealized gains held until death pass to heirs at a stepped-up basis, which means the deferred tax liability becomes zero at transfer.
| Future LTCG rate | Scenario | 10-yr horizon | 20+ yr horizon |
|---|---|---|---|
| 23.8% Current rate | Status quo | Deferral wins | Deferral wins |
| 28.8% Modest hike | +5% | Deferral wins | Deferral wins |
| 43.4% LTCG taxed as ordinary income | +19.6% | Deferral wins | Deferral wins |
| ~70% Historical peak | +46.2% | Approximate break-even | Approximate break-even |
When these strategies make sense.
Diversifying beyond more equity beta
The long/short overlay adds a return stream that has low correlation with the rest of the portfolio. For an investor already broadly diversified in public and private equities, additional equity exposure is largely redundant (the returns move together). A market-neutral long/short strategy can be highly additive as it improves the portfolio's risk-adjusted return profile without layering on more market (beta) risk. For many investors reading about tax-aware long/short, this is a sometimes misunderstood but core benefit of the strategy.
Offsetting a lumpy taxable event
The strategy's loss generation capacity can be particularly valuable when a large gain is on the horizon (carried interest, business/real estate sale, secondary liquidity event, etc.). An investor with an established tax-aware long/short program can use the realized losses to partially or fully defer the gains from elsewhere in their investment portfolio. The strategy needs time to build a loss base, so investors with predictable capital gains events are best served by starting well in advance.
Diversifying a concentrated position
An investor holding large, concentrated equity positions (founder equity, vested employer stock, highly appreciated single stocks, etc.) faces a specific version of the tax problem: selling the concentration generates large immediate gains, but holding it carries uncompensated idiosyncratic risk. Tax-aware long/short can serve as an elegant solution for making diversification economically viable by generating offsetting losses as the investor systematically sells down their overly concentrated position over time.
Long-horizon compounding and estate planning
For an investor with a multi-decade horizon, the deferral advantage compounds every year the strategy runs. The terminal wealth gap between a deferred and a fully taxed portfolio grows nonlinearly with time. For investors planning to pass assets to heirs, the step-up in basis at death can eliminate the remaining deferred tax liability entirely.
What these strategies will not do.
Tax-aware long/short strategies require sufficient time to work properly. Alpha is difficult to generate and loss generation is a multi-year process. An investor expecting meaningful tax benefits within a short period will be disappointed.
Long/short strategies carry material financing and transaction costs, and those costs need to be earned back through real excess return. A long/short manager without the ability to realistically generate alpha (or worse, negative alpha) will potentially produce worse outcomes than a competent direct indexing program would.
While tracking error is expected and intentional, it is not necessarily suitable for all investors. A 150/50 structure can and will deviate from the S&P 500 (sometimes by several percentage points in a quarter10). It is much easier to accept tracking error at the onset of a new strategy, but sticking with it in real time is another story. An investor who loses sleep over intra-year benchmark deviations should be realistic with their expectations and tolerances.
The margin requirements and operational complexity associated with tax-aware long/short strategies are significant. The strategy requires deep quantitative/technical knowledge, institutional infrastructure, and high-quality risk monitoring.
These strategies are not tax shelters. To pass economic substance review under current tax law, a strategy must make sense on its own investment merits, independent of its tax treatment. The clearest evidence that well-designed tax-aware long/short strategies pass that test is that many non-taxable institutions use materially similar long/short equity strategies for their pre-tax merits. In other words, this should be a strategy you want to own on the investment merits alone, with tax alpha as a bonus.
Tying it all together.
Tax-aware long/short combines some of the best aspects of quantitative investment management with a degree of tax efficiency previously not accessible. It allows investors to add sources of uncorrelated return to their portfolio, improve diversification, and significantly enhance after-tax wealth outcomes.
I believe tax-aware strategies will be to the next decade what direct indexing was to the last. The structural argument is just as strong, the academic and empirical evidence is rigorous, and the after-tax gap is much larger than what other approaches deliver.
Even with substantial growth in recent years, tax-aware long/short remains concentrated in institutional and ultra-high-net-worth channels. The strategies are operationally complex, require extensive quantitative abilities to manage, and the underlying math is counterintuitive in places.
My hope is that this primer will serve as a guidepost for those interested in learning more.