A tax-aware long/short portfolio can create substantial flexibility during its life, yet the exit deserves the same level of planning as the entry. A 2026 paper by researchers at AQR Capital Management finds that investors may be able to withdraw meaningful amounts of cash while preserving the economic characteristics of the remaining strategy. Eventually leaving the long/short structure introduces a different challenge because short positions, financing, and appreciated long positions do not have identical tax consequences.
For some investors, the appreciated long securities remaining after deleveraging may create an additional planning opportunity. A qualifying Section 351 ETF exchange can potentially move an already diversified basket of appreciated securities into an ETF without immediate gain recognition. The sequence requires careful investment, tax, legal, manager, and custodian coordination.
| Short Answer: How to Exit a Tax-Aware Long/Short Strategy
An investor may be able to take substantial cash from a tax-aware long/short portfolio without fully liquidating it. A complete exit is different because shorts and financing must eventually be removed. Appreciated long securities that remain may have additional planning options, potentially including an in-kind transfer and a qualifying Section 351 ETF exchange. |
Key Takeaways
The most useful way to evaluate a tax-aware long/short strategy is across its entire ownership lifecycle rather than through a single tax year.
- AQR’s 2026 simulations found that tax-aware long/short portfolios could support substantial partial withdrawals while maintaining target leverage, tracking error, and pre-tax investment efficiency. Withdrawal capacity and complete liquidation are fundamentally different questions.
- A complete exit requires dealing with the short book and financing structure. Tax-efficient withdrawals can also leave the residual portfolio with increasingly large embedded gains relative to its remaining value.
- Appreciated long securities remaining after a transition to long-only may, in suitable circumstances, be transferred in kind and potentially contributed to an ETF through a qualifying Section 351 transaction. Section 351 generally defers recognition of embedded gain through carryover basis. It requires careful qualification and current tax advice.
Why Exit Planning Matters From Day One
Tax-aware investing works best when we view taxes as one component of portfolio construction rather than as an isolated year-end exercise.
A long-only portfolio such as direct indexing can realize losses when individual securities trade below tax basis. Over time, a rising market can leave fewer positions below basis and a larger collection of appreciated holdings. A tax-aware long/short strategy has a broader opportunity set because it owns additional long positions and short positions while managing the portfolio around an investment model.
The underlying tax mechanism deserves precision. Research published by AQR authors in Loss Harvesting or Gain Deferral? found that much of the net capital-loss generation in tax-aware long/short strategies came from deferring taxable gains while continuing to realize losses through normal portfolio turnover, rather than simply trading more aggressively for the sole purpose of harvesting losses.
For readers who want the broader foundation first, our discussion of The Key to Tax-Aware Investing & TALS™ explains how long and short positions expand the portfolio-management toolkit. We also walk through practical tax-planning applications in Tax-Aware Investing in Practice.
The exit question then follows naturally. A portfolio designed to defer gains can become rich in unrealized gains over time. Accessing cash, reducing leverage, or eliminating the long/short structure requires deciding which gains need to be recognized and which can reasonably continue to be deferred.
What AQR’s 2026 Withdrawal Research Actually Found
AQR’s August 28, 2026 working paper, Liquidity Without Liquidation: Tax-Efficient Withdrawals From Tax-Aware Strategies, directly studies whether investors can obtain cash from tax-aware portfolios without fully liquidating them.
The study models direct indexing alongside 150/50, 200/100, and 250/150 long/short portfolios. These labels describe how much the strategy owns long and sells short relative to the investor’s net capital. A 200/100 portfolio, for example, has approximately $200 of long positions and $100 of short positions for each $100 of net asset value.
AQR used the Russell 1000 as the benchmark. Direct indexing was constrained to no more than 1% tracking error. The 150/50, 200/100, and 250/150 strategies targeted 2%, 4%, and 6% tracking error, respectively. Tracking error estimates how much a portfolio’s returns may vary relative to its benchmark; a higher target gives the manager more latitude to differ from the index. The long/short strategies targeted market exposure of 1.0, meaning they were designed to retain roughly full equity-market exposure. They held individual equities and rebalanced monthly.
Key assumptions from the AQR withdrawal study:
| Assumption | AQR Study Design |
| Benchmark | Russell 1000 |
| Simulation period | 26 ten-year histories beginning each January from 1989 through 2014; the last ended in December 2023 |
| Tracking-error targets | Direct indexing up to 1%; 150/50 at 2%; 200/100 at 4%; 250/150 at 6% |
| Market exposure | Targeted at 1.0 for the long/short strategies |
| Rebalancing | Monthly |
| Implementation constraint | Transaction and financing costs included; returns shown gross of management fees |
| Costs | 100 basis points annually per unit of one-sided leverage; approximately 100 basis points annually for the modeled 200/100 strategy |
| Financing assumption | 100 basis points annually per unit of one-sided leverage |
Those assumptions matter. These are historical portfolio simulations, not forecasts of what any investor should expect.
How Large Were The Simulated One-Time Withdrawals?
AQR calculated the largest modeled year-end cash withdrawal consistent with the study’s tax and portfolio constraints. The term “maximum withdrawal” is important. It is not a sustainable spending rate, a recommendation, or a guarantee.
Average maximum one-time withdrawals as a percentage of initial capital (AQR 2026 historical simulations; not TWD performance):
| Strategy | End Of Year 2 | End Of Year 8 |
| 150/50 | 43% | 21% |
| 200/100 | 56% | 46% |
| 250/150 | 65% | 67% |
The denominator is particularly important. A 56% year-two result for the 200/100 strategy means 56% of the investor’s initial capital in the simulation, not 56% of the portfolio’s then-current value.
AQR also tested recurring maximum withdrawals. When those withdrawals began at the end of year two, cumulative average withdrawals over the ten-year investment period reached approximately 102% of initial capital for 150/50, 124% for 200/100, and 139% for 250/150, compared with 62% for direct indexing. Cumulative withdrawals can exceed initial capital because the investment itself can grow during the period.
AQR also tested fixed annual withdrawals equal to 5%, 10%, or 15% of current NAV beginning at the end of year two. Those simulations show why withdrawal design matters: a rigid schedule can eventually force gain realization when the portfolio’s available losses do not align with the required cash withdrawal. In the modeled periods, direct indexing eventually realized gains even at 5% annual withdrawals; the 150/50 and 200/100 strategies handled 5% more comfortably, while the 250/150 strategy accommodated 10% without material gain realization. Higher fixed withdrawals increasingly required gains later in the investment horizon.
Liquidity can also carry an economic cost. In its after-tax wealth analysis, AQR assumed withdrawn capital was reinvested at an 8% after-tax return. Under that assumption, earlier withdrawals generally produced lower ending after-tax wealth because less capital remained in the tax-aware strategy, which had higher modeled after-tax returns. This is a modeling result rather than a forecast. It illustrates the tradeoff between present liquidity and the potential value of leaving capital invested.
Zero Net Capital Gain Requires An Important Qualification
AQR’s maximum-withdrawal process uses the losses generated during the first eleven months of a calendar year to determine how much cash can be raised in December.
When sufficient current-year losses exist, the withdrawal is sized so withdrawal-related gains can use those losses. A net capital loss means realized capital losses exceeded realized capital gains for the period. Prior-year capital-loss carryforwards are assumed to have already been used and are excluded from the calculation.
A year can still finish with a net capital gain. If the portfolio has already generated a net gain during the first eleven months, AQR makes no December withdrawal under the maximum-withdrawal rule. The strategy may still record a net gain for that calendar year.
That detail points directly to household-level planning. An investor with substantial loss carryforwards may rationally accept gains while raising cash. An investor preparing for a business sale or another large taxable event may instead want to preserve losses for that event. The economically appropriate withdrawal therefore depends on the investor’s entire tax picture.
Withdrawal, Transition To Long-Only, And Liquidation Are Different
| In Plain English
A partial withdrawal may be tax-efficient because the manager can use available losses while keeping the remaining strategy intact. A complete exit is harder because the short positions and financing structure eventually have to be closed, and appreciated long positions may carry substantial unrealized gains. |
The word “exit” can describe several very different transactions.
Four different versions of “exit”:
| Planning Decision | What Happens | Does The Long/Short Strategy Continue? | Principal Tax Issue |
| Partial Cash Withdrawal | Selected positions are traded to raise cash, and the residual portfolio is resized. | Yes | How much gain must be recognized to fund the withdrawal. |
| Deleveraging | Long and short extensions are reduced. | Possibly | Gains and losses created while reducing extensions. |
| Transition To Long-Only | Short positions and financing are removed while appreciated long positions may remain. | No | Tax cost of closing shorts and selling enough longs to remove leverage. |
| Full Liquidation To Cash | Longs are sold and shorts are covered. | No | Recognition of embedded gains across the portfolio. |
A withdrawal from a $10 million 200/100 portfolio illustrates the first category. Suppose the investor wants $1 million of cash. AQR’s optimization does more than sell 10% of each position.
After the withdrawal, a $9 million residual portfolio is rebuilt to approximately $18 million long and $9 million short, preserving the 200/100 structure and the 4% tracking-error target relative to the reduced NAV.
That distinction explains why liquidity can be available even when complete liquidation would be expensive.
AQR also found that after a one-time withdrawal, annual net capital-loss generation relative to the remaining NAV generally returned to patterns similar to portfolios that had made no withdrawal. Less capital remains invested, so the absolute future tax capacity is lower, but the remaining portfolio’s tax-management process can continue.
Why The Remaining Portfolio Can Accumulate Larger Embedded Gains
Tax-efficient withdrawals naturally affect what remains.
When an optimizer needs to raise cash, positions with losses or smaller embedded gains are generally more attractive sources of liquidity than highly appreciated positions. As withdrawals accumulate, the appreciated positions can become a larger percentage of the residual portfolio.
AQR observes exactly this result. Built-in gains as a percentage of remaining NAV generally increase with larger withdrawals.
This creates an important lifecycle tradeoff. Current liquidity can be obtained efficiently while future liquidation becomes more tax-sensitive.
The short book adds another dimension. Appreciated long securities can potentially continue to be held, transferred in kind, donated, or qualify for a basis adjustment at death under current federal income-tax rules, depending on the circumstances. Short positions do not receive basis-step-up treatment in the same way.
Closing short positions can also create taxable gains. AQR’s simulations show meaningful short-term embedded gains concentrated predominantly in the short book. Actual tax character depends on the specific positions and applicable short-sale rules, so this portion of an exit requires individualized tax analysis.
How A Tax-Aware Long/Short Portfolio Can Transition To Long-Only
AQR models a transition by eliminating the long and short extensions.
On the long side, tax lots are ranked by their tax impact and sold in a tax-efficient order until enough of the long extension has been removed to repay financing. On the short side, every remaining short position is ultimately covered because an investor cannot reach a long-only portfolio while continuing to maintain the short book.
The resulting long-only portfolio deserves an important qualification. AQR does not assume that it instantly looks like a conventional index portfolio. The study assumes a tax-aware long-only manager can subsequently move the remaining holdings toward the desired benchmark over time.
AQR researchers’ 2023 Beyond Direct Indexing study found that tracking error and leverage can be reduced substantially over time in a tax-aware manner. Read the 2023 research. Separately, a July 31, 2025 AQR educational article discussed one possible unwind that leaves an appreciated long-only portfolio, which could then be managed through direct indexing or potentially contributed to an ETF through a Section 351 exchange. That educational discussion is separate from AQR’s 2026 withdrawal study. Read AQR’s July 2025 lifecycle discussion
For investors evaluating TALS™, our TALS™ service overview describes how we think about integrating long/short investing with broader financial and tax planning.
The central planning question is broader than “When should we liquidate?” We would ask how much liquidity is required, how quickly portfolio risk should change, which tax attributes are most valuable elsewhere, and what the investor should ultimately own after the long/short structure has served its purpose.
Can Appreciated Long Positions Leave The SMA In Kind?
Potentially.
A movement of securities between taxable accounts owned by the same beneficial owner can generally occur without a market sale. The transferred securities generally retain their existing tax basis, so the unrealized gain remains deferred rather than being triggered by the transfer itself.
A long/short SMA adds operational constraints. The securities may be supporting margin requirements. The manager may have rules governing partial distributions. The custodian may have procedures around selected tax lots, collateral, or positions held while shorts remain open.
We would therefore verify the operational path before building a tax plan around it.
If an investor can reach a long-only portfolio containing a suitably diversified collection of appreciated securities, preserving those securities rather than selling them solely to reduce administrative complexity can create additional options. One of those options is a Section 351 ETF transaction.
What Is A Section 351 ETF Exchange?
| In Plain English
Section 351 may allow an investor with an already diversified portfolio of appreciated stocks to exchange those securities for ETF shares without recognizing the gain immediately. The gain generally remains embedded through carryover basis, so the strategy defers tax. It does not automatically eliminate the capital gain. |
Section 351 of the Internal Revenue Code generally provides nonrecognition when one or more persons transfer property to a corporation solely for its stock and the qualifying transferors control the corporation immediately after the exchange.
For this purpose, Section 351 incorporates the control definition in IRC Section 368(c), which generally requires the transferor or qualifying transferor group to own at least 80% of the corporation immediately after the exchange.
That requirement helps explain why ETF Section 351 transactions commonly occur in connection with a new ETF launch or seed. A newly formed ETF is a common practical structure rather than an independent statutory requirement. In an established ETF, a new group of contributors generally would have difficulty obtaining the required level of control.
Transfers to investment companies face an additional restriction: Section 351 nonrecognition generally does not apply when the transfer itself results in diversification, subject to detailed regulatory rules in Treasury Regulation Section 1.351-1.
One important regulatory provision treats a contribution as sufficiently diversified when the transferred portfolio satisfies the applicable 25/50 tests. Broadly stated, no more than 25% of the contributed portfolio may be invested in one issuer, and no more than 50% may be invested in five or fewer issuers, subject to the regulation’s detailed rules. In a transaction involving multiple contributors, this diversification analysis generally applies on a transferor-by-transferor basis.
That point changes how investors should think about Section 351. A highly concentrated single-stock position usually cannot simply be placed into a diversified ETF under the fact pattern described here. The investor generally begins with a portfolio that already meets the applicable diversification test.
When all requirements are satisfied, Section 351 can provide nonrecognition at contribution. The embedded gain remains economically relevant because basis generally carries forward. Under IRC Section 358, the investor generally receives ETF shares with basis derived from the contributed securities. Under IRC Section 362, the ETF generally takes carryover basis in the contributed property.
The result is tax deferral. It does not automatically eliminate the capital gain.
Can Section 351 Fit Into A Tax-Aware Long/Short Exit?
Potentially, and the research boundaries matter.
AQR’s 2026 withdrawal study does not model a TALS-to-Section-351 transaction. Its experiments address cash withdrawals, the residual long/short portfolio, and a modeled transition to long-only.
AQR separately discussed Section 351 in a July 2025 educational article about lifecycle liquidation taxes. In that article, AQR identified one possible unwind in which the long and short extensions are removed, leaving an appreciated long-only portfolio that could potentially be managed through direct indexing or contributed to an ETF through Section 351.
That observation is conceptually useful. It is not empirical validation of a combined transaction.
At True Wealth Design, we think the more valuable contribution is to connect these tools at the household-planning level.
A possible lifecycle can be summarized in six plain-English stages:
| Step | Stage | Planning Purpose |
| 1 | Fund | Taxable capital enters the strategy. |
| 2 | Manage Taxes | The manager seeks to defer gains while realizing losses through investment-driven turnover. |
| 3 | Withdraw | Liquidity is taken when the investor’s tax and cash-flow picture supports it. |
| 4 | Deleverage | Long/short exposure is reduced deliberately rather than through an all-at-once liquidation. |
| 5 | Preserve | Appreciated long positions that remain economically desirable are not sold merely for convenience. |
| 6 | Simplify | A long-term destination portfolio is selected, potentially including a qualifying Section 351 ETF exchange. |
If the residual basket, ETF, transaction structure and investor all satisfy the applicable requirements, Section 351 may then provide a path into an ETF while continuing to defer qualifying embedded gains.
The end portfolio still needs to stand on its own investment merits. Tax deferral has value when the resulting investment supports the investor’s objectives, risk tolerance, liquidity needs, costs, and broader plan.
Current Treasury And IRS Scrutiny Deserves Attention
Section 351 ETF seeding is currently an evolving area.
In July 2026, Treasury and IRS officials publicly discussed ETF transactions involving Section 852(b)(6), the rule that generally permits qualifying in-kind redemptions by regulated investment companies without fund-level gain recognition, including structures paired with Section 351 contributions. Officials encouraged industry participants to provide input as they evaluate these structures. K&L Gates summarized the discussion on July 22, 2026.
The discussion itself did not announce a blanket prohibition on qualifying Section 351 ETF formations or change the governing statute and regulations. As of September 2026, those discussions have not resulted in a blanket prohibition or final published guidance governing ordinary qualifying Section 351 ETF seed transactions. The evolving scrutiny makes transaction substance and current tax review increasingly important.
For an investor considering this route, we would want current specialist tax counsel involved before execution. The contributed portfolio should be appropriate for the ETF’s genuine investment strategy, the Section 351 qualification analysis should stand independently, and the investment should make economic sense without relying on an aggressive future redemption strategy.
A Worked Example: A $5 Million Mature 200/100 Portfolio
Assume an investor has a $5 million mature tax-aware 200/100 SMA and wants $1 million of liquidity. The investor also wants to eliminate the long/short structure over time, and the long book contains a broad collection of appreciated securities.
The first step is to determine how much of that $1 million can be raised under the investor’s chosen gain budget. The answer depends on current tax lots, gains and losses realized during the year, existing capital-loss carryforwards, the value of those losses elsewhere in the household, and market conditions.
Any amount that cannot be raised within that gain budget can be funded from another source or raised while knowingly recognizing additional gains.
The next stage is to manage the long/short transition. Short positions and financing can be reduced deliberately while tax lots on the long side are selected with tax consequences in mind.
Once the long/short extensions are eliminated, the investor may be left with an appreciated long-only portfolio. Selling every remaining security merely to obtain a simpler statement could accelerate gains that serve no investment purpose.
At that point several alternatives can be evaluated. The investor may continue holding the securities, place them with a tax-aware long-only manager, make charitable gifts of selected appreciated positions, retain appropriate assets as part of an estate strategy, sell positions where the investment case justifies the tax cost, or evaluate a qualifying Section 351 ETF contribution.
For Section 351, the analysis would include the 80% control requirement, the investor’s contributed basket under the investment-company diversification rules, the ETF’s strategy, basis and holding-period consequences, manager and custodian procedures, and the current Treasury and IRS environment.
The goal is a coordinated transition from a complex long/short portfolio toward a portfolio the investor actually wants to own, while realizing taxes where they are economically justified and continuing to defer gains where doing so remains appropriate.
Frequently Asked Questions
Can I Withdraw Cash From A Tax-Aware Long/Short SMA Without Realizing Capital Gains?
Potentially. AQR’s historical simulations found substantial withdrawal capacity in many scenarios while attempting to use current-year net capital losses against withdrawal-related gains. Actual capacity depends on the portfolio’s tax lots, realized gains and losses, age, leverage, market conditions and the investor’s broader tax situation.
How Much Could A 200/100 Strategy Withdraw In AQR’s Research?
AQR’s average maximum one-time withdrawal from a 200/100 portfolio equaled 56% of initial capital at the end of year two and 46% at the end of year eight. Those amounts were historical simulation averages rather than expected or recommended withdrawal rates.
What Is The Difference Between A Withdrawal And A Complete Exit?
A partial withdrawal raises cash while leaving the residual long/short strategy in place. A transition to long-only requires the short book and financing structure to be removed. A full liquidation goes another step by selling the remaining long positions and moving entirely to cash.
Does A Large Withdrawal Prevent Future Tax Management?
AQR’s one-time-withdrawal simulations found that subsequent annual net capital-loss generation, measured relative to remaining NAV, generally returned toward the no-withdrawal pattern. The smaller portfolio has fewer dollars working, so absolute future loss capacity is reduced.
Can Appreciated Stocks Be Transferred Out Of An SMA Without Selling Them?
A taxable account can often transfer securities in kind while retaining their basis. A long/short SMA adds manager, custody, collateral and margin considerations, so investors should confirm the actual distribution mechanics before depending on this approach.
What Are The Main Requirements For A Section 351 ETF Exchange?
Section 351 generally requires property to be exchanged for corporate stock and the qualifying transferor group to control the corporation immediately after the transaction. Transfers to investment companies also face diversification rules. In common ETF seed transactions, each investor generally contributes a sufficiently diversified portfolio under the applicable 25/50 framework.
Does A Section 351 Exchange Eliminate Capital-Gains Tax?
A qualifying exchange generally provides nonrecognition at contribution. The investor’s basis normally carries into the ETF shares, which preserves the embedded gain for tax purposes. A future taxable disposition can cause that deferred gain to be recognized.
Can Long Positions Remaining After A TALS Exit Be Used In A Section 351 ETF Transaction?
Potentially. The holdings first need to be capable of leaving the SMA operationally. The contributed basket, transferor group, ETF structure and transaction must then satisfy Section 351 and the investment-company rules. The ETF also needs to be an appropriate long-term investment. AQR’s 2026 withdrawal research does not model this combined transaction.
Designing The Full Tax-Aware Investing Lifecycle
The strongest tax-aware investment plan has an entry strategy, a management strategy, a liquidity strategy and an exit strategy.
Withdrawal capacity is only one component. We also need to understand where future gains may arise elsewhere in the investor’s financial life, how valuable existing loss carryforwards are, how leverage should change over time, which appreciated securities remain desirable, whether charitable or estate-planning opportunities exist, and what the investor ultimately wants the portfolio to become.
That integrated perspective reflects how we approach sophisticated planning at True Wealth Design. Our investment, financial-planning and tax work is designed to operate together so that an individual transaction does not inadvertently undermine a larger objective.
If you hold a substantial taxable portfolio and want to understand how tax-aware long/short investing, withdrawal planning, deleveraging or an eventual Section 351 transaction could fit within your broader plan, you can contact a True Wealth Design professional to evaluate the investment and tax implications in context.
This article is for educational purposes only. The strategies referenced apply to Accredited Investors or Qualified Purchasers per SEC regulations.
