After-tax alpha: where tax-aware investing actually earns its keep

Most performance reports lie by omission. They show a gross return, perhaps a fee-adjusted one, and stop there. For a tax-exempt pension this is honest enough. For a taxable family it is a fiction, because the number that actually compounds in the account is the return that survives the tax authority. The gap between the two is rarely small, and over a multi-decade horizon it is the difference between two quite different fortunes.

Research houses have tried to put a figure on the leak. Morningstar's long-running work on the tax cost of US funds finds that investors in large- and small-cap equities have surrendered roughly 1.8% of annual return to tax, and fixed-income investors close to 1.3%, simply through the distributions and turnover embedded in their holdings. A drag of one to two percentage points sounds modest until it is set against an equity risk premium of perhaps four or five. On that arithmetic, tax can quietly consume a third of the reward an investor is taking risk to earn.

The encouraging corollary is that tax is one of the few areas of investing where skill reliably pays. Markets are close to efficient and most managers fail to beat them after costs. The tax line, by contrast, is full of structural inefficiencies that a disciplined process can exploit. The discipline, not any single clever trade, is what produces after-tax alpha.

The first and least glamorous source of tax efficiency is asset location: deciding which assets sit in which wrapper. A family with a taxable account, a tax-deferred pension and perhaps a tax-free vehicle holds three different tax environments, and the same security behaves very differently in each. Tax-inefficient assets, those that throw off income taxed at high ordinary rates, belong wherever that income is sheltered; tax-efficient assets, broad equity exposure that mostly compounds untaxed until sale, can sit happily in the taxable account.

The principle is simple to state and routinely ignored in practice, because most portfolios are built wrapper by wrapper rather than as a single household balance sheet. Done well, asset location adds value without changing the overall allocation by a single percentage point: the risk is identical, only the tax bill moves. It is as close to a free lunch as the discipline offers, which is why it should be settled before anyone reaches for cleverer tools.

Tax-loss harvesting is the technique that has captured the most attention, and it deserves some of it. The mechanics are mundane: sell a position that is underwater, book the loss to offset gains elsewhere, and reinvest in something similar enough to keep the market exposure intact. The harvested loss is a deferral, not a gift, but a deferral compounded over years has real value. Vanguard's research has estimated the additional after-tax return from systematic harvesting at between 0.5% and 1.5% a year, with a more recent study narrowing the range to roughly 0.47% to 1.27% depending on the investor's circumstances and the path of markets.

The honest caveats matter as much as the headline. Harvesting is worth most in volatile, dispersed markets and early in an account's life, when more positions sit below cost. It fades as a portfolio appreciates and unrealised gains crowd out losses. It depends on rules that are deeply jurisdiction-specific: the US wash-sale rule, the treatment of short- versus long-term losses, the availability of offsets against ordinary income. Transplant the technique to a different tax regime and the maths can change entirely, sometimes to nothing. The benefit is real, but it is neither universal nor permanent.

Direct indexing extends harvesting from the fund to the individual share. Rather than owning an index through a single fund, the investor holds the underlying constituents in a separately managed account. That granularity allows losses to be harvested stock by stock even while the index itself rises, lets a family exclude holdings it cannot own for reasons of concentration or conscience, and gives precise control over which lots are sold and when. It is the most complete expression of treating tax as a managed variable.

The flows tell the story of its appeal. Cerulli put direct-indexing assets at around $462 billion in early 2022 and projected the category to top $800 billion by 2026, growing at roughly 12.3% a year, faster than ETFs, mutual funds or conventional separate accounts. The trajectory has if anything run ahead of forecast: assets closed 2024 near $864 billion, a compound annual growth rate above 22% since 2021. Customisation, once a private-bank indulgence, has become an industrialised product.

Two cautions belong alongside the growth chart. The benefit of direct indexing is concentrated in the same high-tax US context that makes harvesting valuable, and it decays as the account matures and gains accumulate. It also introduces operational weight: hundreds of holdings, tracking error against the index, and management fees that must be earned back from the tax saving. For many families a tax-efficient fund delivers most of the benefit with none of the complexity. The technique is powerful where it fits and oversold where it does not.

The cheapest tax strategy of all is to do less. In most regimes that tax capital gains, a longer holding period attracts a lower rate, and a gain never realised is never taxed. Patience converts a liability into a deferral and, eventually, a deferral into a permanent saving. The investor who churns a portfolio pays in two currencies, transaction costs and accelerated tax, while the one who sits still compounds on the whole balance rather than the after-tax remainder.

Discipline here is behavioural before it is technical. It means tolerating a position that has grown uncomfortably large rather than triggering a gain to rebalance, steering instead with new contributions or harvested losses, and measuring success in after-tax terms so the cost of a trade is visible at the moment of the decision. The point is not to avoid selling, but to make sure each sale is worth its tax.

For families with philanthropic intent, the most appreciated assets are often the most useful. Gifting securities that carry a large embedded gain, rather than selling them and donating cash, can extinguish the gain entirely while still delivering the full market value to the cause, and in many regimes generating a deduction besides. The position most painful to sell becomes the most efficient to give. The same logic governs assets destined for the next generation, where a transfer or an inheritance can reset the cost base and erase a lifetime of deferred gain in a single step.

This is where tax efficiency stops being a portfolio question and becomes an estate one. What to sell, what to hold, what to give and what to bequeath is a single problem, and solving it piecemeal leaves value on the table. The most efficient gain is frequently the one the family never realises itself.

It is worth being plain about the limits. Direct indexing and tax-loss harvesting are largely US phenomena, sharpened by a tax code that rewards them; their value varies enormously by jurisdiction and can shrink to a rounding error under a different regime. None of the above is tax advice, and every technique here turns on rules that differ by country and change with each budget. What travels across borders is not the tool but the habit of mind: that the tax line deserves the same deliberate management as the risk line, measured, owned, and revisited rather than left to chance.

A family that internalises this stops asking what its portfolio returned and starts asking what it kept. The difference between those two questions, compounded across a generation, is large enough to be the whole point. After-tax alpha is not exotic; it is mostly the reward for refusing to ignore a cost that everyone else reports around, a return that no market can take away because it was never the market's to give.

Sources: Morningstar (2023); Cerulli Associates (2023, 2025); Vanguard Advisor's Alpha and tax-loss harvesting research (2020, 2024). Figures are industry estimates, are highly jurisdiction-dependent, and will be revised over time. This article is general education, not tax advice.