ClearViewLesson libraryWhat's new

Learn · Personal Finance · Wrappers, Tax and Risk

Taxes Across the Year

30 min read

Tax in a taxable account is triggered by selling, not by holding, so the household controls its own bill. Losses offset gains dollar for dollar, and a loss harvested with no gain to offset only carries forward — which is why harvesting belongs in the same decision as selling, and why the wash-sale rule can quietly take it all back.

Tax is triggered by selling, not by holding

The defining feature of a taxable account is that the government takes its share when you realise a gain, not while the position grows. An unrealised gain is not a liability you owe today; it is a claim that comes due when you choose to sell, which makes the account a set of decisions rather than a passive container. The first decision is simply whether to sell at all: a position you would otherwise hold for a decade should not be sold to tidy up. The second is the holding period, which changes the rate. Long-term gains — realised more than a year after purchase — are taxed at lower rates than short-term gains, which are taxed as ordinary income. That difference alone is a reason to treat the first twelve months of any position as a waiting period rather than an opportunity. The third is the most powerful and the most ignored: the step-up in basis at death, which resets the cost of an inherited asset to its value on the date of death. For a holding the household intends to leave to heirs, that can eliminate the entire accumulated gain — which is why “never sell the lowest-basis position” is not sentimentality but a tax strategy, and why a charitable gift of appreciated stock beats a cash gift of the same size. Three realisations, three treatments — Held under a year: short-term gain, taxed as ordinary income · Held over a year: long-term gain, taxed at a lower rate · Never sold, then inherited: basis resets to the value at death ← These are U.S. federal rules in outline. Rates, thresholds and the treatment of specific assets change, so the mechanisms are the durable part — the numbers belong in a current reference.

Harvesting a loss, and the rule that undoes it

Losses offset gains dollar for dollar, which makes a realised loss an asset in a year that has gains to absorb. If the losses exceed the gains, up to a fixed amount can offset ordinary income and the rest carries forward to future years — so a loss is rarely wasted, but it is also rarely useful in a year with nothing to offset. That asymmetry is why harvesting belongs in the same decision as selling rather than being a habit of its own. The mechanic that trips people up is the **wash-sale rule**: if you sell at a loss and buy the same or a substantially identical security within thirty days either side of the sale, the loss is disallowed. The intent is obvious — you cannot book a tax loss while keeping the same economic position — and the practical answer is to pair the sale with something genuinely different, or to wait out the window. Buying “similar but not identical” is the grey area; treating a sector fund as a substitute for another sector fund in the same industry is closer to a wash sale than to a harvest. Two disciplines keep this honest. First, do not let the tax tail wag the investment dog: selling a position you want to keep, or delaying a sale you need, in order to save a percentage of the gain, is a bad trade. Second, keep the mechanics cheap: a harvest is a sale and a purchase, so it costs two spreads, and in a small account the tax saved can be smaller than the friction. The harvest, priced — Unrealised gain on the winner: 500 × ($180 − $120) = $30,000 · Unrealised loss on the loser: 300 × ($210 − $180) = $9,000 · Tax selling both, at 15%: ($30,000 − $9,000) × 15% = $3,150 · Tax selling the winner alone: $30,000 × 15% = $4,500 ← The wash-sale window runs thirty days before and after the sale, and it applies across all your accounts — including an IRA. Rebuying in a retirement account does not preserve the deduction; it just destroys the loss.

Asset location: which account holds what

Once a household has more than one kind of account, the question stops being what to own and becomes where to own it — and the answer is decided by how much annual tax each asset generates. A bond fund pays its yield out every year and it is taxed as ordinary income, so it is the most expensive asset to hold in a taxable account. A broad equity index fund yields little and rarely sells anything, so it is the cheapest. A REIT fund pays out a large share of its income and belongs in a shelter. The account you lose the least by filling with a high-yielding asset is the sheltered one, and that is the whole rule. The Roth is the last sleeve to fill and the one to fill with the highest-expected-return asset, because its growth is never taxed and the longer a high-return asset compounds untouched, the more the wrapper is worth. The traditional account is next, holding bonds and other high-yield assets. The taxable account takes the equity index fund — which also gives the household the flexibility to sell something without touching a wrapper, and the step-up in basis if it is never sold at all. One caution keeps the rule from becoming a straitjacket. Asset location should follow the household’s actual accounts, not an idealised grid: if the workplace plan has one good fund and nothing else, that is what goes in it, and a slightly suboptimal placement beats a worse fund. The rule is a tiebreaker between two reasonable options, not a reason to buy a product you would not otherwise hold. A placement grid — Traditional account: bond funds, REIT funds, high-yield assets · Roth account: the highest-expected-return equity position · Taxable account: broad equity index funds, held long and rarely sold ← The grid is a preference ordering, not a law. Accounts differ, workplace menus are short, and a good cheap fund in the wrong account beats an expensive one in the right account (PF6).

The gain side: lots, brackets and the step-up

Losses get the attention because they can be harvested, and the gain side is where the larger decisions are made. Three of them are worth knowing: which lot you sell, which bracket you realise the gain in, and what happens to unrealised gains at the end of a life. The **lot** decision is the one a broker makes by default unless told otherwise. Funds and accounts typically track a purchase history, and the default convention is usually first-in-first-out, which realises the oldest and often the lowest-cost shares first. Choosing **specific identification** lets you select which lots to sell, so a sale can realise a long-held low-basis lot or a recently purchased high-basis one. That choice is not a tax loophole; it is a decision about how much gain to recognise this year, and it is only available if the instruction is given at the time of the sale rather than reconstructed later. The **holding period** then decides the rate. A gain on an asset held beyond a year is taxed at the long-term rate, which in the lower income brackets can be zero — the threshold at which the long-term rate begins is a planning opportunity in its own right, because a household with a low-income year can realise a substantial gain at no federal tax cost. Crossing into the higher brackets adds a surtax on net investment income above a threshold, which raises the effective rate on the marginal gain above the nominal one. The practical point is that the marginal rate on the next dollar of gain is usually higher than the household expects, which is why large one-off realisations are worth spreading when there is any option to. The third decision operates after death, and it is the strongest argument for holding rather than realising in some situations: assets held until death generally receive a **step-up in basis** to their value at that date. A position with a large unrealised gain that passes to an heir can thereby avoid the capital gain entirely. That is a planning tool rather than a strategy to aim at, and it sits in tension with the annual exemption and with loss harvesting — one pushes toward holding, the other toward realising. Where the household also intends to give, the mirror version applies: donating appreciated shares to a charity avoids the gain and produces the deduction at full value, which is strictly better than selling and donating the proceeds. The honest summary is that tax should not drive investment decisions, and it is a real cost on the decisions already made. The three levers above are all about *when* and *which* rather than *whether*: which lot, which year, and whether the asset is held until the basis resets. None of them changes what is worth owning; all of them change what owning it costs. • The default lot convention is not the only one — specific identification is the alternative. • A low-income year can make a long-term gain unusually cheap to realise. • The surtax on net investment income makes the marginal rate higher than the nominal one. • A step-up at death and donating appreciated shares are the two gain-side structures worth knowing. A practical discipline: check the cost basis the broker has recorded at least once a year rather than at the time of a sale. Basis errors are common after transfers between firms, and they are difficult to correct once a return has been filed.

What you'll practise

A $30,000 unrealised gain and a $9,000 unrealised loss, taxed at 15%. What does realising both save?

40 XP in the app · multi select

Sources

Practise this in the app →

Learn content is for education only — not individualized financial advice, a recommendation, or a solicitation to buy or sell any security. Options involve substantial risk. Examples are simplified and historical patterns never guarantee future results.