Tax-Loss Harvesting

Many investors let losing positions linger while rebalancing their portfolios in taxable accounts. That choice creates two clear costs. First, unrealized losses remain useless on tax returns. If you hold a $20,000 position that fell to$14,000, you have a $6,000 unrealized loss. That$6,000 does not reduce taxes until realized. Second, rebalancing by selling winners rather than harvesting losers often realizes gains taxed at short-term rates of 22 to 37 percent or long-term rates of 0 to 20 percent, depending on income and holding period. In a concrete example, imagine a taxable account with $100,000 total, target allocation 60/40, and drift to 70/30. Selling$10,000 in winners could generate a $10,000 long-term gain. IF the investor is in the 24 percent marginal tax bracket AND qualifies for a 15 percent long-term rate, THEN the immediate tax bill could be$1,500 BECAUSE $10,000 × 15$1,500. That $1,500 eats into the capital available for rebalancing. Meanwhile IF the investor instead harvested a$10,000 loss, THEN they could offset that $1,500 gain BECAUSE losses offset gains dollar-for-dollar, and any remaining net loss can offset up to$3,000 of ordinary income per year with carryforward of the remainder. A practical numeric trade-off emerges: realize losses first to reduce current tax drag, then rebalance using tax-advantaged accounts or new cash to limit future taxable events. In Rebalancing we covered using direct new contributions and accounting for threshold rebalancing to reduce realized gains; not doing that here amplifies tax friction.

Start with the mechanics and the math. Tax-Loss Harvesting means selling a security for less than its cost basis to realize a capital loss. Capital losses first offset capital gains in the same tax year. If total capital losses exceed capital gains, then up to $3,000 ($1,500 if married filing separately) may be used to reduce ordinary income in that year. Any remaining excess loss carries forward indefinitely until fully used. Use the algebra: let $L$ = realized loss, $G$ = realized gains, $t_g$ = effective capital gains tax rate (0% to 20% typical for federal), and $t_o$ = ordinary marginal tax rate (10% to 37% typical). Net taxable gain = max($G - L$, 0). Tax change from harvesting equals $-$min($L, G) ×$t_g$plus potential$-$($ ext{max}(L-G,0) ext{limited to } 3{,}000) × $t_o$ in the current year. For example, if $G = 5{,}000$, $L = 8{,}000$, t_g = 15 ext{5}, and t_o = 24 ext{5}: the immediate tax reduction = 5{,}000 × 15 ext{5} + 3{,}000 × 24 ext{5} = 750 + $720 =$1{,}470. Remaining carried loss = $8{,}000 -$5{,}000 - $3{,}000 =$0. If $L$ were $12{,}000$, then carried loss = $4{,}000$ to be used in future years. The main additional rule to incorporate is the Wash Sale Rule. The IRS wash sale prevents claiming a loss if, within 30 calendar days before or after the sale, the taxpayer buys a substantially identical security. That 30-day window means the practical replacement strategy is: IF the investor wants to remain exposed to the same market risk AND plans a replacement within 30 days, THEN use a non-substantially-identical substitute BECAUSE buying the same fund within that 61-day window (30 before, day of sale, 30 after) would disallow the loss. Quantify choices: replacing a sold US large-cap ETF with a different large-cap ETF that tracks a different index avoids the rule in most cases. Remember holding period rules for converting short-term to long-term gains can affect the post-harvest long-term status, changing $t_g$ over time. Also account for transaction costs and bid-ask spreads typically in the range of 0.00 to 0.50 ext{5} of trade value for large ETFs, which can reduce net benefit.

Start with a problem-first question: which action reduces taxes net of costs today and over a 1 to 5 year horizon? Frame the decision in IF/THEN/BECAUSE steps. Step 1 - Determine net taxable position. IF expected realized gains this year are greater than $1{,}000 AND less than$100{,}000, THEN harvesting losses up to those gains may reduce taxes dollar-for-dollar BECAUSE capital losses offset gains at a 1:1 ratio. Use exact numbers: if gains = $20{,}000, harvesting$20{,}000 of losses saves roughly $20{,}000 ×$t_g$in federal taxes today, where$t_g equals 0 ext{5}, 15 ext{5}, or 20 ext{5} depending on income. Step 2 - Compare marginal benefit to transaction costs and fees. IF the expected tax savings is greater than trading costs plus behavioral or tracking risk by at least 20 to 50 ext{5}, THEN proceed, because small percentage advantages evaporate when fees are 0.25 ext{5} to 0.75 ext{5} and spreads are 0.1 ext{5} to 0.5 ext{5}. Numerical example: for L = $10{,}000 and$t_g = 15 ext{5}, tax savings = $1{,}500. If trading cost and tracking slippage total$200 to $400, net benefit remains about$1{,}100 to $1{,}300. Step 3 - Consider timing and wash sale constraints. IF you plan to repurchase similar exposure within 30 days, THEN replace with a different ETF or wait 31 days BECAUSE the **Wash Sale Rule** disallows the loss if you repurchase a substantially identical holding within that period. Step 4 - Decide on automation vs manual harvesting. IF account balance in taxable accounts is under$50{,}000 AND average realized loss events are infrequent, THEN manual harvesting may be cheaper BECAUSE robo or advisor automated harvesting generally charges 0.25 ext{5} to 0.50 ext{5} of AUM or requires an advisory relationship. IF balances exceed $100{,}000 AND trading frequency is high, THEN automated harvesting often uncovers 10 to 30 extra harvestable opportunities per year BECAUSE continuous monitoring and algorithmic identification reduces missed losses and human attention limits. Step 5 - Project carryforward value. IF your marginal tax rate is stable, THEN carrying a$20{,}000 loss forward yields future annual ordinary-income offsets of up to $3,000 per year, which absent other gains represents about$720 per year at 24 ext{5} in ordinary tax reduction, until exhausted.

What this framework does not capture matters. Limitation 1 - tax-bracket timing risk. IF an investor expects to move from a 24 ext{5} bracket to a 12 ext{5} bracket within 1 to 3 years, THEN harvesting now may be suboptimal BECAUSE losses used at a lower future rate produce smaller tax benefits. Quantify: a $10{,}000 loss is worth$2{,}400 at 24 ext{5} but 1{,}200 at 12 ext{5}. Limitation 2 - state and local taxes. This model ignores state tax variation. Some states tax capital gains at 0 ext{5} to 13.3 ext{5}. IF your state rate is 5 ext{5} to 10 ext{5}, THEN harvesting yields extra state tax savings BECAUSE losses offset state gains similarly. Limitation 3 - wash sale complexity with tax-advantaged accounts and third-party transfers. IF an investor sells a loss in a taxable account and buys the same security in an IRA within 30 days, THEN the loss may be permanently disallowed BECAUSE wash sale rules include purchases in IRAs and other accounts for the taxpayer; the disallowed loss is added to the IRA cost basis which often cannot be tracked easily. Limitation 4 - behavioral and market-timing risk. This framework does not model the opportunity cost of being out of a security for 31 days, which historically can change returns by several percentage points in short-term windows. For instance, missing a 3 ext{5} market pop over 30 days on a 50{,}000 position equals $1{,}500 of foregone gain. Limitation 5 - small accounts and tax loss magnitude. IF total expected harvestable losses are under$1{,}000 per year, THEN the gross tax benefit often falls below trading costs BECAUSE tax savings at 15 ext{5} would be under $150 while trading costs might be$0 to $20 per trade plus advisory fees. Finally, this framework does not model multi-generational estate planning where unrealized losses vanish at death due to step-up in basis; in those cases harvesting can sometimes be counterproductive relative to holding until death.

In Rebalancing (/money/rebalancing) we covered portfolio drift and methods like direct new contributions and threshold rebalancing to limit realized gains. That context matters because rebalancing choices change how often Tax-Loss Harvesting is available and which positions produce losses. In Capital Gains (/money/capital-gains) we covered short-term versus long-term rates and netting rules; those rules provide the tax rates $t_g$ and $t_o$ used throughout this lesson. Mastering Tax-Loss Harvesting unlocks more advanced topics like tax-aware asset location (/money/asset-location), Roth conversion timing (/money/roth-conversions), and estate basis planning (/money/estate-basis) because the timing and realization of losses interact directly with income-tax windows, conversion costs, and the step-up in basis available at death.