The Wash Sale Rule Explained

The Sixty-One-Day Window

Section 1091 governs a window of sixty-one calendar days centred on the sale date: thirty days before, the sale date itself, and thirty days after. Any acquisition of a substantially identical security within that window triggers the disallowance, and the rule is symmetric — a purchase that precedes the loss sale by less than thirty days is treated identically to one that follows it. The look-back leg is the one that catches traders most often, because the natural assumption is that a wash sale requires the replacement to come after the loss. It does not.

A worked example fixes the arithmetic. Consider a position of 500 shares of XYZ Corp purchased at $20.00 per share for a total cost of $10,000. The shares decline to $14.00 and the position is closed on March 15 for $7,000, generating a realised loss of $3,000. Absent a controlling fact, that loss would offset other capital gains in the same tax year. If, however, any acquisition of XYZ shares occurs between February 13 and April 14 — the boundaries of the sixty-one-day window — the $3,000 loss is disallowed for current-year purposes.

The thirty-day count is calendar days, not trading days, and the count includes weekends and exchange holidays. The conservative practice is to wait thirty-one full calendar days from the loss-sale date before any reacquisition, which removes ambiguity at the boundary.

What “Substantially Identical” Means

The statute disallows losses where the replacement is “substantially identical” to the security sold, and the IRS has not produced an exhaustive operational definition of the phrase. A body of guidance and case law has developed at the edges, but the core determinations are reasonably settled and the analyst should be able to recognise them.

The clear cases of disallowance include the sale and repurchase of common stock in the same issuer; the sale of common stock followed by the purchase of call options or the writing of put options on the same issuer, both of which establish economic exposure substantially equivalent to ownership; the sale in a taxable account paired with a repurchase in the same taxpayer’s IRA or 401(k), because the rule applies across the taxpayer’s accounts; and a purchase by the taxpayer’s spouse during the window, because the household is treated as a single economic unit for these purposes.

The cases that generally do not trigger disallowance include the sale of a stock and the purchase of a different issuer in the same sector — selling ExxonMobil and buying ChevronTexaco is the textbook illustration; the sale of an individual stock and the purchase of a broad ETF in which the stock is one of many holdings, on the rationale that the basket exposure is materially different; and the sale of preferred stock against the purchase of common stock of the same issuer, where security-class differences in voting, dividend priority, and credit position have been treated as material. Mutual funds follow analogous reasoning where portfolio overlap is partial.

The grey zone is widest with index funds and ETFs that track overlapping benchmarks. Two S&P 500 index funds from different sponsors hold essentially the same constituents in essentially the same weights, and a defensible reading is that the substitution is substantially identical. The IRS has not issued a definitive ruling on this configuration, and the conservative position adopted by most tax preparers is to treat the swap of one S&P 500 fund for another from a different sponsor as exposed to challenge, while a swap from an S&P 500 fund into a total-market fund — whose portfolios diverge meaningfully in mid- and small-cap weight — is treated as safe. The operative point is that the analyst should not assume the IRS view; where guidance is silent, the conservative reading governs.

What Happens to the Disallowed Loss

The wash sale mechanic is, in the typical case, a deferral rather than a forfeiture, and this is the feature most traders fail to recognise the first time they encounter the rule. The disallowed loss is added to the cost basis of the replacement shares, and the holding period of the original shares is tacked on to the holding period of the replacement. The economic loss therefore survives; it is reclassified as a higher cost basis on the new position, recoverable when that position is eventually sold in a transaction that does not itself trigger a fresh wash sale.

Continuing the prior example, the arithmetic is approximately as follows.

The replacement shares now carry a cost basis of $19.00 per share rather than the $13.00 actually paid. If the position is later closed at $22.00, the recognised gain is $3.00 per share on 500 shares, or $1,500, rather than $9.00 per share, or $4,500. The $3,000 disallowed loss has been embedded in the higher basis and is recovered through a smaller future gain. The taxpayer is, on the available arithmetic, made approximately whole — subject to the disclaimer that recovery has been deferred, and that the deferral has a present-value cost equal to the time value of the foregone current-year deduction at the taxpayer’s marginal rate. In a 32% bracket, deferring a $3,000 deduction by twelve months at a discount rate of, say, 5% costs approximately $46. In a multi-year deferral the cost compounds.

Why the Rule Bites Active Traders

A buy-and-hold investor encounters Section 1091 rarely and incidentally. The rule is structurally hostile to day traders and swing traders who recycle the same names through the calendar, because the thirty-day window forecloses precisely the trading pattern they rely on. The cascade can become non-trivial.

Consider a trader who works XYZ stock three times in January. On January 3, 1,000 shares are purchased at $8.50 and sold on January 5 at $9.20, generating a $700 realised gain. On January 10, 1,000 shares are purchased at $9.00 and sold on January 12 at $8.20, an $800 realised loss. On January 18, 1,000 shares are purchased at $8.00 and sold on January 22 at $8.60, a $600 gain. The economic net is $500. The tax accounting, however, treats the January 18 purchase as a wash-sale repurchase against the January 12 loss, because eighteen days separate the two events. The $800 loss is disallowed and added to the basis of the January 18 lot. The reported gains are $1,300 and the deferred loss is $800; assuming the January 22 sale of the adjusted-basis lot is itself not followed by another acquisition within thirty days, the loss recovers in the same tax year and the trader is approximately whole.

The cascade is dangerous when the sequence does not stop. A trader who continues to recycle the same name through February will trigger fresh wash sales on each loss-followed-by-repurchase pair, with each disallowed loss rolling forward into the basis of a position that may itself be closed at a loss within thirty days, generating another disallowance. By year-end, a trader with a net economic loss can have a substantial reported gain on the 1099-B because the unrecovered disallowed losses sit in basis on positions still open as of December 31. The tax bill is real; the corresponding offsets are deferred into the following year. (One sees this most clearly in active client books reconstructed at year-end, not in the broker’s headline 1099 figure, where gross trading P&L and reported taxable P&L can diverge by figures that surprise the principals.) The practical implication is that an active book requires its own reconciling ledger maintained alongside the broker statements, not in place of them.

How to Plan Around the Rule

The legitimate avoidance techniques are well established, and the analyst’s preference among them depends on what is being preserved — the deduction, the economic exposure, or the specific security.

The simplest technique is also the cleanest: realise the loss and wait thirty-one calendar days before re-establishing the position. The cost is the foregone exposure during the window, which is non-trivial if the security rebounds. The benefit is that the rule is unambiguously not triggered.

Where the underlying thesis is sectoral rather than name-specific, substituting a similar but not substantially identical security preserves the economic exposure. A loss in ExxonMobil is harvested by selling and immediately purchasing ChevronTexaco; a loss in a single pharmaceutical name is harvested into a sector ETF. The exposure is broadly maintained, the deduction is captured, and after thirty-one days the original position can be re-established without consequence. This is the standard practice in tax-loss harvesting programmes run by RIAs and wealth managers, and it generally produces the cleanest outcome.

Year-end calendar discipline is the third element. A loss realised in early December and not repurchased until mid-January captures the deduction in the current tax year and clears the wash sale window without leaving the position uncovered for an unreasonable period. The conservative practice is to identify the harvest candidates in early November, schedule the sales for the first week of December, and treat the calendar as a binding constraint rather than a soft target. December crowding produces meaningful price effects in lower-liquidity names, and earlier execution is, in the typical filing, the preferred path.

None of the above survives without recordkeeping discipline. Brokers report wash sales on Form 1099-B with reasonable accuracy within a single account, but the reporting does not capture cross-broker activity, cross-account activity outside their visibility, or activity in a spouse’s accounts at a different institution. The taxpayer’s own ledger is the reconciling document, and it is the analyst’s responsibility to maintain it. A serviceable record requires the trade date, the security identifier, the shares transacted, the price, the holding period, and the running adjusted basis on each lot. Spreadsheets handle this for a small portfolio; specialised tax software handles it for an active book. There is no acceptable shortcut.

Wash Sales Across Account Types

A common misunderstanding in client onboarding is the assumption that a loss harvested in a taxable account can be circumvented by repurchasing the same security in an IRA. The structural appeal of the manoeuvre is obvious — the IRA is, in tax terms, a different vehicle — and the structural defect of the manoeuvre is total. The IRS extended the wash sale rule to cross-account purchases in Revenue Ruling 2008-5, and the disallowance applies whether the replacement account is a Traditional IRA, a Roth IRA, or a 401(k).

The asymmetry that makes this configuration uniquely costly is that the replacement-shares basis adjustment, which in a taxable repurchase preserves the loss for eventual recovery, has no operative meaning inside an IRA. The IRA does not track lot-level cost basis for taxable purposes because distributions from a Traditional IRA are taxed as ordinary income on the gross amount and Roth distributions are tax-free. There is no future event in which the higher embedded basis produces a smaller recognised gain, because no recognised gain ever occurs in the account. The disallowed loss therefore disappears. It is one of the few configurations in which the wash sale rule destroys economic value rather than deferring it, and the analyst should treat the same-security-into-IRA pattern as a hard avoid rather than a planning option.

What the Broker Reports and What It Misses

Brokers are required to report wash sales on Form 1099-B and to apply the corresponding basis adjustments to the affected replacement lots. The year-end statement will identify the disallowed loss, the adjusted cost basis on the replacement shares, and a wash sale code (typically “W”) on the relevant line. Within a single account at a single broker, the reporting is generally reliable.

The reporting fails in three configurations the taxpayer is responsible for catching. Cross-broker activity — a loss sale at Broker A followed by a repurchase at Broker B within thirty days — produces a wash sale that neither broker will identify on its own 1099. Cross-account activity, where a sale in a taxable account is followed by a repurchase in an IRA at the same or a different institution, escapes broker reporting entirely. Cross-spouse activity, with the taxpayer’s account on one side and a separately held spouse account on the other, is similarly invisible to the brokers but visible to the IRS through aggregated 1099 filings. The risk is not that the omission goes undetected indefinitely; it is that detection occurs at audit, several years after the original transaction, by which time reconstructing the basis adjustments is materially more difficult.

The Operative Point

Section 1091 does not prevent a trader from working the same names repeatedly. It prevents a trader from claiming a current-year loss while retaining the economic position that produced it. The rule is, in evidentiary terms, an anti-fungibility doctrine: the same security held continuously in substance is, for tax purposes, the same security regardless of the intermediate paper transactions. The IRS reads through the form to the substance, and the disallowance follows.

For most active accounts, the practical consequence is administrative rather than economic, provided the trader plans the harvest calendar with the thirty-day window in view, maintains a reconciling ledger across all accounts and brokers, and treats the cross-account-into-IRA pattern as a permanent loss of the deduction rather than a deferral. The rule is older than most of the assets it now governs — it predates the modern ETF, the discount broker, and the tax-deferred retirement account — and the analyst should expect periodic IRS guidance refining its application at the edges. The core mechanic, in my working assumption, will not change. The disciplined response is to plan around it.