Tax-Loss Harvesting Isn’t a December Panic. It’s an October Rule.
Picture December 27.
Your CPA just forwarded the gain report. It's bigger than you thought. RSUs vested in a strong market. You sold a slice of company stock in an open window last spring. A private fund K-1 showed up with a capital-gain allocation you didn't see coming.
So you scramble. You sell the losers. You buy something that "looks the same" so you stay invested. You tell yourself you harvested.
Then January hits.
Your broker flags a wash sale. Or your spouse's account bought the same fund on autopilot. Or the "similar" ETF was treated as substantially identical. Or — the quiet killer — you bought the same security back inside an IRA, and the loss is gone for good. No basis add-back. Just a permanent forfeit.
I watch some version of this every December. Smart people. Real tax bills. A harvest that fails the one rule that matters.
Tax-loss harvesting isn't a December panic. It's an October rule.
What tax-loss harvesting actually is
Strip the jargon.
You sell an investment for less than you paid. That capital loss offsets capital gains dollar-for-dollar. If losses exceed gains, up to $3,000 can offset ordinary income ($1,500 if married filing separately). The rest carries forward indefinitely until you use it.
That's the whole mechanism. Not a trading strategy. Not a market call. A tax calendar tool.
Done early and clean, it lowers the tax on gains you already took — or will take — this year. Done late and sloppy, you sell the loss and still don't get to use it. Or you get to use it later when you didn't need it. Or you permanently lose it.
The difference between those outcomes is almost never "good stock picking." It's whether you respected the wash-sale window — and whether you designed the replacement before you sold.
Why December fails for executives
December isn't hard because the math is hard. December is hard because your life is already full.
Trading windows for company stock often close ahead of earnings. Autopilot dividend reinvestment is still running. Spouse and trust accounts hold overlapping funds. IRAs and Roths sit next to taxable brokerage. Year-end bonuses, RSU vests, and capital calls are stacking. Your attention is gone — kids home, travel, closing the year at work.
So the harvest becomes a Friday-afternoon order ticket: sell the losers, buy something close, hope it clears.
Hope is not a wash-sale plan.
The wash-sale rule is simple on paper and ugly in practice. If you buy the same or a substantially identical security within 30 days before or after the loss sale — a 61-day window — the loss is disallowed. That includes purchases in your spouse's accounts. It includes certain options and contracts. And if the repurchase happens inside an IRA or Roth, Revenue Ruling 2008-5 is clear: the loss can be permanently forfeited. No basis add-back. Gone.
In a taxable account, a disallowed wash loss usually adds to the basis of the replacement. Deferred, not destroyed. In an IRA repurchase path, deferred is not on the menu.
December leaves you with 61 days that don't fit on the calendar. October does.
Wash-sale traps high earners actually hit
These aren't theoretical. They're the patterns I see when a "harvest" gets unwound in February.
Company stock overlap. You sell a loss lot in a taxable account. Inside the same 61-day window, an ESPP purchase settles, RSUs vest, or you exercise and hold. Same issuer. Same security. Wash sale.
"Same" ETF siblings. You sell one broad-market fund and buy another that tracks the same index with a different ticker. Similar exposure can be the right investment idea. Substantially identical is a tax question — and it's not something you want to settle with a Friday Google search.
Taxable + IRA. You harvest a loss in brokerage, then your automatic IRA contribution or a rebalance inside the IRA buys the same (or substantially identical) fund. The IRA path can kill the loss permanently.
Spouse and household accounts. The wash-sale rules look across related accounts. Separate login credentials do not create separate tax realities.
Auto-DRIP and auto-invest. Dividend reinvestment and recurring buys keep purchasing after you sell. If those buys land inside the window, they can wash the loss you just booked. Turn them off before you sell — not after.
None of this means you shouldn't harvest. It means you shouldn't invent the replacement on December 27.
Replacement design: similar is not identical
The point of tax-loss harvesting is usually to realize a loss without abandoning the exposure you still want.
That requires a replacement that is economically similar enough for your plan — and not substantially identical for wash-sale purposes.
That is an advisor-and-CPA conversation, not a DIY ticker swap.
What "similar but not identical" means depends on the holding, the index, the share class, options overlays, and how your household accounts are set up. Getting it wrong costs you the loss. Getting it late costs you the window.
October is when you still have time to:
Inventory every taxable lot with an unrealized loss (and the ones with gains you might realize this year).
Map realized gains year-to-date — equity sales, fund distributions you can estimate, K-1s if you have them early.
Design replacements before you sell — with names that fit the plan and clear the substantially-identical test for your facts.
Pause DRIP / auto-invest on anything in the harvest set.
Put the 61-day window on a calendar that includes spouse accounts, IRAs, ESPP settlement dates, and RSU vest dates.
Rules over feelings. The calendar is the rule.
The October checklist
Dated. Concrete. Run it with your CPA and advisor — not as personalized advice from a blog post, but as a shared agenda.
This week (week of October 12)
Pull a lot-level report for every taxable account: cost basis, unrealized gain/loss, acquisition dates.
List realized gains year-to-date. Rough is fine; a range beats a blank cell.
Flag household accounts that could create a wash: spouse brokerage, trusts, IRAs/Roths, managed accounts with auto-rebalance.
Pause DRIP and recurring buys on candidates you might harvest.
By October 31
Agree on which losses are worth harvesting this year (size vs. gains, carryforward usefulness, transaction costs).
Lock replacement designs — written, not verbal. What you sell. What you buy. Why it is not substantially identical for your situation.
Note every purchase that could land in the 61-day window: ESPP buys, RSU vests, scheduled IRA contributions, trust rebalances.
November
Execute harvests with runway. Prefer earlier over later so the window clears before year-end chaos.
Confirm settlements. Confirm DRIP stays off. Confirm no "helpful" rebalance ran overnight.
Update the gain/loss worksheet after each trade.
Early December
Final check: any remaining losses worth taking? Any new gains from November activity?
Confirm the 61-day clocks from November trades — what still cannot be repurchased, and where.
Stop ~December 20
New harvests this late leave almost no clean runway before year-end and the January autopilot season.
If you didn't start in October, December is for damage control with your advisors — not freestyle order tickets.
Boring is the goal. A boring harvest is one that still works in February.
The decision rule (three questions)
Before you sell a loser for tax reasons, answer these out loud:
What gain am I offsetting — this year — and do I actually have it? Harvesting into a thin gain year just builds a carryforward you may not need soon.
Where could a repurchase happen in the next 30 days — including accounts I don't trade every day? Spouse, IRA, ESPP, RSU, DRIP, advisor rebalance.
Is the replacement designed, written down, and cleared as not substantially identical — or am I improvising?
If you can't answer all three, you don't have a harvest. You have a hope.
Not "no." Just "not without these answers."
Want a second set of eyes?
If your gain year is already loud — company stock sales, equity compensation, K-1s — a 30-minute conversation in October beats a wash-sale letter in February.
Happy to be a second set of eyes on the checklist. Educational conversation, not a product pitch.
Disclaimer
This article is for educational purposes only and is not investment, tax, or legal advice. Forecast Capital Management is a registered investment adviser. Your facts matter — wash-sale treatment, replacement design, and tax outcomes depend on your accounts, filing status, and holdings. Consult your own CPA, attorney, and investment adviser before taking action. Past performance is not indicative of future results. No specific securities are recommended here.
