Three Holdings Was All It Took to Cut My Tax Bill by $2,340

At 8:55 on the last Sunday of December, with my street quiet and a space heater running under the desk, I closed my portfolio app, opened a blank note, and wrote three numbers: one down 41 percent, one down 9 percent, one up 63 percent. That was the entire empire. Three holdings, total value under $14,000, and a decision about whether loss harvesting was a rich person's game or something I could actually run at my scale. The standard advice assumed index funds and six figures. I had neither. The stake was a number I could name. My accountant's early projection had me owing $2,340 more than I wanted to part with in April, and the down holding, snagged in stages thru the spring, carried enough grasped loss potential to change that if I acted before midnight on the 31st.

the rule I nearly got backwards

The trap I studied for an hour: sell the loser, book the loss, and buy it back immediately, and a wash-style rule kicks the loss down the road. Stocks have a bright-line 30-day rule; the treatment of my crypto purchases was murkier, and murkier is not the same as safer. My accountant's instruction was blunt. Do not buy the same asset inside the window at all. Wait it out or rotate into something truly different, and let the paperwork stay clean. Clean cost me thirty days of exposure and snagged me a deduction I did not have to defend.

what I actually sold

I sold the 41 percent loser in full at 9:31 AM Monday, morning volume, thin spreads, a fill within $6 of the price on my screen. The loss on paper was $1,780. The 9 percent holding I trimmed by half, booking another $310 of loss while keeping a position I still believed in. The 63 percent winner I did not touch, which was the whole emotional test. Harvesting losses is not an apology for the winners. It is housekeeping around em.

the thirty days of watching the door

January was longer than usual. The sold asset chopped sideways for two weeks, then jumped 11 percent in the last week of the window, and I watched that jump from outside like a person watching a party through glass. That 11 percent would have been worth about $470 on my old position. Against a $2,340 reduction in tax owed, I still came out ahead by nearly five to one, and knowing both numbers is the only way this math stays honest. Anyone quoting the upside of staying in and forgetting the deduction side is doing theatre.

the mistake I almost made first

My opening plan, sketched in that same December note before I talked to my accountant, was to sell the loser and rebuy it the same afternoon, cuz the price looked like a floor and I did not wanna miss the bounce. That single instinct would have tainted the entire loss and wasted the only clean number in my portfolio. The bounce I feared missing never came in January anyways, and even if it had, chasing it with a disqualified loss is paying a deduction for the privilege of round-tripping a position. Writing the plan down before sending it to anyone was what exposed how much of it was superstition.

where the proceeds sat instead

For the thirty-day window the proceeds went into a high-yield savings account earning about 4 percent, which generated a laughable $14 of interest and total peace of mind. The point was not the yield. The point was that the money existed, was not rotting in the same asset to spite the rule, and was one transfer away when February opened. I rebought the position on the 31st at a price 6 percent below my December sale, which nobody can plan for and I will not pretend I did.

the paperwork pile that made it real

The transaction itself took eleven minutes. The documentation took an evening. Screenshot of the sale fills, screenshot of the rebuy in February, the savings statement showing the proceeds parked for exactly the window, and a note in my own words explaining the sequence, dated, so the file reads as a plan rather than an accident. My accountant later told me the note was the difference between a five-minute review and a call back. Loss harvesting at any portfolio size is mostly an evidence project, and the people who skip the evidence are the ones who end up unable to defend a legitimate number in April.

what my accountant said in April

The loss carried against my gains exactly as designed, the filing took twenty minutes less than the prior year, and the effective saving landed within $85 of projection, which is about as close as reality gets to a spreadsheet. She now uses my file as her example of a modest portfolio doing this correctly, which means she uses it as her example of a client who waited the window. That is the entire trick. The strategy is not cleverness. It is calendar compliance.

the Sunday in December, closed

The note with three numbers is still pinned in that app, and I reread it every December now like a checkup form. The down-41 holding, rebought in February, is up 22 percent as of this month, and I feel no superstition about it, just the quiet satisfaction of having sold a loss on purpose and snagged back a position on schedule. The space heater is still under the desk. The only thing that changed is that the last Sunday of the year belongs to the calendar now.