I Sold My Best Stock at $94 Because It Got Too Big

The limit order filled at exactly $94.10 on October 3, 2023, and I remember staring at the confirmation screen in a Phoenix Marriott elevator bank around lunch, briefcase in one hand, thinking: congratulations, you just sold your retirement house to buy balance.

Bulgarian grandmother context matters here more than the ticker does. My grandmother Mira, strict about debts and suspicious of paper wealth, mailed that check with instructions in shaky cursive: buy somethin intelligent, then leave it alone. Following half her instruction made money. Following both, it turns out, required adulthood.

This was my AMD position. Snagged originally at $12.84 in July 2016 with $14,000 of wedding-gift money my Bulgarian grandmother had earmarked for "something intelligent," grown thru five years of Lisa Su's relentless execution into approximately $110,000 by that autumn afternoon. Somewhere between those prices, my clever side bet quietly became my largest holding, then my scariest one, then an obsession wearing a ticker.

Concentration math reads beautifully upward and terribly downward. With AMD at 38% of investable assets, a single earnings miss, a China export restriction headline, a product-cycle stumble: any of them could erase more wealth overnight than eleven years of careful saving across everythin else combined. My wife Daniela ran our household budget on spreadsheets; nowhere did variance that size fit a column.

trimming felt like betrayal

Selling winners carries odd grief. Every financial psychology paper nods at this politely, but living it means watching daily quotes of something you owned with pride while knowing proceeds earn 4% Treasury yield instead of fame.

My trim plan executed in thirds across three months per self-imposed rule: sell down til no single name exceeds 15% of portfolio, proceeds redeployed mechanically into plain-broad-market ETF choices, VTI plus SCHD, zero discretion permitted after market hours. Final third hit during December 2023 strength near $98.

Then came 2024. AMD reached almost $227 in March before settling back thru the year, today trading around $165, meaning my full exited stake would currently run about $180,000 versus the $107,000 grasped. Paperwork difference of $73,000. Friends who watched me exit early still bring up Nvidia comparisons at barbecues like consolation prizes.

seventy-three thousand of tuition for insurance nobody sells

That $73K is the tuition bill for insurance nobody sells: guaranteed portfolio survival purchased against single-company ruin I couldn't underwrite myself, regardless of how much I admired the CEO or understood chip architectures as a former datacenter technician.

why concentration math breaks people

hold, hedge, or trim

Weighed alternatives rationally for weeks beforehand: hold fully (risk catastrophic), hedge via options (complex, expensive at that scale), trim partially (chosen). What surprised me most involved emotional mechanics rather than analytics — ownership itself breeds advocacy. Every conference call became fandom, every short-seller piece personal insult, every dip an affront requiring immediate defense via additional purchase apparently.

Objectivity dies incrementally inside winning positions, which explains veteran investors' boring insistence on written rules decided calmly. Mine live taped inside my desk drawer since 2023: maximum 15% any single stock, quarterly review dates only, trades execute per calendar never per headlines, grandmother's gift money stays permanently diversified now cuz she deserved intelligence plus safety equally.

bringing daniela into the position review

We ran a household stress test on paper first: portfolio down 50% concentrated versus down 20% diversified, next to our actual monthly obligations printed from her budget workbook. Seeing both scenarios beside the number in column F ended the debate before I finished making my case for keeping everything.

Daniela's involvement changed outcomes too. Presenting the concentration situation as household risk rather than individual investing choice transformed trim from surrender into protection. Our mortgage refinancing conversations abruptly included honest discussion about whether one company's revenue cycles should dictate family solvency margins.

rebuilding dignity after selling winners

deploying without ceremony

Proceeds went to work immediately per plan without lingering in cash temptation. VTI slices deployed weekly over six months, SCHD chunks similarly spread, international allocation filled last at whatev valuations January 2024 offered. Everythin automated thereafter through Schwab transfers scheduling themselves monthly whether markets celebrated or sulked.

Portfolio behavior changed visibly post-trim. Drawdowns during spring 2024's tech correction dented statements mayb 8% versus previously projected 14% for equivalent market moves. April 2024's earnings dip passed without a single browser refresh, which my old self would have weighed impossible. Sleep improved measurably. Arguments about money reduced frequency dramatically within two quarters, correlation worth confessing publicly here since money arguments rarely announce their causes honestly upfront.

Barbecue commentators keep score anyway, and accounting for their applause costs nothin since my drawdown tolerance stopped being their decision the day that first trim order filled.

Would partial retention have been optimal historically? Evidently, looking backward through charts. Position sizing exists precisely because nobody occupies hindsight in real time when decisions carry weight.

Elevator-bank version retold whenever somebody asks about hefty winners: keeping every winner til cosmic disappointment isn't loyalty, it's leverage disguised as conviction, sized correctly conviction compounds safely whereas misplaced sizing destroys compounding machines entirely. Grandmother's $14,000 multiplied generously thru one great company already; protecting multiplication required becoming adult enough to sell applause-worthy stories.