Average Cost vs FIFO: Which Tax Method Should You Use?
Here's something that surprises a lot of investors the first time they see it: the exact same sale — same stock, same shares, same price — can produce two genuinely different taxable gains, depending entirely on which accounting method you use to figure out your cost basis. Neither number is wrong. They're just answering the question differently.
The problem: you bought the same stock more than once, at different prices
If you buy a stock in one single purchase and sell it in one single sale, there's no ambiguity — your gain is just proceeds minus what you paid. The complexity shows up the moment you buy the same stock more than once, at different prices, and then sell only part of your position. Which shares did you actually sell — the first ones you bought? A blend of all of them?
Average Cost: blend everything together
Average Cost takes every share you've ever bought of that stock, averages the price across all of them, and uses that single blended number as your cost basis for any sale — regardless of which specific purchase the shares "came from." It's simple, and it's the default most basic tracking tools use because it doesn't require keeping track of individual purchase lots.
FIFO: the oldest shares go first
FIFO — First In, First Out — assumes you sell your oldest shares first, tracked lot by lot. If you bought 10 shares in 2024 and 10 more in 2026, and then sell 15 shares, FIFO says you sold all 10 from 2024 plus 5 from 2026 — not a blend of both.
A real example where it matters
Say you bought 10 shares at $100 back in 2024, then bought 10 more at $150 in early 2026. You sell 15 shares in mid-2026 at $200.
| Method | Cost Basis Used | Taxable Gain | Holding Period |
|---|---|---|---|
| Average Cost | $125/share (blended) | $1,125 | Long-term (one blended sale) |
| FIFO | $100 for 10 sh, $150 for 5 sh | $1,250 total | Split: $1,000 long-term + $250 short-term |
Notice two things: the total gain is different between methods ($1,125 vs $1,250), and FIFO correctly recognizes that part of this sale was short-term while Average Cost lumps the whole thing into one holding-period bucket. Neither is "cheating" — they're legitimately different accounting conventions, and the gap between them can be meaningful once you're selling partial positions built up over several purchases.
Why this is easy to get wrong without proper lot tracking
Average Cost is simple precisely because it throws away information — it doesn't need to know which shares you're selling, just the blended average. FIFO requires actually tracking each purchase as a separate lot, with its own date and price, and correctly consuming the oldest lot first, splitting a single sale into multiple pieces if it spans lots with different holding periods. Doing this by hand across a few years of trading is exactly the kind of bookkeeping that gets error-prone fast.