An investor who buys shares more than once owns them at more than one price, and the average cost per share is the single figure that summarises the position: the total amount spent, divided by the total number of shares held. It determines the break-even price, it frames every decision about buying more or selling, and it is the figure most brokerage screens display. This guide explains how the average is computed, how it moves when shares are bought at different prices, what averaging down and averaging up really mean, how corporate actions and reinvested dividends change the figure, and why the average must never replace the lot-level record that tax returns require. No live prices appear here, and nothing in this guide is investment advice: the examples use round, illustrative prices.

What average cost means

Average cost per share is the total amount invested divided by the total shares held. Buy 100 shares at $50: cost $5,000, average $50. Buy 100 more at $40: cost $4,000 more, total $9,000, shares 200, average $45. The average is the break-even price: sell all 200 shares above $45 and the position profits; sell below $45 and it loses, before fees and taxes. It is also the reference point for every subsequent decision — a price above the average feels like a gain, a price below it feels like a loss — which is why the figure’s psychological weight exceeds its informational content, and why it should be checked against the position’s actual prospects rather than treated as a decision rule.

The weighting is by shares, and that matters. Buy 100 shares at $50 and 10 shares at $40: total cost $5,400, shares 110, average $49.09 — barely moved by the cheaper purchase, because only 10 shares were bought at the lower price. Buy 10 shares at $50 and 100 at $40: total $4,500, shares 110, average $40.91. The same two prices, opposite weightings, produce averages $8 apart. The lesson is that averaging is a function of quantities as much as prices: a small purchase at a bargain price barely moves the average, and a large purchase near the top barely lowers it either.

Averaging down: the maths, and the risk

Averaging down is buying more as the price falls. The maths is attractive: each purchase below the current average pulls the average down, and the break-even price falls with it. A position of 100 shares at $50 that falls to $40 can be averaged to $45 with 100 more shares; if it falls to $30, 200 more shares bring the average to $40 (total cost $5,000 + $4,000 + $6,000 = $15,000 over 400 shares = $37.50 — check: 100×50 + 100×40 + 200×30 = 5,000 + 4,000 + 6,000 = 15,000; 15,000 ÷ 400 = $37.50). The position now needs only $37.50 to break even, against $50 for the original purchase alone. The fall that looked like a 40% loss on the first purchase is a 25% loss on the averaged position — at the cost of four times the capital at risk.

That last clause is the risk, and it is the whole argument against indiscriminate averaging down. The technique assumes the fall is temporary and the company sound; when the fall reflects a permanent impairment — a broken business model, a fraud, a market that has moved on — averaging down converts a small loss into a large one, and the lower break-even price is cold comfort, because the price may never return to it. The discipline is to decide in advance: a maximum position size, a maximum number of averaging purchases, and a fundamental test the company must still pass before each purchase. Averaging down on a schedule, without re-examining the thesis, is not a strategy; it is the sunk-cost fallacy with a spreadsheet.

Averaging up, and buying in tranches

Averaging up is the mirror image: buying more as the price rises. The average cost rises with each purchase, and the break-even price rises with it, but the position is in profit throughout, and each purchase expresses confidence confirmed by the market. The risk is buying the top: a rise that reverses leaves a position whose later purchases were made at the highest prices, and whose average sits above the new price. The defence is the same as for averaging down — position sizing decided in advance, and a thesis re-examined at each purchase — plus the recognition that averaging up concentrates in a single position just as averaging down does.

Buying in tranches — splitting a planned investment into several purchases over weeks or months — is averaging without a price view: the purchases happen on a schedule, and the average is whatever the prices over that period produce. It reduces the risk of investing the whole amount on a single bad day, at the cost of the expected return forgone while the later tranches wait in cash. Whether that trade is worthwhile depends on the investor’s view of the market’s direction and their tolerance for regret — the regret of a lump sum that falls immediately versus the regret of tranches that miss a rise. The average-cost figure is the same in all three approaches; what differs is the process that produced it, and the process is where the risk lives.

Corporate actions and reinvested dividends: when the average moves without a purchase

A stock split changes the average without any money changing hands. A 2-for-1 split doubles the share count and halves the average cost per share: 100 shares at $50 become 200 shares at $25, total cost unchanged at $5,000. A 3-for-2 split multiplies the count by 1.5 and divides the average by 1.5. A reverse split does the opposite. A bonus issue (free shares) increases the count without increasing the cost, lowering the average. A spin-off distributes shares in a new company and allocates part of the original cost basis to them, lowering the average cost of the original shares. Each of these requires recomputing the average from the new share count and the unchanged (or reallocated) total cost, and each is routinely mishandled by investors who update the price but not the count.

Reinvested dividends are purchases. A $100 dividend reinvested at $40 buys 2.5 shares and adds $100 to the total cost; the average moves toward $40 by the weight of those 2.5 shares. Over years of reinvestment, the reinvested dividends become a substantial part of the cost basis, and omitting them understates the basis and overstates the eventual gain. Dividend reinvestment plans and fractional shares make the arithmetic fiddly but not difficult: every reinvestment is a lot with a date, a price, and a quantity, and it belongs in the record alongside the deliberate purchases.

The record behind the average, and the tax position

Keep the lot-level record: date, quantity, price, fees, and the resulting cost of every purchase, every reinvestment, and every corporate action. The average is derived from the record; the record cannot be derived from the average. When shares are sold, the gain or loss depends on which lots are deemed sold, and the available identification methods — first-in-first-out by default in many jurisdictions, specific identification where the broker and the rules allow — produce different results from the same sale. The holding period of the lots sold determines whether the gain is short- or long-term, with different rates. None of this is visible in the average cost figure, and a tax return prepared from the average alone will be wrong wherever lots differ in price or age.

Fees belong in the cost basis: commissions and charges paid on purchase increase the cost, and charges on sale reduce the proceeds. In many jurisdictions, the cost basis for tax is the total amount paid including purchase fees, and the proceeds are the sale amount less sale fees. The calculator on this page works with the amounts entered; whether fees are included is the investor’s choice at entry, and the choice should be consistent with the tax treatment the investor will claim. This guide is educational and does not constitute tax or investment advice; the tax treatment of share sales varies by jurisdiction and by the investor’s circumstances, and professional advice is appropriate where the amounts are material.

Frequently asked questions

How do I calculate my average stock price?

Total cost divided by total shares. 100 shares at $50 ($5,000) plus 100 at $40 ($4,000) = $9,000 ÷ 200 shares = $45 per share. Include purchase fees in the total cost if your tax treatment counts them as part of the basis.

What is averaging down in stocks?

Buying more shares as the price falls, which lowers the average cost per share and the break-even price. It increases the total amount invested in one company, so it concentrates risk — set a maximum position size and re-examine the investment thesis before each additional purchase.

Does a stock split change my average cost?

Yes, without any money changing hands. A 2-for-1 split doubles the share count and halves the average cost per share; total cost is unchanged. A reverse split does the opposite. Recompute the average from the new share count after every corporate action.

Do reinvested dividends affect average cost?

Yes — each reinvestment is a purchase at the reinvestment price, adding shares and cost. Omitting reinvested dividends understates the cost basis and overstates the gain when the position is sold. Record every reinvestment as a lot.

Is average cost the same as cost basis for tax?

Not exactly. The average summarises the position; tax is computed lot by lot, and which lots are deemed sold determines the gain, the holding period, and the rate. Keep the lot-level record — the average cannot reconstruct it.

Should I average down or cut my losses?

There is no universal answer: it depends on whether the fall reflects a temporary price move or a permanent impairment of the company, and on the position size relative to the portfolio. Decide the maximum position size and the thesis test in advance, and apply them mechanically — averaging down without re-examining the thesis is the sunk-cost fallacy.