What a blended cost basis is, and what it is not
Your blended cost is the share-weighted average of everything you paid, including commissions. Two hundred shares acquired as 100 at $50 and 100 at $35 cost $8,500 in total, which is $42.50 a share. That figure is your break-even price: sell above it and you have made money on the position, sell below it and you have not.
The weighting is by shares, not by trades. If you bought 900 shares at $50 and 100 at $35, the average is $48.50 rather than the $42.50 that averaging the two prices would give. That is the single most common arithmetic error in this calculation, and it always flatters the position when the smaller lot is the cheaper one.
What the blended cost is not is a signal. A position at $42.50 with the stock at $38 tells you nothing about whether the stock is cheap. The market does not know what you paid and does not care. Cost basis matters for two things only: measuring your result, and calculating your tax. Treating it as a decision input is anchoring, and it is exactly the bias that averaging down exploits.
It is also not automatically your tax basis. For US federal tax, the average-cost method is available for mutual fund shares and certain dividend reinvestment plan shares. Individual stock lots default to first-in-first-out unless you specifically identify which shares you are selling at the time of the sale. IRS Publication 550 sets out both rules, and the difference can change your tax bill substantially even though it never changes your economics.
The three formulas behind the outputs
The average. Add up every lot's quantity times its price, add the commissions, and divide by the total share count. Commissions belong in the numerator because they are money that left your account to acquire the position; leaving them out understates your break-even.
The break-even move. This is the percentage change in the current price needed to reach the average, (P̄ ÷ P − 1) × 100. Note it is not the same as the percentage you are down. A position down 10.59% needs an 11.84% rise to recover, because the rise is measured from the lower base. That asymmetry grows fast: down 50% needs +100%, and down 80% needs +400%.
The top-up requirement. Suppose you hold Q shares costing C in total and buy n more at price P, wanting the new average to be exactly T. Then (C + nP) ÷ (Q + n) = T. Multiply out: C + nP = TQ + Tn, so n(P − T) = TQ − C, giving n = (TQ − C) ÷ (P − T).
Read what that expression tells you. The denominator is P − T, so as your target approaches the current price the share count you need goes to infinity. That is not a quirk of the algebra — it is the actual behaviour of averaging. Every share you buy at $38 pulls the blend toward $38 and can never pull it past. If your target is below the price you are buying at, no quantity achieves it, and the calculator returns a dash rather than a misleading number.
Worked example: 100 shares at $50, 100 more at $35, now trading at $38
These are the calculator's defaults, so you can check each line on the page.
- Cost of the first lot. 100 × $50 = $5,000.
- Cost of the second lot. 100 × $35 = $3,500.
- Total cost and shares. $8,500 across 200 shares.
- Blended average. $8,500 ÷ 200 = $42.50. The second purchase pulled the average down from $50 by $7.50.
- Market value. 200 × $38 = $7,600.
- Unrealised result. $7,600 − $8,500 = −$900, which is −10.588% of cost.
- Break-even move. $42.50 ÷ $38 = 1.118421, so the stock must rise 11.842% to get you back to flat — more than the 10.588% you are down.
- Top up to a $40 average. n = (40 × 200 − 8,500) ÷ (38 − 40) = (8,000 − 8,500) ÷ (−2) = 250 shares, costing 250 × $38 = $9,500.
Step 8 is the one that changes minds. Shaving $2.50 off a $42.50 average costs $9,500 — 112% more capital than the $8,500 already committed — and it more than doubles your exposure to a single position. The average falls to $40 and your break-even move falls from 11.842% to 5.263%. The dollar loss at $38 does not move at all — it stays at $900, because shares bought at the market price carry no gain or loss of their own — but it is now spread across 450 shares instead of 200, so the loss per share falls from $4.50 to $2.00. What actually grew is exposure: $18,000 of cost in one name instead of $8,500.
That is the trade being made whenever anyone averages down: a lower break-even price bought with a larger position. Neither number is the whole story on its own.
When averaging down is a plan and when it is a reflex
Averaging down is defensible when it is the plan you wrote before the price moved. If you sized a position at half your intended weight and specified in advance that you would add the rest on weakness, then buying more at $35 is executing a plan. Your target weight was always the full amount and the market simply gave you a better entry.
It is a reflex when the reason for buying more is that you already own it. The tell is that the size of the purchase is set by how far the price fell rather than by conviction or by position limits. A stock at $35 that you would not buy today, at that price, with fresh money is not a stock you should be buying because your average is $50.
Three checks make the difference concrete. First, position limit: after the top-up, what share of your portfolio is this one name? The default example takes a position from $8,500 to $18,000 of cost. Second, the fresh-money test: would you initiate this position at today's price if you owned none? Third, thesis versus price: has anything changed in the business, or only in the quote? A price fall that came with a broken thesis is a reason to reduce, and averaging down into it is how a small mistake becomes a large one.
Also watch the tax rules if you are selling anything at a loss in the same window. Buying substantially identical shares within 30 days before or after a loss sale triggers the wash-sale rule: the loss is disallowed and instead added to the basis of the replacement shares. That interaction catches people who sell to harvest a loss and simultaneously average down in the same name.
Cost of pulling a $50 average down, buying at $35
| Target average | Extra shares needed | Extra capital at $35 | Total position | Rise from $35 to break even |
|---|---|---|---|---|
| $45.00 | 50 | $1,750 | 150 sh | 28.57% |
| $42.50 | 100 | $3,500 | 200 sh | 21.43% |
| $40.00 | 200 | $7,000 | 300 sh | 14.29% |
| $38.00 | 400 | $14,000 | 500 sh | 8.57% |
| $37.00 | 650 | $22,750 | 750 sh | 5.71% |
| $36.00 | 1,400 | $49,000 | 1,500 sh | 2.86% |
The capital required is a hyperbola in the target, not a straight line: halving the remaining gap roughly doubles the shares needed, and reaching $35 exactly would take infinitely many.
Details that change the answer
- Commissions and fees belong in the basis. On a small position they matter: $20 of fees on 200 shares adds $0.10 per share, moving a $42.50 average to $42.60.
- Dividends received do not reduce your cost basis unless they were reinvested, in which case each reinvestment is a new lot at that day's price.
- Return of capital distributions do reduce basis, which matters for some funds, trusts and partnerships. Check the year-end tax statement rather than assuming.
- Splits change the share count and the per-share price but not the total cost. After a 2-for-1 split, double every historical quantity and halve every historical price before entering the lots.
- Currency matters for foreign shares. Your basis is in your home currency at the exchange rate on each purchase date, so a stable share price can still produce a gain or loss.
- Selling costs are not included here. The break-even price above recovers your purchase outlay; add any sale commission to clear the position exactly flat.
Averaging down, dollar-cost averaging and position sizing
Averaging down and dollar-cost averaging are frequently confused and are not the same thing. Dollar-cost averaging is a schedule fixed in advance and indifferent to price: the same amount goes in every month, which mechanically buys more shares when prices are low. Averaging down is a discretionary purchase triggered by a price fall. The first is a rule that removes judgement; the second is a judgement that borrows a rule's respectability.
The framework that actually resolves the decision is position sizing. Decide the maximum share of your portfolio this holding may reach, decide it before the price moves, and let that cap govern every top-up. The portfolio rebalancing calculator will tell you what a proposed purchase does to your weights, which is the number that constrains you when the story is compelling and the price is falling.
When you do eventually sell, the basis you calculated here drives the tax. Which lots you sell can matter more than when: selling the $50 lot realises a loss while selling the $35 lot may realise a gain, even though the blended average sits between them. That election is only available if you identify the shares at the time of sale under Publication 550's specific identification rules. Run the outcome through the capital gains tax calculator before you place the order, and check the realised return with the CAGR calculator so you are judging the position on an annualised basis rather than on a raw percentage.