Overview / Key Takeaways
– A weighted average (weighted mean) assigns different levels of importance (weights) to data points before averaging, making the result reflect relative significance.
– Formula (general): weighted average = (Σ (weight_i × value_i)) / (Σ weight_i). If weights sum to 1, the denominator can be omitted.
– Common finance uses: cost-basis of shares, portfolio returns, inventory valuation (weighted-average method), and the weighted average cost of capital (WACC).
– Advantages: more accurate representation when items differ in importance; reduces undue influence of outliers when weights downweight extreme values.
– Disadvantages: weights can be subjective and introduce bias; results may be sensitive to small weight changes; interpretation can be less intuitive than a simple mean.
– Source: Investopedia (see link at end).
What Is a Weighted Average?
A weighted average multiplies each data point by a predetermined weight representing its importance, sums the weighted values, and divides by the sum of weights. It is used whenever data points do not contribute equally to the overall figure.
Core formula
– If weights may not sum to 1:
weighted average = (Σ (w_i × x_i)) / (Σ w_i)
– If weights are normalized (Σ w_i = 1):
weighted average = Σ (w_i × x_i)
When to use a weighted average (purpose)
– Different observations have different sizes, frequencies, or relevance (e.g., share lots, survey strata, component sizes in a fund).
– You want to reflect economic significance, not just count each item equally.
– To produce fair representations when sample strata are imbalanced (survey weighting).
– To compute averages where quantities are directly additive (e.g., average price per share across multiple purchases).
Practical steps to calculate a weighted average
1. Define the value(s) to average (x_i). Examples: prices, returns, grades, inventory unit costs.
2. Choose objective weights (w_i) reflecting importance: quantities, market values, frequency, predetermined grade weights, etc.
3. Ensure units align: weights should correspond to the value’s underlying quantity (shares, dollars, units).
4. (Optional but recommended) Normalize weights so Σ w_i = 1 by dividing each w_i by total weights. This makes interpretation easier.
• Normalized weight_i = w_i / Σ w_i
5. Compute weighted sum: Σ (w_i × x_i)
6. Divide by Σ w_i if you haven’t normalized. Result is the weighted average.
7. Validate: check dimensions (units) and run sensitivity checks (change weights slightly) to see how stable the result is.
Worked examples
Example 1 — Weighted average purchase price (shares)
– Purchases: 100 shares at $10; 50 shares at $40.
– Values × weights: (100 × 10) + (50 × 40) = 1,000 + 2,000 = $3,000.
– Total shares (Σ weights) = 150.
– Weighted average price = $3,000 / 150 = $20 per share.
Example 2 — Portfolio weighted return
– Holdings market values: Stock A $6,000, Stock B $4,000. Returns: r_A = 8%, r_B = 12%.
– Weights (by market value): w_A = 6,000 / 10,000 = 0.6; w_B = 0.4.
– Portfolio return = 0.6 × 8% + 0.4 × 12% = 4.8% + 4.8% = 9.6%.
Example 3 — Weighted Average Cost of Capital (WACC)
– Let E = market value of equity, D = market value of debt, V = E + D.
– WACC = (E/V) × R_e + (D/V) × R_d × (1 − T_c)
where R_e = cost of equity, R_d = cost of debt, T_c = corporate tax rate.
– Weights are based on market values of capital components.
Example 4 — Inventory (weighted-average method)
– Purchase 100 units at $10, then 50 units at $12.
– Weighted average unit cost = (100×10 + 50×12) / (150) = (1,000 + 600) / 150 = $1,600 / 150 = $10.67 per unit.
– Use that cost to value cost of goods sold and ending inventory under the weighted-average inventory method.
Weighted Average vs. Arithmetic vs. Geometric Mean
– Arithmetic mean (simple average): sum of observations divided by count. Use when each observation is equally important and data are independent.
arithmetic mean = (Σ x_i) / n
– Geometric mean: nth root of the product of values (or exp(mean of logs)). Use for multiplicative processes like compounded returns.
geometric mean of returns = (Π (1 + r_i))^(1/n) − 1
– Weighted average complements both: you can have a weighted arithmetic mean or a weighted geometric mean. For compounded returns of a portfolio held continuously, use weighted arithmetic for compositional returns and geometric mean for multi-period compounded returns.
Is weighted average “better”?
– It’s not universally better; it’s more appropriate when data points have differing importance. For equal-importance data, the simple arithmetic mean is sufficient and simpler.
– For returns across time (compounding), geometric mean is the appropriate average.
Best practices and practical tips
– Use objective, measurable weights (quantities, market values, population shares) when possible to avoid bias.
– Document and communicate the weighting scheme to users of the result.
– Normalize weights for clearer interpretation.
– Check units: avoid multiplying units that don’t match (e.g., weighting percentages by counts without converting).
– For portfolio and capital-structure measures, prefer market values over book values when available (WACC and portfolio weights), unless there is a specific accounting reason not to.
– For returns over multiple periods, use geometric mean to capture compounding; for computing a weighted portfolio return at a point in time, use weighted arithmetic mean by market-value weights.
– Run sensitivity analysis: change weights a bit to see how much the weighted average shifts.
Common pitfalls
– Using arbitrary or subjective weights without justification.
– Forgetting to divide by the sum of weights (if weights are not normalized).
– Mixing units or mismatching the weight’s basis (e.g., weighting by number of trades vs. dollars invested).
– Applying an arithmetic weighted average to compounded growth problems — use geometric mean instead when compounding matters.
– Using book values instead of market values when market values better reflect current economic weights.
Advantages and disadvantages (summary)
Pros
– Reflects real-world importance of components.
– Reduces distortion from small but extreme observations if those are downweighted.
– Flexible and applicable across finance, statistics, and operations.
Cons
– Potential subjectivity in choosing weights.
– Sensitive to weight changes; small adjustments can yield different conclusions.
– Slightly more complex to interpret than simple averages.
When you should be cautious
– If weights are chosen to achieve a desired outcome rather than to reflect objective importance.
– When the underlying data are volatile and weights are derived from unstable quantities.
– When comparing weighted averages across datasets with different weighting schemes—ensure comparability.
Further examples of uses in finance
– Mutual fund returns (value-weighted average of holdings’ returns).
– Corporate valuation (WACC used as discount rate).
– Index construction (e.g., market-cap-weighted indexes).
– Earnings-per-share when computing weighted-average shares outstanding (accounting).
Practical checklist before publishing a weighted average result
1. Are the values and weights clearly defined and measurable?
2. Do weights reflect the intended importance (quantity, market-value, frequency)?
3. Are weights normalized or is the denominator included in the calculation?
4. Have you checked units and time alignment?
5. Did you test sensitivity to plausible changes in weights?
6. Have you documented the method and rationale for the weights?
References and further reading
– Investopedia — “Weighted Average” (Paige McLaughlin)
Editor’s note: The following topics are reserved for upcoming updates and will be expanded with detailed examples and datasets.