Weighted average: when arithmetic mean lies and how to fix it
How to calculate weighted average and when it matters: GPA, portfolio returns, blended interest rates, and the cases where simple averaging quietly produces the wrong answer.
Arithmetic mean adds values and divides by the count. Weighted average adds values multiplied by their importance and divides by the sum of importances. Most real-world averaging problems need weighted average, not simple mean. The error of using simple mean when weights matter shows up in GPA calculations, portfolio returns, blended interest rates, and any aggregation where items differ in size.
How to calculate weighted average
The formula is straightforward:
`weighted_avg = (v₁ × w₁ + v₂ × w₂ + ... + vₙ × wₙ) ÷ (w₁ + w₂ + ... + wₙ)`
Multiply each value by its weight, sum the products, then divide by the sum of all weights. The weights normalize automatically — they do not need to add to 1 or 100. If they do (like proportions), the denominator is 1 and the formula simplifies to just the sum of products.
Worked basic example: course grades. Three exams worth different weights. Exam 1 score 85, weight 20%. Exam 2 score 92, weight 30%. Final exam score 78, weight 50%.
Weighted average = (85 × 20 + 92 × 30 + 78 × 50) ÷ (20 + 30 + 50) = (1700 + 2760 + 3900) ÷ 100 = 83.6.
Compare with simple arithmetic mean: (85 + 92 + 78) ÷ 3 = 85. The weighted result is lower because the final exam (lowest score) has the largest weight — simple averaging would have hidden that the heavy-weight piece dragged down the result.
Three real cases where it matters
GPA calculation. A college transcript shows: Calculus (4 credits, grade A = 4.0), History (3 credits, B+ = 3.3), Physics (4 credits, B = 3.0), Spanish (2 credits, A- = 3.7).
Simple mean of grade points: (4.0 + 3.3 + 3.0 + 3.7) ÷ 4 = 3.5.
Weighted by credits: (4.0 × 4 + 3.3 × 3 + 3.0 × 4 + 3.7 × 2) ÷ (4 + 3 + 4 + 2) = (16 + 9.9 + 12 + 7.4) ÷ 13 = 3.49.
Close in this case because credits are similar. But a 1-credit "easy A" course skews the simple mean and barely moves the weighted GPA. Universities use weighted because credits represent the time and rigor commitment, and that should drive how much each grade counts.
Portfolio returns. Investor holds two assets. Stock A: $1,000 invested, returned 50% over the year. Stock B: $100,000 invested, returned -10%.
Simple average: (50% + (-10%)) ÷ 2 = 20%. Looks like a great year.
Weighted by dollar amount invested: ($1,000 × 0.50 + $100,000 × (-0.10)) ÷ ($1,000 + $100,000) = ($500 − $10,000) ÷ $101,000 = -9.4%.
The actual portfolio lost about 9.4%. The 50% gain on the small position cannot offset the 10% loss on a position 100x larger — reporting your "average return" without weighting by position size is a textbook way to deceive yourself about portfolio performance.
Blended interest rate. Two loans. Loan A: $50,000 balance at 4.5% APR. Loan B: $20,000 balance at 9.0% APR.
Simple average: 6.75%. Weighted by balance: ($50k × 0.045 + $20k × 0.09) ÷ $70k = ($2,250 + $1,800) ÷ $70k = 5.79%.
Your effective rate on combined debt is 5.79%, not 6.75%. This matters when comparing against a refinance offer. A 6% consolidation loan looks worse than current setup if you use the simple average; it looks better than the actual blended rate.
When you actually use this
Investment performance reporting. Any time you report a return number, ask whether it is weighted by capital deployed or simple-averaged across positions. Time-weighted return (used for evaluating fund managers) and money-weighted return (used for evaluating personal investor outcomes) treat the same underlying data differently. The IRR (Internal Rate of Return) is the money-weighted version that accounts for both timing and amount of cash flows.
Course grading and academic standing. GPA is weighted by credit hours. Honors thresholds, scholarship eligibility, and graduate school admissions all use weighted GPA. The simple unweighted GPA most students calculate informally is wrong if courses have different credit loads.
Inflation indexes. CPI and PPI are weighted averages of price changes across categories, with weights set by what consumers actually spend money on. Housing typically gets around 40% weight in CPI, so housing price changes dominate the headline inflation number. Knowing the weight composition helps interpret official inflation against your personal spending mix.
The compounding error. Averaging multi-period returns by simple arithmetic mean produces the wrong number even before weighting by position size. A portfolio gaining 50% one year and losing 50% the next has arithmetic mean return of 0% — but actual end-state is $1.00 → $1.50 → $0.75, a 25% loss. Geometric mean (CAGR) is the right calculation here.
Try the calculator
Calculators mentioned in this post:
Weighted Average Calculator (GPA, Portfolios, ENEM)
Compute weighted average vs simple average. Use cases: GPA (course grade × credit hours), portfolio allocations, survey aggregation. Shows per-item contribution and weighted-vs-simple delta.
Percentage
Calculate percentages three ways: what is X% of Y, X is what percent of Y, and percentage change.
Compound Interest
Calculate compound interest with monthly contributions. See how your money grows over time.
Frequently asked questions
When do I use simple average vs weighted average?
Simple average works when every item carries equal importance. Five test scores treated as equal counts: simple mean is fine. Weighted average is needed when items have different sizes, durations, importance levels, or amounts. Weights of 1 each (equal) reduce to simple mean. Weights varying means the simple mean is biased.
Do weights need to sum to 1 or 100?
No. The formula divides by the sum of weights, so any consistent units work. Weights expressed as percentages (40%, 35%, 25%) sum to 100. Weights expressed as raw counts (5, 3, 2) sum to 10. Same answer either way. The convention of using percentages is just convenient when communicating; the math doesn't care.
What if some weights are zero?
Items with zero weight don't contribute to the numerator and aren't counted in the denominator. Effectively excluded. Useful when you want to compute a weighted average over a subset without removing items from the data. Some calculators silently exclude zero-weighted rows; others throw errors. Check what your tool does.
What's the difference between weighted average and weighted mean?
They're the same thing, just different names. "Weighted mean" is more common in statistics textbooks; "weighted average" is more common in finance and everyday usage. Both refer to the same formula: sum of (value × weight) divided by sum of weights. Some technical contexts also use "weighted arithmetic mean" for full precision (as opposed to weighted geometric mean or weighted harmonic mean, which use different aggregation rules).