How this instrument works
Portfolio beta blends the market-sensitivity figures of two holdings you already know into one number for the combined position. It doesn't measure anything new about either asset on its own — Beta of asset 1 and Beta of asset 2 are inputs you bring in from a broker's data page, a screener, or your own regression. This instrument only answers what happens once you own both together, in the proportions you actually hold them, which is a different question than either individual beta answers alone.
The formula is a weighted average, but the denominator is written as w1 + w2 rather than assumed to equal 1, and that choice matters in practice: the two weight fields don't need to be pre-converted into portfolio percentages first. Enter 60000 and 40000 for two positions worth those dollar amounts, or 0.6 and 0.4 if you've already worked out the split, and the ratio behind Portfolio beta comes out identical either way, because dividing by the sum of the weights normalizes whatever units you used.
The result inherits every limit of the two betas fed into it. If Beta of asset 1 was estimated from two years of weekly returns and Beta of asset 2 from five years of monthly returns, blending them produces a figure that looks precise but rests on two different clocks. A weighted average of beta also says nothing about how the holdings' non-market risk interacts with each other — two portfolios can share the same Portfolio beta yet swing very differently in practice, depending on whether the company-specific part of each return tends to move together or cancel out.
- Enter Weight of asset 1 — the size of your first holding, as a portfolio fraction or a raw dollar amount.
- Enter Beta of asset 1 — that holding's own beta, sourced from a broker, a data provider, or your own calculation.
- Repeat with Weight of asset 2 and Beta of asset 2 for the second holding in the mix.
- Read Portfolio beta — the size-weighted blend, showing how the combined position tends to move against the market.
- Keep both weight fields in the same unit — mixing one dollar figure with one percentage will not normalize correctly, since the formula only divides by their sum.
Worked example — a 60/40 growth-defensive split
Take the default case: Weight of asset 1 is 0.6 with Beta of asset 1 at 1.2 — a growth stock — and Weight of asset 2 is 0.4 with Beta of asset 2 at 0.8, a steadier holding. The weighted numerator is (0.6 x 1.2) + (0.4 x 0.8) = 0.72 + 0.32 = 1.04, and because the two weights already sum to 1, dividing by (0.6 + 0.4) leaves that figure unchanged, so Portfolio beta reads 1.04.
That 1.04 sits between the two inputs, closer to 1.2 than to 0.8 because the growth stock carries the larger weight — the combined position is expected to move a little further than the market on average, but nowhere near as sharply as the growth stock alone would. Shift the split to an even 50/50 with those same two betas and Portfolio beta lands at exactly 1.0, which shows how the weighting, not just the pair of betas themselves, decides where the blended figure falls.
Questions
Do Weight of asset 1 and Weight of asset 2 need to add up to 1?
No — the formula divides by their sum, so raw dollar amounts work exactly like pre-computed fractions. Enter 60000 and 40000 for two positions worth those amounts, or 0.6 and 0.4 if you've already found the split; Portfolio beta comes out identical either way because the denominator normalizes whatever units the two weight fields are in.
Why not just average the two betas without weighting them?
An unweighted average treats a small position the same as a dominant one, describing a portfolio that doesn't exist. A $95,000 stake in a beta-1.2 stock next to a $5,000 stake in a beta-0.4 stock behaves almost exactly like the larger holding alone, close to 1.2, not like the simple average of 0.8. Weighting by size is what makes Portfolio beta describe the position you actually hold.
Does a lower portfolio beta mean the portfolio is safer overall?
Not entirely — beta only measures sensitivity to market-wide swings, not the stock-specific risk each holding still carries on its own. Two pairs of holdings can share the same Portfolio beta yet behave differently in practice, depending on how their company-specific returns interact; a low blended figure says the combination tracks the market closely, not that it carries no risk.
Can Portfolio beta ever fall outside the two individual betas?
No. A weighted average always lands between its two inputs, so Portfolio beta can never exceed the larger of Beta of asset 1 and Beta of asset 2, or fall below the smaller one, no matter how the weights are split. In the default example, 1.04 sits between 0.8 and 1.2 for exactly this reason — the weighting decides where in that range it lands, not whether it can escape the range.
If both holdings have a beta of 1.0, can the portfolio beta differ?
No — the weighted average of two identical numbers is that same number regardless of the weights used, so any mix of two beta-1.0 holdings returns a Portfolio beta of exactly 1.0. This is why adding more names to a portfolio doesn't by itself change its beta: the figure only moves if the added holdings' own betas differ from what is already in the mix.
Where do Beta of asset 1 and Beta of asset 2 come from?
Not from this instrument — each is sourced beforehand, either from a broker's stock page, a data provider's published figure, or computed directly from a stock's covariance with the market divided by the market's own variance. Portfolio beta only combines two betas you already have; it does not estimate either one from raw price history.
References
- U.S. SEC Investor.gov — diversification and portfolio risk basics
- NYU Stern (Damodaran) — industry beta tables for sourcing individual betas
Read this first: This instrument shows arithmetic, not advice. Real offers add fees, taxes and terms that vary by lender and place — verify the figures against your actual paperwork before deciding anything.