Systematic and unsystematic risk

Total risk. The return on a share or a project varies for two kinds of reasons, so its total risk has two parts. Not all risk is equal: some risk can be diversified away, and some cannot. Diversification lets good and bad events in different holdings cancel out, reducing total variability without reducing expected return.

Unsystematic (diversifiable, specific) risk. Risk peculiar to a company or industry: a strike, the loss of a major customer, a product failure, a management change, a lawsuit. Such events affect one firm and not the market as a whole, so in a well-diversified portfolio losses on some holdings are offset by gains on others. Because investors can eliminate it cheaply by diversifying, the market does not reward them for bearing it.

Systematic (non-diversifiable, market) risk. Risk that affects all firms together: changes in interest rates, inflation, recession, tax policy, political events. It cannot be removed by diversification, because all securities move with the market to some degree. It is measured by beta (β), the sensitivity of a share's return to the market's return: β = 1 moves with the market, β > 1 more than the market, β < 1 less.

Why the distinction matters. Since only systematic risk remains in a diversified portfolio, only it is priced. Under the CAPM the cost of equity is the risk-free rate plus a premium for systematic risk: Ke = Rf + β(Rm − Rf), the premium being the excess of the diversified market portfolio's return over the risk-free return, scaled by beta. A firm's cost of capital, and so its investment and financing decisions, therefore depend on its systematic risk.