The well-known capital asset pricing model asserts that only a single number—an asset's "beta" against the market index—is required to measure risk. Arbitrage pricing theory asserts that an asset's riskiness, hence its average long-term return, is directly related to its sensitivities to unanticipated changes in four economic variables—(1) inflation, (2) industrial production, (3) risk premiums, and (4) the slope of the term structure of interest rates. Assets, even if they have the same CAPM beta, will have different patterns of sensitivities to these systematic factors. The central focus of portfolio strategy is the choice of an appropriate pattern of sensitivities. This choice will depend upon the economic characteristics of the beneficiary of the portfolio's income—in particular, upon whether it is more or less concerned with the economic factor risks than the broad average of investors. An organization, for example, may expect to spend less on food than the average investor. To the extent that food prices coincide with general inflation and are somewhat independent of productivity risk, the organization's optimal portfolio could have a lower inflation beta and a higher productivity beta than the broad-based average has. Such concerns have tactical implications. The organization unconcerned about inflation in agricultural prices would also wish to skew its portfolio holdings out of this sector. Conversely, a sensitivity to energy costs might lead it to skew its holding in the direction of the energy sector. Implementation of the chosen strategy might be carried out directly by the fund itself, or it might be delegated to select investment mangers who follow established investment policy guidelines.