This chapter asserts that the management of the assets of defined benefit (DB) plans has been guided by a simple three-part paradigm. First, designers assumed that the generous equity risk premium observed in the past will be available in the future. Second, they have posited that a constant asset mix policy portfolio such as 60-40 or 70-30 (in stocks) will provide adequate protection against shorter-term equity markets volatility. Third, given the first two assumptions, it can be deduced the bulk of a fund’s human and financial resources can be spent attempting to add modest additional return to the fund’s policy portfolio return by taking on a modest amount of additional risk. We show that within this old paradigm framework, the average US DB pension fund performed reasonably well over the 10-year period ending 2002. But we also show that the choice of policy portfolio largely determined total DB plan balance sheet mismatch risk, and that on average, this major risk went unrewarded over the 10-year measurement period. Our findings raise questions about the appropriateness of the old pension paradigm; instead, we assert that a new paradigm is required which include a defensible set of investment beliefs which include a time- varying equity risk premium; an integrative investment model that directly links a varying return opportunity set to DB balance sheet stakeholder income needs and risk tolerances; and a human systems-based decision-making protocol that can dynamically integrate the first two elements into the production of measurable stakeholder value over time.