After considering risk control through VaR-based limits and risk-adjusted performance measurement, we can now discuss the final step of the capital allocation process. While capital allocation has sometimes been addressed in more theoretical settings by a number of important papers (Froot and Stein 1998; Perold 1999; Stoughton and Zechner 1999, 2000; Denault 2001), in practice it represents an organizational process that is deeply intertwined with the strategy formation process. Just like corporate strategy, it can alternate between periods of sharp radical changes and periods of incremental changes. The former may lead to an abrupt reallocation of capital and other resources to new businesses that are perceived to offer the potential for higher risk-adjusted return. The latter may imply marginal (but sometimes repeated) capital reallocations deriving from the bank s learning process about which businesses appear to be able to deliver superior performances to shareholders. As a consequence, reallocations should always be driven by future expected performances. But past performances may also still play a role. In fact, a capital reallocation may derive from a strategic decision that has little to do with past numbers and is based largely on a forecast (think about a decision concerning an acquisition or the choice to expand geographically or to enter into a new product niche). But sometimes, instead, a capital reallocation may, at least in part, be the reaction to the satisfactory or unsatisfactory performances of certain existing business units.

For these reasons, we devote this chapter to two main topics. The first is...