The literature on security design has traditionally assumed that control of investment and production rests with the entrepreneur even after flotation. In this article I relax this assumption, which is both unnecessarily strict and unrealistic. The contribution is twofold. First, I show that when control is delegated during flotation, financial innovation is a trade off between agency costs (arising from principal agent relationships) and targeting costs (arising from the inability to split the cash flow). Second, I show that unlevered equity or the use of debt which can be issued without defaulting in equilibrium, are optimal means of financing. The strand of finance which deals with security design seeks to explain theoretically why certain contracts are used as means of corporate financing. If there are zero costs of issuing securities, we might well expect to find a myriad of contract designs. Empirically, there are however just a few contract designs which seem to dominate, namely debt and equity contracts. It is the simple designs which appear to survive, as understandably simplicity makes it easy to evaluate the risk of the securities. Within the class of simple contracts there still are unresolved theoretical issues, however, some of which I address in this paper. Perhaps the most puzzling aspect of corporate financing is why controlling securities have residual payoffs and non-controlling securities have fixed payoffs. This is puzzling in two respects; first, why is control associated with residual payoff, and second, why is risk not spread onto more than on security. In particular, it can be harmftLil to give controlling stakeholders too much discretion as it leads to agency costs (Green, 1984), and risk sharing concerns would normally imply that splitting the risk, giving a greater span in the market, is the optimal route. In this paper I investigate financial innovation in a context where both concerns apply, and where indeed the optimal security structure has the properties observed. The mnain reason for this, which is also the point at which the paper distinguishes itself from most of the literature, is that here the floatation is associated with delegation of control to the new owners in the market. In particular, I consider the problem of optimal security design for an entrepreneur who intends to float a firm when she must delegate the control of the firm to the new owners. Delegation of control implies two major concern for the entrepreneur. First, she seeks to design securities which take maximally advantage of existing investor clienteles in the economy. As an example we can think of firms that increase their value by issuing tax-efficient