RT Generic T1 CAPM-like formulae and good deal absence with ambiguous setting and coherent risk measure A1 Balbás, Beatriz A1 Balbás, Raquel A2 Universidad Carlos III de Madrid. Instituto para el Desarrollo Empresarial (INDEM), AB Risk measures beyond the variance have shown theoreticaladvantages when addressing some classical problems of Financial Economics, atleast if asymmetries and/or heavy tails are involved. Nevertheless, in portfolioselection they have provoked several caveats such as the existence of good dealsin most of the arbitrage free pricing models. In other words, models such asBlack and Scholes or Heston allow investors to build sequences of strategieswhose expected return tends to in nite and whose risk remains bounded ortends to minus in nite. This paper studies whether this drawback still holds ifthe investor is facing the presence of multiple priors, as well as the propertiesof optimal portfolios in a good deal free ambiguous framework.With respect to the rst objective, we show that there are four possibleresults. If the investor uncertainty is too high he/she has no incentives to buyrisky assets. As the uncertainty (set of priors) decreases the interest in riskysecurities increases. If her/his uncertainty becomes too low then two types ofgood deal may arise. Consequently, there is a very important di¤erence betweenthe ambiguous and the non ambiguous setting. Under ambiguity the investoruncertainty may increase in such a manner that the model becomes good dealfree and presents a market price of risk as close as possible to that re ected bythe investor empirical evidence. Hence, ambiguity may help to overcome somemeaningless ndings in asset pricing.With respect to our second objective, good deal free ambiguous modelsimply the existence of a benchmark generating a robust capital market line.The robust (worst-case) risk of every strategy may be divided into systemicand speci c, and no robust return is paid by the speci c robust risk. A coupleof betas may be associated with every strategy, and extensions of the CAPMmost important formulas will be proved. SN 1989-8843 YR 2011 FD 2011 LK https://hdl.handle.net/10016/12636 UL https://hdl.handle.net/10016/12636 LA eng NO Partially supported by "RD-Sistemas S.A", Welzia Management SGIIC SA and "MICINN" (Spain, Grant ECO2009-14457-C04) DS e-Archivo RD 20 may. 2024