Prices and Portfolio Choices in Financial Markets: Theory, Econometrics, Experiments

Abstract

Many tests of asset-pricing models address only the pricing predictions, but these pricing predictions rest on portfolio choice predictions that seem obviously wrong. This paper suggests a new approach to asset pricing and portfolio choices based on unobserved heterogeneity. This approach yields the standard pricing conclusions of classical models but is consistent with very different portfolio choices. Novel econometric tests link the price and portfolio predictions and take into account the general equilibrium effects of sample-size bias. This paper works through the approach in detail for the case of the classical capital asset pricing model (CAPM), producing a model called CAPM + . When these econometric tests are applied to data generated by large-scale laboratory asset markets that reveal both prices and portfolio choices, CAPM + is not rejected.

Competition in Portfolio Management: Theory and Experiment

“…This replicates the findings in past asset pricing experiments where investors trade assets directly (e.g., Plott 2004, Bossaerts et al 2007). The coexistence of CAPM prices and a unimodal distribution of investor holdings centered at the market portfolio is theoretically founded by a variant of the CAPM that relaxes the assumption of quadratic utility, allowing for individual deviations that wash out in aggregate (CAPM + , Bossaerts et al 2007). Figure 2 also shows, importantly, that this unimodal distribution of investor final wealth is maintained to a large extent 22 The median DistanceAD across managers and periods was 1.23 for type X, whereas it was 0.83 for type Y , and 0.77 for type Z.…”

“…Moreover, we know that the largest fund in period 4 was of type Y , not X, and the largest fund in period 5 was of type Z, not X. For the CAPM pricing to obtain robustly, it is known that individual idiosyncrasies need to average out, and this can only occur if there are a sufficient number of participants, all of whom only hold a small fraction of wealth (Bossaerts et al 2007). Consistent with the hypothesis that concentration has a detrimental effect on pricing, we discovered that the correlation between concentration measures and the statistic is negative, albeit only marginally significant.…”

“…25 We focus on the weak CAPM, but one may be interested to know whether also the strong CAPM obtains. The CAPM requires the assumption that investors have quadratic utility (or that a quadratic function is a good approximation of the true utility; see Bossaerts et al 2007). In our setting where shares and fees are computed using expected payoffs, even with the assumption of quadratic utility, prices gross of fees satisfy the CAPM only approximately.…”

“…Absent delegation, the setup is one that had been tested extensively in the laboratory and the experiments had provided robust support for the theory (Asparouhova et al 2003, Bossaerts and Plott 2004, Bossaerts et al 2007). As such, there exists a solid benchmark against which to calibrate an experiment with delegation.…”

“…The design closely followed that of earlier experiments that had reliably generated weak CAPM pricing in the absence of portfolio delegation (Asparouhova et al 2003, Bossaerts and Plott 2004, Bossaerts et al 2007. During all replications, 32 subjects participated as managers.…”