急！！！关于capm！请大家都进来看看~~[悬赏]

CAPM tells us that the expected return is solely determined by beta. But we are all accustomed to hearing that well-managed firms will provide higher rates of return. Is CAPM wrong?
如果可以的话，请尽量回答地具体些~~谢谢！

Alpha = Return of the fund - [Risk free rate + Beta * (Return of the market ptfl - Risk free rate)]

Hey, plz refer to the above, you see the second part on the right is the return predicted by the CAPM. so alpha on the left (Jensen's alpha, to be precise) is the excess return of the fund relative to the return of the benchmark (the market ptfl). For instance, if CAPM estimates that a portfolio should earn 6% based on the risk of the portfolio but the portfolio actually earns 8%, the portfolio's alpha would be 2%. This 2% is the excess return over what was predicted in the CAPM model. ppl usually use it as a measure of fund manager's skills

{:3_60:}怎么米有人来回答捏。。。。。

顶顶顶顶顶~~~不能就这么沉了啊~~~~

PS: Most finance theories are based on the efficient market hypothesis, and under this assumption, the alpha would be zero. But in reality, it's never the case, alpha could be negative or positive

I think the up floor tell the answer,  CAPM only works on specified assumption .

CAPM模型是在若干假设的条件下成立的，是理想化的模型。它只是解释了系统性风险和单个股票收益之间的关系。相比较而言APT模型更为合理化。对于楼主说的最后的问题我觉得在短期来讲可能会出现这样的情况，但从长期的趋势来说，没有人能持续战胜市场。CPAM模型本身并没有错，只是在应用上有局限性。说的不对请多多原谅

用roll's critique也可以解释啊
capm没有错，the only legitimate test of capm is you choose the correct market portfolio。 之前的market portfoliO 是对的，那么就不会有deviation from sml， 反之，如果市场组合选错了，那么any order of ranking of return is possible。
camp 本身去验证基本是不可能的。所以才有apt 啊

{:3_59:}哈哈~谢谢各位喽~~~~~不知道还有没有别的见解捏？？

8# sausau

个人觉得Roll's critique只说明了past attempts to test capm virtually did nothing and capm is not testable due to the unobtainability of the true market portfolio，所以他觉得一个test 不了的model不是好model，所以他提倡用apt. 而且Roll的观点只是过去的那些attempts其实都没有在真正test capm这个模型，而他即不能证明那些人是对的，也不能证明那些人是绝对错的，他只是证明了如果alpha不是0，那么用来做market proxy的那个portfolio肯定不是tagent portfolio而已，但虽然capm不能因为此而被证明是错的，却也无法因为此被证明是对的，原因就是true market portfolio is not observable.