I wrote a post on November 29, 2015 on how difficult it is to determine conditional probabilities for candidates to win the general election, should they get their party's nomination. The key issues are: (1) there is a range of possible conditional probabilities, which is wide for candidates with low probability of winning their party's nomination (2) in order to calculate this I need to make a strong assumption on how the probability of winning the general updates with the probability of winning the nomination (3) any possible third party runs will inflate the conditional probability (i.e., Donald Trump could be inflated).
Hillary Clinton is consistently in the mid-60's. Marco Rubio is consistently in the high 30's. Ted Cruz is consistently in the low 30's. Trump is moving up and down around Rubio. The markets are giving him a non-negligible probability of running as an independent. Even 1% probability of winning as an independent would inflate his conditional probability of winning the general election. So, it is hard to disentangle that result, but it is very possible that markets view Trump as on par with Rubio, but having more variability.