This is a terrible example of statistical thinking -- especially from a Bayesian point of view. What's the prior probability that you assign to a chicken coming from one the rocks? If it isn't zero, then... I don't know what to say.

...and modeled badly (e.g., anything quants have done in the last, well, ever). Even Oracle of Omaha has an implicit model when buying conservative stock that society needs (e.g., toothpaste). Models are everywhere... What an odd thing to assert about financial markets from interesting article on bacterial protein synthesis.