Arguably, the myth is that markets are efficient. That is to say that the clearing price is the expression of all the information available to participants is a

Arguably, the myth is that markets are efficient. That is to say that the clearing price is the expression of all the information available to participants is a fiction. This idea of the Turing completeness, or maybe complexity class of your transaction logic determining whether it is an endogenous logical arbitrage trade, or an exogenous manipulation scheme may have some really appealing features. Hypothetically, if the steps of your transaction logic operate on or recurse over feedback into and from the market, you are in fact, "manipulating," it. I'd wonder how describing manipulation in terms of recursion limits and feedback would impact the definiton. Whereas, if you are precalculating or front running some periodic market function, you are arbitraging it with endogenous market information and that makes it "legit." Where this guy might be vulnerable in that model is the question of how far upstream of his actual transaction did he get before the feedback loop he was operating over is not considered a part of that market - and whether his arbitrage was legit because it was between markets.