Speculative risk grows with financial innovation

Interesting work by Alp Simsek of MIT in his paper, "Speculation and risk sharing on new financial assets". He  decomposing average portfolio variance into two parts, uninsurable variance (that which cannot be diversified away) and speculative variance. He argues that financial innovation through new assets will decrease unisurable or actual risk but there will be an increase in speculative variance. This is caused by differences in beliefs across investors that can be traded when there are new assets - innovation.

Belief disagreement which can be expressed through new assets will increase speculative variance. New assets will create new bets or will amplify existing bets through the new sets of instruments available for trading. Additionally, if there is less disagreement there will more speculative variance when there are more unique ways to express a view.

So what is he telling us? Financial innovation is  bad thing and should not be allowed? Derivatives are inherently bad because it allows for speculation that was not possible before?  The math may be sound but the result are not what one would expect.