Deployed correctly, they can spur competition. Finance has no concept of intellectual property. If a small shop invents a security, and it does well, the big shops copy it overnight. This is good, from a product-diversity perspective. It is bad, however, from a firm-diversity perspective. Balancing those perspectives is what this debate is about. Slanting to one extreme or another isn't helpful.
How is it bad from a firm-diversity perspective? If multiple companies develop the same product then a diverse range of businesses sell that security.
Of course, one could argue that "innovation" in financial markets has had a lot of bad consequences. There are very few truly innovative and solid ideas that don't cause huge amounts of risk. Black-Scholes algorithm is one example of good innovation that comes to mind, but even that caused problems for their inventors.
Collateralised debt obligations, credit-swaps, contracts for difference, most derivatives, high-frequency trading and binary options are all hugely high risk and/or distort markets. HFTs in particular are egregious, they are hard to analyse and rely on huge numbers of transactions being made and, to my mind, allow traders to make money not on the qualities that allows commerce to do good for society, but only benefits those who aren't the first to attempt to buy or sell a share based on what others are doing, but by looking at 3hat they are doing and jumping in first. The only winners there are those looking to improve high speed Internet transactions, but even then the gains are limited.
Deployed correctly, they can spur competition. Finance has no concept of intellectual property. If a small shop invents a security, and it does well, the big shops copy it overnight. This is good, from a product-diversity perspective. It is bad, however, from a firm-diversity perspective. Balancing those perspectives is what this debate is about. Slanting to one extreme or another isn't helpful.