One of the reasons that there are currently so many stock exchanges in the US is that while you aren't allowed to quote at subpenny levels, if you have several exchanges with different levels of make/take fees, you effectively have the ability quote (or remove liquidity) at subpenny price levels net of fees and rebates. In that scenario the proliferation of exchanges is less about competition and more about finding clever ways of negotiating the regulatory environment.

Yesterday Levine posited that less fragmentation of exchanges makes more economic sense:

http://www.bloombergview.com/articles/2016-02-24/oil-loans-a...

I found it odd given what we saw in the US exchanges was the opposite of that before fragmentation, but the argument is that the network effects of large exchanges reduces the overall costs.

On exactly what front are exchanges competing? Order types? Fees? Volume? Differences in any of these are very often opaque to the customer.

From what I've seen, exchanges are basically functionally equivalent, since they provide a utility-like service. And, as Vanguard points out, more trading venues means higher market frictions and costs.

TBH I'm still trying to forming a strong opinion on the issue, as indicated by "probably" in my initial comment. And I think it's a topic worth discussing.