Simon Johnson is right to worry about the SEC being overmatched by existing interests in securities markets. I think the biggest problem is that the people with the access to the data that needs to be analyzed are embedded in the status quo.
So I think the CME — in the interests of proving the liquidity benefits of HFT — should release the tape of the 100,000 or so trades that took place on the June e-mini contract on May 6, 2010 from 2:43 pm to 2:50 pm with the trader names removed, but identified as trader 1, trader 2, trader … etc. Ideally the tape should include orders that were outstanding at 2:43 and orders that were placed, but not filled. Of course the tape needs to include precise time tags down to the microsecond, if that is relevant to the sequencing of trades.
Note that the release of trade and order data (without any trader tags) for all exchange traded products should be considered the norm — after all these are public exchanges and none of this data should be in any way privileged. The information on traders is clearly a much more delicate matter — but if high frequency traders wish to defend their role in markets they will have to do so on the basis of data and not on the basis of unverifiable claims.
Failure to release this data for analysis will mean that HFT remains a complete black box and that the public is being asked once again to trust entrenched interests to tell the truth about what goes on in these markets.
Update: It’s very disturbing to realize that according to an SEC commissioner (via Alphaville) the SEC even now may not have this data set. Thus, the challenge for the SEC isn’t just one of analyzing millions of trades, but of gathering the raw data to analyze. Not good.