What Flash Crash Report Will — and Won't — Say

The Flash Crash report is imminent. The good news: it will finally be out, likely tomorrow. The bad news: the SEC is unlikely to point to a single "smoking gun" as the cause of the decline, and the laundry list of immediate "action items" is likely to be fairly short.

These include:

1) Tweaking of the single stock circuit breakers — they will likely adopt a "limit up-limit down" approach, which will allow stocks to train within bands when certain limits are hit, rather than being halted outright;

2) Elimination of stub quotes: market makers will not be allowed to fulfill their obligations to make markets by simply putting in a place marker bid of $0.01. They will be required to make markets within bands that are determined by the price of the stock.

3) A consolidated audit trail that will allow the SEC to quickly access recent trades — including all bids and offers. This will allow much quicker analysis of trading patterns.

4) banning of naked access, which allows high-frequency traders to to use a broker's "membership card" to get to the exchange a fraction of a second earlier. The downside: it's not clear who exactly is making the trades, since the HFT is using the brokers' "pass" to access the markets, and checks to make sure that trades are meeting parameters set by the exchanges are not being made.

5) increased regulatory scrutiny of high-frequency traders. While some are already registered with the SEC because they are market makers, most traders expect the SEC to seek to have ALL high-frequency traders above a certain volume threshold register with the SEC and be subject to risk compliance checks.

I also expect somewhat vague discussion on:

1) the impact of the growing "unlit market" (i.e., dark pools and internalizers who do not expose their bids and offers to the outside trading world);

2) limiting the number of exchanges and trading venues;

3) creating some kind of tax on high-frequency traders who put in large numbers of bids and offers that are never executed;

4) how fast is fast enough, that is, whether some minimum quote duration (1 second? One-tenth of a second?) should be required between the time an order is posted and the time it is executed.

For more, see our Man versus Machine section.

Bookmark CNBC Data Pages:



Questions? Comments? tradertalk@cnbc.com