January 26, 2005
Dear Securities and Exchange Commission:
In my capacity as CEO of Tradebot Systems, Inc., I would like to thank the commission for re-publishing Regulation NMS for public comment. Tradebot is one of the leading providers of limit orders to electronic market centers such as NASDAQ, Brut, Inet, Archipelago, Attain, and Track. We currently choose not to trade in markets that do not provide reliable automated executions. All of our trading is done on a principal basis, and so is free of the principal/agent conflicts cited throughout the release.
Overall, I support Reg. NMS as proposed by the commission because I believe it is better than the current outdated rules. Whatever the commission decides, I am quite confident that our trading systems can be modified to adapt to the new market structure. I would however, strongly urge the commission to reconsider two parts of the proposal: the trade through rule and the related prohibition against displaying an order that locks the market.
It is impossible to eliminate all apparent trade throughs because of the transmission delays between competing markets. Archipelago already has technology to try to prevent trade throughs, but the study done by the commission shows about 2% of Archipelago's trades are 'false positives'. Even if everyone uses the best available smart routing technology, there will be tens of thousands of trades per day that hit the consolidated tape out of order. How much time will be spent by regulators and SRO's investigating these 'false positives'? Even if you try to comply, how much will it cost to defend your order routing decisions against these accusations?
The economic cost to a limit order of sometimes getting unfairly traded through is less than the economic benefit of getting a lucky fill that you should not have received. For this reason, rather than increasing limit order production as sighted in the report, a trade through rule actually works to reduce the number of displayed limit orders. My firm is currently a large producer of near the market limit orders. I estimate that the new trade through rule will reduce my firm's production of displayed limit orders by about 20%. Using the questionable numbers cited in the proposal causes me to estimate that similar behavior across the industry would actually cost long term investors over $3 billion annually and benefit short term liquidity providers by a similar amount.
There is another effect that will come into play that is even worse. Rather than being forced to route an order to an undesirable market center, many automated participants will just send non-displayed price improving orders or refrain from price improving until the slowest automated quote is removed out of the way. Because these orders are non-displayed, internalized orders and orders in other markets will trade through these invisible prices that by regulation, are not allowed to be displayed. Of course, this is good for specialists, internalizers, and short-term traders, but the costs to long-term participants make this horrible public policy.
Having a trade through rule also makes the securities markets much more vulnerable to inevitable technical glitches. The proposal tries to address this with the self-help remedy, but does not go far enough. What if the technical problem is on our end or in the transmission line between us? Are we allowed to continue trading? What if their matching engine has already crashed four times this week and left our orders in unknown states? Do we have to route there the moment they claim to be back on-line? The current OTC rules provide disincentives to market centers with unreliable technology - namely, people avoid trading there. Being forced to route to the best automated quote, even if it is not working very well, eliminates this helpful feedback mechanism. Ironically, the proposal works to reward the market centers with the worst technology.
Incidentally, with manual quotes, there is a general consensus that the trade though rule currently benefits the NYSE on listed stocks. It is not widely understood yet, but I think a trade through rule with automated quotes would have just the opposite effect of increasing fragmentation, hurting the larger market centers, and helping the very smallest market centers.
Sincerely,
David Cummings
Chief Executive Officer
Tradebot Systems, Inc.