Someone with a Dungeons and Dragons background must have named them “dark pools.” But they’re neither mysterious nor scary. Dark pools are just off-exchange automated trading networks for stocks.
They exist for two reasons:
–the old school method of having a trader in a money management firm call up a broker and place a buy or sell order by phone is expensive. And money managers have a legal obligation to obtain the lowest cost execution of their orders on behalf of clients. So they have a positive obligation to seek out cheaper ways of doing business–which automated networks are.
–brokerage house traders won’t keep a money manager’s order secret unless the manager is exceptionally diligent. This is a real hassle, and very time-consuming for the money manager’s trading room. But if you don’t pay extraordinary attention, your secret trading plans–which, after all, are your stock in trade–will be all over Wall Street in a nanosecond.
Automated trading networks have one–no, make that two–defects:
–they can be relatively illiquid, so that very large positions may not be able to be moved quickly, and
–many of the biggest of them are run by investment banks/brokerage houses.
This second characteristic is the reason for the current SEC investigation.
In a recent post, I wrote that Fidelity was exploring the possibility of forming its own automated trading network with other money managers, cutting out brokers altogether. Its reason, I thought, was that computer=based high frequency traders were able to deduce Fidelity’s trading plans by analyzing dark pool data–and that Fidelity wanted to create a venue where they’d be banned.
It appears I may have been too high-tech in my approach. The SEC investigation appears to focus on two possibilities, both of which are decidedly old-school brokerage behavior and both of which would violate the guarantees the automated network operators give to clients:
–the first is that the operators may have taken undisclosed fees from high-speed traders to allow their buys and sells to have priority over other order–essentially letting them front-run or scalp other participants
–the second is that operators may have taken the supposedly anonymous trading activity of high-profile participants and sold its details to others. I say “sold” but in my experience, the compensation for such information would normally not be in cash but either in increased trading volume or higher per-trade fees.
Personally, I don’t think dark pools themselves are the issue. I view them as part of the solution to a problem with how traditional brokerage/investment banks are run. And the fact that the old system is breaking down makes these firms even less willing than normal to give clients an even shake.
It will be interesting to see how the SEC investigation progresses.