In markets for goods and services, dealers display price quotes in order to attract customers away from other dealers. A car buyer, for example, needs to know which dealers offer which prices in order to route or match his order to the most competitive dealer.
In a security market where liquidity is offered through a consolidated book of orders, display is not important for order routing. The market’s electronics and rules can ensure thatconsumers of liquidity are matched with the liquidity providers who offer the best prices. Nevertheless, most security markets give preference to displayed over hidden orders. Many markets do not allow fully hidden orders, but instead require that a portion of such orders are displayed (as reserve or “iceberg” orders). Displayed orders are also commonly allowed to step ahead of existing hidden orders at the same price..
Some recent research argues that allowing liquidity providers to use fully hidden orders enhances the quality of a security market. This suggests that discriminating against hidden orders is unnecessary and potentially harmful.
Offering preference to displayed orders seems appealing because, in classic microstructure models, the informed exploit their informational advantage by hiding among the uninformed. By handicapping hidden orders, informed traders are placed at a disadvantage in exploiting private information, which limits the losses suffered by the uninformed. In the classic models, the informed trade as liquidity demanders.
Boulatov and George (2011) and Moinas (2011) show that this intuition is incomplete if the informed can trade as liquidity providers. As potential dealers, the informed are drawn into liquidity provision by the prospect of capturing the bid-ask spread. However, if their orders are displayed, they hold back to preserve their informational advantage. Some may not trade as liquidity providers at all. Alternatively, if orders can be hidden, more of the informed are attracted into providing liquidity and they trade more aggressively. Allowing orders to be fully hidden therefore intensifies competition among liquidity providers. When the informed can trade as liquidity providers, hiding orders has two effects on market quality—it preserves informational advantages and it intensifies competition among liquidity providers. Which effect prevails?
Boulatov and George show that the intensity of competition dominates. In their model, the informed can trade either as liquidity providers or as liquidity demanders. Allowing hidden orders draws the informed into providing liquidity, and the intensity of their trading yields profits that are actually lower in equilibrium than if they had traded as liquidity demanders. They are better off resisting the pull to capture the spread by trading as liquidity providers. But their incentives are similar to those of cartel members who expand output beyond the level that maximizes profit for the cartel as a whole. Trading aggressively as a liquidity provider is privately optimal for each informed trader, but profit is lower than if they could coordinate to resist. Resisting is what happens when orders are displayed because the threat of having their informational advantage expropriated by display causes traders to hold back.
Market quality is enhanced by allowing hidden orders because competition narrows spreads and incorporates more private information into the price schedule. The spirit of Moinas’s conclusion is similar. In her model, the informed trader is constrained to providing liquidity, but he can refrain from trading. She shows that the market is more resilient to the extent of adverse selection (and remains open over a greater range of parameter values) when hidden orders are allowed than when they are not.
The weakness of the simple intuition is in not accounting fully for how giving preference to displayed orders changes the behavior of the informed. Informed traders can effectively hide their orders by demanding liquidity rather than supplying it. Discriminating against hidden orders causes traders to exit liquidity provision in favor of using market orders, which reduces the intensity of competition among liquidity providers. This in turn widens spreads and decreases the information incorporated into quoted prices.
This perspective might also be useful for thinking about how constraints on high frequency trading could affect market quality. A trader can effectively hide by canceling limit orders that do not execute the instant after submission. This keeps his strategy out of the order book.1 Kirilenko, Kyle, Samadi and Tuzun’s (2011) study of high-frequency traders (HFTs) suggests that HFTs are able to anticipate the direction of order flow a tick or two ahead. Being allowed to add and remove orders quickly enables them to provide liquidity without so obviously revealing their private information about order flow.
The intuition of Boulatov and George and Moinas applied to this setting, suggests that intensifying competition dominates the advantage of hiding in how high frequency trading affects market quality. Indeed, the results of Kirilenko, Kyle, Samadi and Tuzun indicate that even during the flash crash, only about 20% of the orders of HFTs in their sample were aggressive (marketable sells as prices dropped). The vast majority provided liquidity by bidding to buy as prices dropped.
1 See Chester Spatt’s comments in Does High Speed Trading Hurt the Small Investor? Wall Street Journal, October 10, 2011.
These models are highly stylized. Nevertheless, they suggest that innovations in trading technologies that seem to favor the informed may actually enhance market quality. Allowing orders to be hidden enables the informed better to exploit their informational advantages, but it also induces them to compete more aggressively as liquidity providers. Policies that are designed to limit traders’ ability to hide orders, and possibly also policies that place constraints on high frequency trading, run the risk of impeding competition and impairing market quality.
Boulatov, Alex and George, Thomas J., 2011, Hidden and Displayed Liquidity in Securities Markets with Informed Liquidity Providers. SSRN abstract=1923946
Kirilenko, Andrei A., Kyle, Albert S., Samadi, Mehrdad and Tuzun, Tugkan, 2011, The Flash Crash: The Impact of High Frequency Trading on an Electronic Market. SSRN abstract=1686004
Moinas, Sophie, 2011, Hidden Limit Orders and Liquidity in Order Driven Markets, University of Toulouse working paper.