PrefBlog

This is a very, very long post. Forgive me. It’s as much for my own reference for a future article as it is for you guys!

Pegged Orders were last discussed on PrefBlog in the post Brown & Holden on Pegged Limit Orders.

For those who came in late, a pegged order is defined as:

Marketplaces have also introduced market pegged orders (also referred to as reference priced orders) that are priced and re-priced to a reference price such as the national best bid (offer) or a marketplace’s best bid (offer). One type of market pegged order is the primary peg order. A primary peg order is a visible order that is automatically priced (and then subsequently re-priced as necessary) to equal either the best bid, in the case of a buy, or the best offer in the case of a sell.
…
Marketplaces have introduced other types of market pegged orders. These include market pegged orders that are fully-hidden and market pegged orders that either peg to a price above or below the NBBO [National Best Bid & Offer], or are eligible to execute at the mid-point of the NBBO.

I have taken the view that pegged orders will be a godsend to non-professional preferred share traders and will increase the liquidity of the Canadian preferred share market.

The OSC has published comments on the paper. Most commentators are more interested in the Dark Pool part of the paper than Pegged Orders, but we’ll just look at the responses to Questions 16-19, which are:

Questions relating to market pegged orders

Question 16: Please comment on the actual or potential impact if any, of market
pegged orders on:
a) Price discovery
b) Fairness

Question 17: Although this paper has not specifically addressed pegged orders that execute at the mid point of the NBBO, in your view, should market pegged orders be allowed to execute at prices unavailable to transparent orders (e.g. at a price between the bid and the ask when the spread is a single trading increment)?

Question 18: Although this paper has not specifically addressed pegged orders that are fully-hidden, in your view are there any issues that arise due to fully hidden market pegged orders?

Question 19: Are there other issues that should be considered with regard to market pegged orders?

Unfortunately, the OSC has merely scanned many of the responses into PDFs, without running them through an Optical Character Reader (or, even better, indicating a preference for electronic responses in the first place, as is the practice in civilized countries). So … let’s just say that some of the following will be paraphrases, and before getting angry at a particular comment, check the original! Then get angry!

TD Asset Management: At present there is no clear evidence supporting any material impact of market pegged orders on price discovery or fairness in the marketplace.

Investment Industry Association of Canada: Good question! But what’s with all this public consultation bullshit? Let’s discuss it privately, with a fat per-diem.

Investment Counsel Association of Canada:

Questions Relating to Market Pegged Orders:

Question 16: Please comment on the actual or potential impact if any, of market pegged orders on:
a) Price discovery
In a truly Dark Pool, market pegged orders should have no impact on price discovery.
b) Fairness
From the point of view of fairness, pegged orders not as fair. They are able to take advantage of the visible orders (free ride), without making a contribution.

Question 17: Although this paper has not specifically addressed pegged orders that execute at the mid point of the NBBO, in your view, should market pegged orders be allowed to execute at prices unavailable to transparent orders (e.g. at a price between the bid and the ask when the spread is a single trading Increment)?

Among our members, there were mixed opinions on this issue. There are pros and cons of allowing market pegged orders to be able to execute at prices unavailable to transparent orders. On the one hand, there is the opportunity for price improvement, no matter how small. However, there is the counter argument that allowing market pegged orders to execute at prices that are not available to the lit market is unfair, and does not provide the right incentives.

Question 18: Although this paper has not specifically addressed pegged orders that are fully-hidden, in your view are there any issues that arise due to fully-hidden market pegged orders?
We have no comment on this question.

What a thoroughly ridiculous response. Thank God I’m not a member of the ICAC.

National Bank Financial:

[16] a) Price discovery Pegged orders are already being used today by dealers. They are either created by the front end trading systems or via dedicated algorithmic trading systems. As such, it shouldn’t matter if a market offers this as a value added order type. Price discovery would be the same as if the order was pegged by a dealer’s system.

b) Fairness
As long as the market based pegged orders are made available to all participants of the marketplace, fairness is maintained. Fairness is not impacted if even if a market chooses to charge a premium for these orders. It is up to the dealer and trader to decide whether they want to pay for this service. Other markets may not charge for the same feature and users are therefore free to use that venues pegged orders instead. The choice is held by the dealer. The advantage to the market based pegged order is its proximity to its own market data and order entry which makes it quicker than external sources. This will allow the market based pegged order a better opportunity at capturing priority at the new price level (more so if an aggressive peg strategy is permitted). However this advantage is limited to the market itself, assuming peg updates must adhere to the trade through obligation, routed out orders would be subject to the same forces. To reinforce this point, all market based orders should be forced to meet the same UMIR obligations that dealer and vendor generated order must.

[17] Market pegged orders should not be allowed to execute at prices not available to other order types. This is already mandated under UMIR 6.1 where the rule and policy basically states that all orders should be entered at full tick increments unless specifically exempted under UMIR. Such exemptions include Basis Order, Call Market Order or a VWAP Order, where they may execute at the prescribed increment established by the marketplace of execution.

[18] All dark orders should be pegged to the NBBO and not allowed to trade outside of this. Ultimately it will come down to an order’s trading limit, and as long as the order is between the NBBO and at or better than the limit, the trade should be allowed to take place. The alternative is force all pegged order updates (either fully hidden or visible) to pass through a SOR [Smart Order Router].

In a single marketplace environment regulators must often consider what is “fair”, and in such a situation, we would agree that this would not be “fair” to most people. However we find ourselves in a multi-market place environment where dealers have the choice as to where they want to send their orders. They can therefore “vote” on fairness with their orders, relieving the regulator from doing so. If an exemption is to be made for Dark Pools and their order types, the CSA should decide what the minimum trading increment can be on a dark market (ie half the existing trading increment[ .005 on a visible .01 increment], or 10% of the increment [ .001 on a visible .01 increment]. Either way there should be a min increment specified, and whether a dark order can take place AT the NBBO, instead of inside. If trades can occur at sub penny increments, orders should also be allowed to be entered at this sub penny increment, subject to the min trading increment (suggested 10% of visible trading increment).

Highstreet Asset Management: Should be a minimum size requirement for dark pegged orders that can only interact with other dark orders.

Greystone Management Investments: Not enough experience to comment

TMX:

Marketplaces should not be permitted to offer visible market pegged orders. Marketplaces that simply regenerate orders from their competitors are in essence counterfeit quoting. The effect of counterfeit quoting is the same as all other counterfeiting schemes where ultimately the value of the original currency is diluted. In the case of market pegged orders, the value of the original displayed order is diluted.

Marketplace visible pegged orders have a negative impact on market structure because they result in significant messaging increases that place unnecessary strain on marketplace and regulatory infrastructure. Visible pegging results in increased messaging because the price of the visible market pegged orders must be changed (and therefore the pegged order is cancelled and rebooked) each time the primary price changes. These repeated price changes result in extremely large message traffic not only to downstream users of the marketplace data, but they also greatly increase the message traffic in the marketplace’s regulatory feed received by IIROC. This large strain on infrastructure is not warranted given that the visible market pegged order is offering no new value to the market’s price discovery process.

Further, any marketplace that offers visible market pegged orders will have an unfair informational advantage and functional advantage over competing dealer/vendor pegging systems. This informational advantage is achieved when the marketplace receives a quote changing order and then updates its own visible pegged orders before publishing the quote change. This practice allows the marketplace to update its own book before allowing any other updates to ensure that its pegged order stays at the front of the queue.

Dark pegged orders do not have the same negative market impact, as there are no external messages that are derived from the changing quote on a dark pegged order. Dark pegged orders can bring additional liquidity to visible marketplaces and enable visible marketplaces to, among other things, provide customer executions at the mid-point between the national best bid and offer. Market pegged orders in Dark Pools and dark pegged orders on displayed marketplaces should be permitted since these do not unfairly compete with broker pegging systems and Dark Orders do not generate the high externality (messaging) costs associated with displayed marketplace pegging.

[17] The CSA and IIROC should ensure a fair and even application of regulation across visible and dark venues regarding trade execution prices. As we have asserted in our response to Question 9, Dark Pools have an unfair advantage in that they can print fractional price executions, while fractional prices cannot be posted on visible marketplaces. If the CSA continues to permit Dark Pools to exist and allows visible marketplaces to offer Dark Orders, then it would seem inconsistent and inappropriate to allow trades to execute at differing execution increments between a visible and a dark marketplace.

[18] As stated above, we believe that fully-hidden market pegged orders do not raise the same concerns as visible market pegged orders. We believe that visible marketplaces, particularly exchanges, can serve a useful function by offering dark market pegged orders to their customers, while providing fully disclosed rule sets that allow all participants to utilize these dark order types. With the adequate rule set in place, dark market pegged orders will not take priority away from displayed orders.

Liquidnet: Mid-peg orders are great and everybody should buy our product! You can get sub-increments on pricing with a little jiggery-pokery on the pricing, so what’s the problem with putting the actual order in that way?

Newedge Canada: Huh?

RBC: Pegged orders are great! Sub-incrementally priced orders are also great! Dark pegged orders are great!

Omega ATS:

There are two fairness issues with visible pegged orders, and both are free rider problems of a sort.

Firstly, visible pegged orders subvert the traditional expectation that price discovery is rewarded with executions. On a single marketplace, the first to post liquidity at a price will be the rewarded with the first fill. So a second trader who simply follows the lead of the first trader by posting a quote at the same price level has to wait in line for an execution.

In a multiple marketplace trading scenario, where time priority does not exist across marketplaces, a trader attempting to “copy” the price discovery generated on another marketplace has to contend with the risk of latency delays between watching the quote move and then acting accordingly. But visible pegged orders reduce the latency risk and permit traders to add liquidity at a price level in competition with the quotes that originally generated the price discovery. Pegged orders become an effective way for a trader that does not wish to post an at-the-market quote in a principal marketplace (where time priority forces him to wait in line) to instead post at-the-market quotes on another marketplace and so have a better chance of participating in “active” (market/marketable limit order) flow.

Secondly, the marketplace hosting the visible pegged order type is flooding other marketplaces and dealers’ order-execution systems with new market data messages every time the reference price moves. The message-to-execution ratio of a visible pegged order is significantly higher for visible pegged order types than any other order type. This taxes the overall market data communications systems in a manner disproportionate to the price discovery provided by visible pegged orders.

In traditional auction markets, no auction house would permit a trader to peg to the best bid/offer of other traders, or to “penny” the best bid/best offer by pegging one price increment higher/lower – this would be allowing one set of clients to beggar the price discovery generated by other clients.

However, in a multiple marketplace context, we would expect new marketplace entrants would be willing to host this order type in order to steal away business from the dominant incumbent.

To address the first fairness issue, we have the following recommendation to the CSA and IIROC: allow traders to fully discipline pegged order “free riding”. We have heard anecdotal reports that pegged order types are not widely used in the United States because savvy traders can quickly spot a trading strategy utilizing pegged orders and so trade against it. By using pegged orders, a trader signals to the market that he will move his quotes up or down in response to the actions of others; a savvy trader can therefore cause the pegged trader to move his quote up and down and then execute against the pegged trader at inferior prices (for the pegged trader). We understand from IIROC that a strategy of using a trader’s pegged orders against him in this fashion could be viewed as manipulative trading contrary to UMIR 2.2. We would recommend, to the contrary, that IIROC allow any trader to capitalize on the information gleaned from observing a pegged order trading strategy. In situations where a proportionally small number of traders are pegging at the best bid/best offer in a highly liquid stock, there will be little obvious free riding that can be traded against by a savvy trader. However, where a trader is “pennying” the best bid/best offer or is pegging to quotes in an illiquid stock – in precisely those situations where the free riding is at its most obvious and so most egregious – then a savvy trader ought to be permitted to trade against this activity.

We believe that the second issue described above can be dealt with commercially. That is, if one marketplace is inordinately taxing the “common grid” of market data communications, the other marketplaces can be expected to charge a higher price for their market data used to support pegged order types.

[18] As a preliminary point, a fully hidden order must be pegged at all times to the best bid/best offer or else it risks executing outside the best bid/best offer as the visible market moves. For example, assume the market in a stock is $4.80 x $4.82 and a fully hidden offer is booked at $4.81. A bid for $4.82 (marked bypass) executes against all offered volume at that price and books the remainder of the bid. The market has now moved to $4.82 x $4.83. The next incoming bid for $4.83 (assuming it is not marked by-pass) will match against the hidden offer at $4.81, with any balance executing against the visible $4.83 offer. The $4.81 execution would technically constitute a trade-through under the Order Protection Rule (the crossed market exemption being unavailable because the $4.81 price was hidden). To avoid this scenario, the hidden order must be pegged to the best bid/best offer “dynamically” (that is, not only at the moment of order entry but continuously as long as the order is posted and available for execution).

Unlike visible pegged orders, the cost of maintaining this facility is borne entirely by the hosting marketplace – i.e., it does not flood other marketplaces with market data messages as the pegged order moves up or down. Moreover, the “free riding” that occurs off others’ price discovery is less egregious than is the case with visible pegged orders precisely because the quote is hidden and so doesn’t compete directly for active orders.

[19] As articulated above, there can be concerns associated with messaging increases derived from marketplace pegging. The CSA must ensure that dealers, vendors, and regulators do not become swamped by large increases in messaging traffic due to marketplace pegging. We submit that it may be useful for the CSA and IIROC to study, and perform quantitative analysis on, marketplace pegging to determine how these order types are used (by whom and for what purpose) and whether investors are disadvantaged by these order types. We also suggest studying alternatives to marketplace pegging, such as vendor pegging systems and dealer proprietary pegging systems, to determine what the impact of these systems is on the market as a whole.

What a great answer, with respect to UMIR 2.2! Imagine, allowing smart traders allowed to punish dumb ones without going to jail! Nirvana!

Penson:

Pegged and visible on the NBBO adds to price discovery … Pegged and visible on the NBBO is fair for all participants to see

[17] A fair marketplace should apply tighter restrictions around dark orders, not more lax as priority needs to be given to the visible orders. A transparent order should be able to compete in the same space as dark, between the NBBO, if executions are being allowed within this space.

[18] Transparency in a visible market may result in higher investor confidence and understanding of what is actually happening to an order when it is placed in the market and can help in the trading decision if all information is available.

RBC Asset Management: Sorry, not interested.

ITG Canada:

Even though primary pegged orders are “free-riding” on the contribution of limit orders, the underlying principle of what can be done manually, should be able to be done electronically, applies in this circumstance. The pegged orders should be displayed in the quote as part of the total volume available at that price point to provide price and volume discovery…Limit orders posted at the same price as pegged orders should get priority in all circumstances. The pricing of pegged orders should be at the full increment to ensure that they cannot step ahead of protected limit orders without narrowing the quote.

[17] As noted above a pegged order should trade at the full increment so in the case of single increment spread the pegged order should be required to queue behind all the displayed bids/offers or be required to immediately cross the spread.

[18] As noted above with Dark Orders, we believe that a portion of the order should be displayed when entered on a transparent market that displays protected quotes.

[19] None at this time.

Canadian Security Traders Association: Both buy-side and sell-side believe allowing primary pegged orders to execute ahead of limit orders is unfair. However, the buy-side things that otherwise pegged orders are good, while the sell side disdains the free-riding.

CIBC:

[16a] Market pegged orders should have no adverse effects on price discovery, provided that at least some portion of the order is visible if it is entered on a visible market with a discretionary price or that it is priced deterministically. Pegged orders simply automate functionality that occurs in traders minds, on their desks, and on order and execution management systems. Pegging has been termed a “free riding” on the price discovery mechanism strategy. The reality is that this “free ride” comes at the cost of potentially putting in effort without a fill, if the order does not have priority because it is at or on the outside of the quote. Pegging is a more risk-averse way to trade alongside the evolving spread.

[16b] In terms of fairness, market pegged orders are simply a logical evolution of algorithms which react to quote changes and penny other participants. Market pegging simply allows a participant to hand over the pegging function to the exchange that can achieve the same result faster. This is a fair evolution as it is available to all participants. If anything, this levels the playing field for firms unable to make the expenditures required to compete on speed and technology. The only negative impact arises when fully hidden discretionary pegged orders are permitted on lit markets. As discussed in question 10(a), we believe that if they are permitted they will enjoy an unfair advantage over the same orders placed on dark markets; discretionary, fully hidden inside pegs allow for automated and systematic free riding to occur since rules will force others to interact with them.

[17] We believe that market pegged orders should be permitted to execute in sub-tick increments, even when these prices are not available to be quoted by transparent orders. We note, however, that although transparent orders cannot carry sub-tick limit prices today, they can execute in sub-tick increments under UMIR6 so it is our view that sub-tick executions are in fact available to transparent orders as well. When a transparent order hits a mid-point peg limit, it can fill at a sub-tick price and enjoy the same price improvement as the hidden order. If two parties are willing to transact subject to the rules of a dark pool (or dark order type on a lit market) at the current quote, then the midpoint result could not be fairer. Each side pays half the spread. If orders were not allowed to transact at the midpoint of the NBBO, quotes with a one tick spread would face restricted trading options. We are seeing one-tick spreads more consistently, on more symbols as the Canadian market evolves.

Prohibiting sub-tick executions for deterministically priced matches would compromise the viability of dark pools; it would not be possible to provide price improvement on Canada’s most liquid symbols, and without the prospect of price improvement dark pools become unattractive to a large share of order flow on liquid names.

Consequently we believe that all orders should be permitted to execute at sub-tick prices, but that visible quotations should continue to be limited to full-tick increments. Transparent orders should continue to have an equal opportunity for a mid-point fill, but valid reasons to prohibit sub-tick quotations on visible markets remain; simplicity, orderliness of markets, and reduced gaming because getting ahead in the queue by improving the quote in an insubstantial manner is not possible.

As discussed in question 10, there is a fundamental tradeoff between visible and hidden orders. Dark markets offer reduced transaction cost, and lit markets offer increased certainty of a fill and protection of orders in the allocation queue. Permitting sub-tick executions and prohibiting visible orders price in sub-tick increments is congruent with this trade-off. On a dark market, the opportunity costs of an execution are reduced by pegged orders that execute in sub-tick increments, enhancing the cost reduction benefits of trading dark. On a lit market, full-tick quotations increase the cost of getting ahead in the queue, enhancing the certainty benefits of holding a position in a visible book. The balance of interests between dark and lit markets is maintained.

[18] If dark pools are allowed to function and the logic of visible price-time priority over dark price-time priority holds, then fully hidden market pegged orders should be fully acceptable. Where such orders have a discretionary price they should be restricted to dark pools, and where they are deterministically priced they can be permitted on visible books as well.

[19] It is important to note that, in addition to automating trader behaviour pegged orders improve, to some degree, operating efficiency for all market participants. Pegging without a dedicated order type involves posting an order, then sending CFO after CFO in reaction to each change in the quote. With ever-accelerating markets, the message traffic this (common) strategy generates is staggering. That traffic has a cost. Marketplaces need throughput capacity, dealers need increasingly powerful data feed readers, and the networks that link them are increasingly burdened.

Alpha ATS: No problems with dark pegged orders. Could be problems with visible pegged orders if they have an advantage over other visible orders.

Instinet: [16a] Pegged orders are great!
[16b] How can an order type available to everybody possibly be unfair? Also, it’s only formalizing something available through algorithms anyway.
[17] Yes
[18] No
[19] No. We support the use of pegged order types.

Penson #2 [16 a&b] Pegged and visible is good!
[17] Priority needs to be given to visible orders.
[18] Transparency is good

Chi-X:

[16]To provide some perspective and context as to how pegged orders came about it’s helpful to go back in time. In the old days, investors had to call their dealer and ask for a revision to the price of a limit order to buy or sell stock each and every time they wanted to change the limit. The dealer would write up a new order ticket with the new price and call down to the exchange floor where a runner wrote up yet another ticket and took that to the floor broker who then added the order to his book until he found a willing counterparty. The process was manually intensive and inefficient. With the electronification of stock trading, investors are able to enter orders and send re-pricing instructions to their brokers electronically, or even directly to the markets via direct market access (DMA), significantly increasing the efficiency of the process. The importance placed on the ability to re-price an order in a timely and efficient basis has significantly increased over the past decade due to many events that have made trading faster and more automated.7 The introduction of competition and multiple markets has increased the likelihood of the mispricing risk8 facing limit orders, particularly for retail customers. Many of the technologically more sophisticated markets have responded to this growing issue by implementing “pegging” functionality to automate the re-pricing process using algorithms that react dynamically to changing market conditions. This removed or reduced many of the inherent risks and inefficiencies in limit order re-pricing.

In the U.S. reference priced orders (e.g. pegged orders), have been used without controversy for more than 10 years. There are currently at least nine displayed markets in the US that offer pegged orders so while it’s only recently that Canadian markets began offering this functionality, it is certainly not a recent innovation. Just as noteworthy, over the past decade brokers have developed algorithms that re-price limit orders in response the prevailing inside market. The most widely used algorithms, such as “VWAP, TWAP and Volume Participation” react to market data in very similar fashion to the pegging functionality offered by markets. In fact broker algorithms and the pegging functionality offered by markets have a symbiotic relationship in which pegging enables the broker algorithms to be much more efficient with its re-pricing logic. Like most other market center functionality, pegging is available without discrimination to any participant who wishes to use it. It is also important to note the intra-market priority of pegged orders. Typically pegged orders will sit behind visible limit orders at the same price. This ensures that limit orders responsible for generating the price of a pegged order are not disadvantaged and are appropriately rewarded for “price setting” (in contrast to broker-preferencing which allows the price setter to be disadvantaged, see response to question 8).

There have been independent studies noting that “pegging enhances price discovery and liquidity by reducing the mispricing risk and making limit orders more profitable, and as a result pegging increases the quantity of limit orders submitted”9. By continuously joining the NBBO pegged orders help create price discovery in at least two ways: 1) Displayed limit order quotes add to the price discovery process by increasing the number of data points on which the fair market value of a security is derived. The additional volume created at the NBBO enhances the supply and demand curve thus yielding a better approximation of fair market value. 2) When executed, these orders result in trades, which are disseminated through recognized market data vendors. In addition to displayed quotes that express investors’ pre-trade willingness to trade at a given price, the trade prints add to the price discovery process by showing the prices at which transactions are actually taking place.

The concern raised by those who either don’t fully grasp the benefits of pegged orders orthe inter-relationships between markets and their customers is that pegged orders are a form a “free riding” on limit orders posted on other markets thus violating inter-market price/time priority. In a multiple market environment investors should be empowered to decide where to route an order and not be beholden to inter-market price-time priority. This is a deliberate characteristic of a competitive market environment and crucial to achieving the maximum benefits from multiple markets. A rule set that supports intermarket price-time priority creates a totally centralized system that loses the benefits of vigorous competition among individual markets. The point of inter-market price-time priority and a “virtual CLOB” was addressed by the SEC in Regulation NMS:
“The essential characteristic of a CLOB is strict price/time priority. Such a facility would greatly reduce the opportunity for markets to compete by offering a variety of different trading services. Price priority alone, however, would not cause nearly as significant an impact on competition among markets because it allows price-matching by competing markets”.

If required to follow strict inter-market price-time priority participants would, given two or more markets with identical quotes, be forced to execute with a market because a quote appeared on that market first. If followed, this market mechanic would impair a participant’s ability to achieve Best Execution as it disregards the key factors that determine how participants prioritize which markets to route to when multiple markets have equally priced orders:
• Price – in the form of execution fees
• Performance – in terms of the market’s trading system speed and functionality
• Reliability – in terms of the market’s stability, knowledge of its staff, and support

Ideally an order will be routed to the market that offers the lowest execution fees, the highest performance, and the greatest reliability. The potential danger of inter-market price-time priority is that it does not penalize trading venues for charging higher fees and operating sub-par systems because participants would be forced to trade with them based solely on time priority.

The point of allowing competitive forces to determine the preferred trading centers of market participants was addressed by the Securities and Exchange Commission in Regulation NMS: “Market participants and intermediaries responsible for routing marketable orders, consistent with their desire to achieve the best price and their duty of best execution, will continue to rank trading centers according to the total range of services provided by those markets. Such services include cost, speed of response, sweep functionality, and a wide variety of complex order types. 242 The most competitive trading center will be the first choice for routing marketable orders, thereby enhancing the likelihood of execution for limit orders routed to that trading center. Because likelihood of execution is of such great importance to limit orders, routers of limit orders will be attracted to this preferred trading center. More limit orders will enhance the depth and liquidity offered by the preferred trading center, thereby increasing its attractiveness for marketable orders, and beginning the cycle all over again. Importantly, Rule 611 will not require that limit orders be routed to any particular market. Consequently, competitive forces will be fully operative to discipline markets that offer poor services to limit orders, such as limiting the extent to which limit orders can be cancelled in changing market conditions or providing slow speed. 243 Simply stated, a participant should have the ability to choose the venue that presents them with the highest probability of executing that order at the best price.

In the final analysis the argument that pegged orders are unfair has no more merit than an argument that email has an unfair advantage in transporting messages versus a mail courier that physically delivers letters. Of course email has an advantage! But it is strictly a competitive advantage that stems from superior technological capabilities that create efficiencies. To restrict email or in this case pegged orders, is tantamount to restricting progress by reducing the competitive landscape to the lowest common denominator and a Harrison Bergeron11-like equality based on handicapping the more capable, more sophisticated, and more efficient systems.

Pegged orders are part of a larger trend to improve the efficiency of the trading process by automating it. The automatic updating of pegged limit orders is a substitute for the human time and attention devoted to monitoring and updating regular limit orders.12 The fact that any market has invested in the technical capabilities and infrastructure to offer pegged orders may indeed create a competitive advantage that helps it succeed in attracting liquidity over its competitors, but that is purely a result of the fact that not all markets are created equal.

TD Securities: What a great question! Let’s discuss it in private!

CNSX:

[16a] Market pegged orders add nothing to price discovery and free ride on published limit orders. Similar to Dark Orders, as the use of pegged orders rises, the price discovery mechanism becomes less and less functional.
[16b] The Request For Comments cited the article “Pegged Orders: An Unfair Trade” by Jeffrey MacIntosh, which contains compelling arguments against the use of “parasitic” pegged orders. He describes the multiple market phenomenon of a virtual single market and argues that the effect of pegged orders negates the benefits that this phenomenon creates.

The reality is that while price priority can be preserved in a multi-market environment, time priority is market specific. Pegged orders allow a marketplace to use the disclosure of participants on another marketplace for commercial benefit. Both the second marketplace and the user of the pegged order are simply trying to use the best bids and offers displayed elsewhere without granting them time priority.

[17] No. If there is a benefit to investors to be able to trade at increments less than the current minimum trading increments, then the increments should be reduced on all marketplaces. Otherwise, to allow such executions appears to be more an effort to avoid transparency than to provide a benefit to investors. Such orders intercept order flow that would otherwise trade with posted visible orders, thus not only providing another disincentive to post orders in a visible market, but also creating a “best execution” incentive (which may be technically correct, but practically of little value given the narrowing of spreads) to use such pegged orders and thereby undermine market integrity.

[18] Fully-hidden pegged orders are no different than any other Dark Orders that use public prices as a reference. Any order that is placed on a light market should contribute to the price discovery mechanism, even if it is only present briefly.

[19] The argument that market participants can always create their own pegging methodology should not be the basis for allowing marketplaces to offer pegged orders. The dealer responsible for the order owes a best execution duty to the client and if a pegged order is seen to meet that, it is between the dealer and its clients. It should not be at the marketplace level, where the only interest in the order is trapping it to increase trading on that venue. This underscores the issue of where the right line is with respect to time priority and whether there is a resulting public good that warrants ignoring it: given today’s technology, we do not believe there is a benefit that warrants allowing such system-enforced mechanisms for circumventing time priority. Further, it is unclear whether the market data agreements in place among marketplaces in Canada permit the use of a price posted by another marketplace in establishing a reference price for pegged orders. At this point, two marketplaces have declared selfhelp against each other, based on this issue. Although there are associated proprietary rights concerns on the part of the originating marketplace, the continuing uncertainty could be addressed through a regulatory ban on the use of pegged order types.

Connor Clark & Lunn:

As a general comment to the questions in this section, our view is that market pegged orders are not materially different from Dark Orders, and thus they should have a similar impact on price discovery, liquidity, etc. As for fairness of such orders, if all participants are able to utilize such orders, we believe that the type of order is fair. That is, even if Pegged Orders are given priority ahead of displayed orders on an exchange or ATS, so long as all participants can choose between using Pegged Orders and displayed orders we believe the order types are fair.

BMO-Nesbitt:

[16] Pegged orders have long been offered by marketplaces and broker dealers. We see no reason to restrict trading venues from offering similar products.
[17] As stated above, we don’t believe that orders should be allowed to execute at sub-tick increments. Allowing such matching penalizes those participants that have placed visible bids and offers and endured the risks associated with such order placement. We strongly advocate that visible orders must be protected or we risk dis-incenting market participants from placing such orders.
[19]Not at this time.

Assiduous Readers will remember that the purpose of regulation is to obtain a competitive advantage for your firm by ensuring that the rules reward your strengths and minimize your weaknesses. Thus, I find the comment by the Canadian Security Traders Association to be very revealing: the buy-side wants it, the sell-side hates it. I suggest that this has more to do with the sell-side seeking to maintain a competitive advantage, exploiting their universal availability of technology, than any consideration of fairness or efficiency.

To my mind, those who decry Pegged Orders as parasitic have to explain why algorithmic trading, which can do (and often does) the same thing a fraction of a second slower, is not also parasitic. The similarity is pointed out several times in the above comments, but the question is not discussed by the order type’s opponents.

This entry was posted on Wednesday, February 24th, 2010 at 10:50 pm and is filed under Regulation. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.