Notice of Approval – Alpha Exchange Inc.

1. Introduction of New Order Books -- Alpha-X and Alpha DRK

1.

Two commenters were generally supportive of the proposed new order book. (BMO, CSTA) and one commenter was of the view that taken together, the Smart Limit order type, order processing delay and TMX QDS create a very favourable environment for providers of passive liquidity and is beneficial to institutional traders who are concerned about adverse selection. (CSTA)

Alpha thanks the commenters for their feedback.

2.

One commenter was supportive of the two new order books being explicitly separate books that do not interact with one another. (Nasdaq)

Alpha thanks the commenter for its feedback.

3.

Two commenters were unsupportive of the proposed order books (Scotia, FIA PTG) and one was of the view that novel order types (including the Smart Limit and Smart Peg order types, especially if the randomized order processing delay is preserved and potentially lengthened) can be implemented on the existing Alpha Exchange. (Scotia)

While there are similarities between the order types and trading functionalities on Alpha Exchange and the New Order Books, Alpha is of the view that there are differences that require the new order books to be implemented on a separate marketplace.

For example, Alpha currently uses an inverted fee model whereby rebates are offered to liquidity demanders and fees are charged to liquidity providers. The new Smart Peg and Smart Limit order types, which are intended to attract natural, liquidity-posting investors, require a make-take fee model whereby transaction rebates are offered to those who provide liquidity while charging customers who take that liquidity, and therefore presents a change to the inverted fee model currently on Alpha Exchange.

In addition, the proposed duration of the order processing delay on Alpha Exchange will be shorter than the order processing delay on the New Order Books, thus requiring a separate marketplace.

Lastly, Alpha is of the view that introducing a separate market for the new books provides a venue to provide novel and strategic solutions to improve execution quality for retail investors and other participants with slower execution speeds without negatively impacting Alpha Exchange.

4.

One commenter was of the view that the introduction of the new order books comes with certain considerations for the investment community:

As participants formulate ways in which to maximize their economic benefits when executing trades, while the new order books are available to all market participants, some market participants may choose not to execute trades on the New Order Books based on their trading strategies.

•

additional overhead costs;

•

further liquidity fragmentation; and

•

added complexity to the Canadian market structure landscape.

The commenter was of the view that the proliferation of new trading books in Canada continues to add to infrastructure, support and maintenance burdens for broker dealers that ultimately result in an increase in costs to the investment community. Further, creating new books may in fact hinder the adoption of new features because novel features on new marketplaces may require their own bootstrapping, instead of receiving the benefit of established order flow patterns and order handling practices that already touch existing marketplaces. (Scotia)

While the new order books constitute an addition to the multi-market Canadian equity trading ecosystem, given that the connection, order entry and market data workflows are the same for the existing Alpha market, Alpha is of the view that the effort required for existing Alpha participants choosing to execute trades on the New Order Books is an incremental one. Alpha believes that these factors may help simplify the incorporation of the new order books into clients' trading environments.

5.

One commenter was of the view that, although the new order books are on an unprotected marketplace, a broker must consider all venues as potential sources of liquidity in an effort to provide clients with the very best possible prices available. As such, unprotected markets require consideration and therefore add development work and complexity to routing decisions. (RBC)

While there are additional considerations, complexity and development work that may arise as a result of the New Order Books, Alpha is of the view that the overall benefits to the broader trading ecosystem as a whole outweigh these concerns and that the New Order Books do not pose any harm to traders.

Please also see our response to comment #5 above.

6.

One commenter questioned why Alpha is proposing to introduce two order books, independent of one another, that will result in connectivity costs incurred by Members. (Nasdaq)

Alpha points out that the two order books are accessible via a single new set of order entry sessions incorporated into existing session bundles at no extra cost.

7.

One commenter was of the view that the new order books favour the resting/liquidity provision side of trade, and in particular, that the new order types, as proposed, would benefit liquidity providers at a significant expense to liquidity takers, including retail investors. (RBC)

While the New Order Books, and more specifically Alpha DRK and the Smart Peg order type, may benefit liquidity providers, liquidity providers, as slower market participants are typically disadvantaged in other conventional marketplace structures which generally benefit liquidity takers. The New Order Books aim to provide a venue which democratizes the trading experience among faster and slower participants, and offer a platform where the natural investor can post liquidity with protections against adverse selection and more favourable markouts as a result.

We strive to create marketplaces that benefit and improve the Canadian capital markets as a whole, and ultimately democratize trading to ensure optimal results for the end investor. The advantages that some market participants have over others (including technological advantages) as well as their respective trading strategies are developed over time. Market participants will always formulate ways in which to maximize their economic benefits when executing trades, and will always gravitate towards strategies that meet that end goal. These strategies continuously change as new order types or marketplaces are introduced. In that context, while the new order books are available to all market participants, market participants may or may not choose to trade on the new order books as a result of decisions based on their trading strategies.

8.

Several commenters were supportive or appreciative of the new order books being on an unprotected marketplace (Nasdaq, FIA PTG, CSTA), and one commenter was of the view that as traders are free to choose their interactions with the new order books (or avoid them entirely), any potential harms of the proposed functionality are minimal. (CSTA)

Alpha thanks the commenters for their feedback.

9.

One commenter was unsupportive of the use of anonymous broker preferencing on Alpha-X, being a lit order book. (CSTA)

The Request for Comment contained an error with respect to the allocation of trades for Alpha-X. The correct allocation priority is as follows:

1.

Price

2.

Broker (excluding orders marked as anonymous)

3.

Time

This has been corrected and reflected in a correction to the Request for Comment which was published in the OSC Bulletin and on the Alpha website on March 13, 2023.

10.

One commenter was supportive of the allocation priority on Alpha DRK. (CSTA)

Alpha thanks the commenter for its feedback.

2. Introduction of New Order Types -- Smart Limit and Smart Peg Orders

11.

One commenter was unsupportive of the introduction of the two new order types and expressed concern about the negative impact of these order types on overall market quality. The commenter generally opposes exchange provided discretionary order pricing functionality. The commenter stated that in this case, the exchange assumes some of the order-handling, including price movement and best execution responsibilities historically left to the broker-dealer. (FIA PTG)

Alpha is of the view that the introduction of the new order types will have a positive impact on the markets and improve execution quality. We expect that new order types will offer protection from latency arbitrage, and may result in some participants receiving better pricing on their orders. It is expected that by using Smart Limit or Smart Peg orders, participants will gain confidence in their execution quality and be able to post larger sized orders. This may improve the depth of liquidity in the markets, have a positive impact on price discovery, and benefit the market as a whole.

The Smart Peg and Smart Limit order types do not alleviate, and market participants continue to be responsible for, best execution requirements under National Instrument 23-101 -- Trading Rules ("NI 23-101") and the Universal Market Integrity Rules ("UMIR").

Smart Limit Order Types

12.

Three commenters expressed concerns with, and were not supportive of, the priority that is maintained by the Smart Limit order as described in the Request for Comment (CSTA, Nasdaq, Scotia) and one commenter stated that this proposed feature is at odds with the principle that in order to encourage healthy liquidity provision that participants that are willing to take on the economic risk of exposing a quote to the market at a new price level should be rewarded with execution priority. (Nasdaq)

Based on feedback received, Alpha has withdrawn its proposed amendment regarding Smart Limit orders retaining their time priority when they are repriced as originally set out in the Request for Comment.

One commenter was of the view that the priority allocation proposed was unfair and that time priority should be reset at all times when the Smart Limit order type improves the tradeable limit price of an order. (Scotia)

Please see "Summary of the Final Amendments" above.

13.

One commenter was of the view that, while the Smart Limit order type may provide benefits to liquidity providing orders, liquidity takers, including retail clients, may not be able to receive such benefits and that the Smart Limit feature may have negative impacts on active/liquidity taking orders by facilitating quote fading. The commenter stated that adding more facilities to enable quote fade can cause negative outcome and confusion, resulting in a lack of investor confidence. The commenter stated that managing different latency between lit venues would require sophisticated technology that may not be available to all firms. (RBC)

Please see our response to comment #8 above.

In addition, Alpha is of the view the introduction of the Smart Limit and Smart Peg order types will have a positive impact on markets. We expect that these order types will offer protection from latency arbitrage, and may result in some members receiving better pricing on their orders. It is expected that by using Smart Limit or Smart Peg orders, participants will gain confidence in their execution quality and be able to post larger sized orders.

Lastly, given the static nature of the order processing delay between Alpha Exchange and Alpha-X, we believe that natural participants will be able, using existing technology already at play in our markets, to time orders in such a way so as to capture liquidity on all venues.

14.

One commenter expressed concerns over the potential for information leakage. The commenter was of the view that the most likely users of the Smart Limit order type are institutional traders who frequently trade large parent orders. Disproportionately high volume on Alpha-X or concentrated activity by a single participant on Alpha-X could signal other traders that a large institutional order is present and may lead to suboptimal prices. The commenter suggested that a potential remedy to help mitigate information leakage is to create a market-by-price display rather than a market-by-order display. (CSTA)

A participant using any order type is able to discern, or make an educated guess, regarding certain information about the contra side of the trade. Note that in any transaction between an active and a passive order, information is revealed in the form of the trade price. For example, in the case of a Smart Peg order trading at a discretionary price above its booked price, the trade price is the limit price of the active order. In conventional circumstances, the trade price is the price of the resting order where the active order is discerning the price of the resting passive order. In either case, the price of a previously undisclosed order is now known to the participant interacting with that order and happens in the normal course of trading as part of the price discovery process.

15.

One commenter was of the view that Alpha is blurring the line between marketplace and dealer by with discretionary re-pricing of Smart Peg orders to achieve best execution by Alpha Exchange rather than by a router or algo strategy being used to trade an order.

The Smart Peg and Smart Limit order types are intended to improve execution quality on Alpha DRK and Alpha-X, respectively. The Smart Peg and Smart Limit order types do not alleviate, and market participants continue to be responsible for, best execution requirements under NI 23-101 and UMIR.

The commenter was of the view that Alpha Exchange would take order execution discretion, without any requirement to adhere to best execution policies & procedures -- or accountability for failing to comply with a best execution standard, and that the dealer entering orders on Alpha-X would be subject to a best execution standard involving an order type they do not have a complete understanding of (i.e. the TMX QDS).

Dealers routinely use third party routers and algorithms that do not have best execution obligations. These routers and algorithms execute discretion in handling orders. The introduction of smart order types in no way introduces new best execution risks to the dealer community.

The commenter stated that in the traditional algorithmic trading context, dealers directly address issues surrounding algorithmic malfunctions or best executions concerns, and was of the view that this is not possible when the malfunction is at a marketplace, given the proprietary nature of TMX QDS and because Alpha Exchange is not required to be capitalized adequately to account for the possibility of malfunction.

As noted above, our experience with Alpha Exchange suggests that agency dealers are less well suited to take advantage of order processing delays on their own. As such, many global marketplaces have introduced similar features to help facilitate order management for natural investors.

The commenter acknowledges that there are times when a trading venue has some discretion to reprice a limit order, typically related to an auction, risk protection mechanism or other participant controlled order type, but was of the view that the Proposed Amendments would be the first time that a trading venue would be responsible for repricing to improve execution quality based on a prediction being made by the trading venue's controlled signal, without transparency into the logic involved. (Scotia)

Smart Peg Order Types

16.

One commenter expressed concern regarding the Smart Peg order feature that permits such order to provide price improvement inside the NBBO (up to the midpoint) only where a contraside order does not cross the spread. The commenter was of the view that this feature would allow a passive order the advantage of maximizing its execution price without taking the same risk as firm midpoint orders and in doing so, would disadvantage orders that are willing to cross the spread. Active orders would interact with any price improvement opportunities in between the spread on other dark books in the market. The commenter stated that this feature would result in inferior executions for institutional investors, who are expected to use these order types. (Nasdaq)

Please see our response to comment #14 above.

We also note that the priority of Smart Peg orders trading at discretionary prices falls behind that of orders trading at their booked prices. Therefore, orders with urgency will continue to be posted at the midpoint in Alpha DRK.

17.

One commenter was of the view that active retail traders and passive liquidity providers would be the biggest beneficiaries of Alpha DRK. (CSTA)

Alpha thanks the commenter for its feedback.

This is consistent with our approach on helping natural participants.

18.

One commenter was of the view that the Smart Peg order type features (i.e. offers the minimum amount of price improvement required to achieve a trade) is a departure from acceptable norms where limit prices contribute to price discovery by establishing the most aggressive price the investor is willing to pay, thus providing public signaling of the value of the securities being traded. The commenter stated that being able to peg to the near side while allowing the order type to at its discretion match at the midpoint will encourage participants to walk up (or down) the book to ultimately seek out limit order pricing, and is inefficient. Participants will also have to justify executions which are made at the discretionary price from a best execution perspective, adding to the burden of accessing the Alpha DRK market. The commenter was of the view that these elements, when combined, create for a more complex and potentially unfair marketplace. (Scotia)

With respect to walking up or down the book, Alpha notes that this is the case for any Dark market. Alpha is of the view that this does not impose an additional burden on dealers attempting to capture dark liquidity in a multiple market environment.

TMX Quote Decay Signal ("TMX QDS")

19.

One commenter was supportive of using public market data in the construction of the TMX QDS, however suggested that all lit books are used and not only data from TMX marketplaces. The commenter expressed concerns over fair access relating to the potential for bespoke versions of the TMX QDS. While the commenter did not object to marketplaces creating tools to help clients better manage order flow, it was of the view that these tools should be accessible to all and on fair terms.(CSTA)

The TMX QDS will consume only public market data. Our research indicates that good predictive results are obtained using data from TSX alone, thanks to TSX's large market share, and TSX's pricing structure leading to TSX being at or near the bottom of a typical liquidity-taking market sweep.

Please also see our response to Comment #5 above.

20.

Three commenters suggested that Alpha provide further public disclosure on the TMX QDS, including information on:

Alpha intends to publicly disclose a general overview of the model, and disclose the following information on the TMX QDS on a periodic basis:

•

the model design, accuracy and updates of the TMX QDS. (CSTA)

•

statistics on efficacy of the signal; and

•

the signal's construction (Scotia, BMO), such as if, and how it will change over time (and provide the market with 30 days prior notice), or details on its efficacy. (BMO)

•

revisions and upgrades to the signal (in advance of the release).

•

the machine learning model around suitability, model risk, and maintenance schedules. (Scotia)

We believe that this level of disclosure will provide market participants with the level of information needed to determine whether the Smart Limit order type is appropriate, and aiding, in their trading strategies.

Because of the proprietary nature of the TMX QDS signal, and concerns with intellectual property infringement and gaming of the TMX QDS, Alpha does not intend to publicly disclose specific details regarding the TMX QDS, including its input parameters.

3. Order Processing Delay -- Alpha-X and Alpha DRK

21.

Three commenters were unsupportive or questioned the duration of the 10ms order processing delay on the new order books and suggested that Alpha reconsider introducing an extended delay on these two new order books or consider further analysis. (FIA PTG, BMO, CSTA)

Based on the feedback received, as well as on further analysis conducted by Alpha, we are reducing the duration of the order processing delay on Alpha-X and Alpha DRK from 10 ms to 3 ms. After careful analysis, we believe that 3 ms will provide participants adequate time to manage their orders throughout the post-trade market movement, and minimize adverse selection by delaying incoming liquidity-seeking orders for that duration. We believe that the 3ms would not be controversial for market participants as 3 milliseconds is currently at the top end of our current randomized order processing delay on Alpha.

We will continue to analyze and monitor the duration of the order processing delay in order to determine whether we are still achieving our goal of execution quality. We may determine, from time to time, to amend the duration of the order processing delay on Alpha-X and/or Alpha DRK in future releases as provided for in the Alpha Rule Book.

4. Other Comments Received

22.

Several commenters were generally supportive of innovative market models and new marketplace features that provide participants with greater options for making trading decisions, (Nasdaq, RBC, CSTA, Scotia, BMO), and one commenter stated that while the Proposed Amendments may be complex in nature, that the cost of this additional complexity is worth the benefit and that there is not any significant harm to traders or to the broader trading ecosystem. (CSTA)

Alpha thanks the commenters for their feedback.

23.

One commenter requested that the proposed trading fee structure for Alpha-X and Alpha DRK be made public before regulatory approval is obtained (CSTA), and one commenter noted that the fee schedule, a critical element, has not yet been made public. (Scotia)

The amended Alpha Trading Fee Schedule reflecting the New Order Books will be published prior to the production launch date, and will be subject to regulatory approval. The Alpha-X and Alpha DRK fee structures will be make-take and take-take, respectively, with a premium charged for Smart Limit and Smart Peg orders.

24.

One commenter was supportive of Alpha not charging for market data at launch. (CSTA)

Alpha thanks the commenter for its feedback.