Understanding Passive Rebate Arbitrage

Question

I was reading a BMO paper which offered the following example of passive rebate arbitrage:

"For example, if BBD.b is trading at $4.71 - $4.72 with multiple players on each side, one might place a 'passive' order on TSX to buy 5000 shares at $4.71 Upon receiving a fill, then proceed to place a 'passive' offering on another marketplace that does not have a bid at $4.71, thus locking the market:

Bid                         Offer
TSX 5,000 $4.71           $4.71 5,000 Pure
TSX 10,000 $4.70          $4.72 15,0000 TSX

Counting on one of the bidders on the TSX to move their order to Pure and buying their stock. This achieves buying and selling at 4.71 and collecting passive rebates.

If I understand correctly this would only work without Reg NMS? So you place a passive bid in TSX, lets say. And then as you get filled, you place a corresponding best offer at the same price level in a different exchange. This locks the market and without Reg NMS in place you could be getting filled in both places? Is this one of the reasons why Reg NMS bans locked markets? And also, if a market is locked with Reg NMS - is the only way of it unlocking by a match happening in the same venue or by a price slide?

Answer 1

You are correct that this cannot happen under America's RegNMS. Of course, TSX is in Canada. This "strategy" also makes generous assumptions about the likelihood of getting filled on passive orders. I doubt the author of this paper ever had much success with this.

If a passive order were to lock the market in RegNMS, then that order is usually routed to the exchange with the best price for an immediate fill. This information is then disseminated as a trade.

However, if the order contains special instructions to only provide liquidity at that particular venue, then the order is either hidden or slid.