Pragma urges order routing clarity

Buy-side firms should demand higher levels of transparency from their brokers over which venues they are executing on, according to a new report by trading technology provider Pragma.

Buy-side firms should demand higher levels of transparency from their brokers over which venues they are executing on, according to a new report by trading technology provider Pragma.

Pragma, which specialises in providing trading algorithms, studied US exchanges that offer rebates for taking liquidity (the taker-maker model), instead of the usual rebates for posting liquidity (maker-taker).

Paying to post liquidity effectively allows market participants to 'hop the queue', or gain execution priority over orders posted at other exchanges that are more expensive for the taker. Hopping the queue could provide an increased probability of fill before an adverse price move, with a resulting improvement in execution shortfall, Pragma found.

Order routers are programmed to route to venues that offer rebates for taking liquidity – the research showed that Nasdaq BX, which pays a rebate to liquidity takers, received 35% of orders when it showed the best offer – roughly ten times its overall market share. In other words, it is easier to get a fill on a taker-maker venue that offers best price than it is on almost any other venue.

“Aggressive traders can often choose between several venues that are offering the same price,” said Eran Fishler, director of research at Pragma. “It makes sense that in such a situation, they will route to the cheapest venue for them.”

But this only works in certain situations. Overall, the total cost to post on maker-taker and taker-maker exchanges is almost identical – meaning that a buy-side firm would have to be highly selective about when to route to a taker-maker exchange to gain a significant benefit from the technique.

Pragma developed a pricing model that can work out the value of ‘hopping the queue’, identifying opportunities when the value of hopping the queue exceeds the explicit cost of doing so. The model discovered that the higher the volatility, the higher the incentive to hop the queue, as waiting even a short period of time without a fill might result in a large adverse price move.

The upshot is that decisions such as where limit orders are posted can have a significant effect on execution quality. Buy-side firms should take that into account when dealing with their brokers, insists Fishler.

“Institutional clients should understand how their brokers approach the tradeoff between the client’s shortfall and the broker’s execution expenses,” he said. “An algorithmic broker who has no policy and has done no research may be optimising their own economics at an unknown cost to the client in execution quality.”

«