Institutional PFOF – Removing Conflict of Interest of Broker Routing

Over the past year(s), there has been much discussion around the topic of retail payment for order flow (PFOF). The issues have been well-documented elsewhere surrounding the practice of brokers monetizing retail flow, but there has been less discussion around institutional flow.

There are fundamental differences between retail and institutional flow, with the most obvious being the size of orders – institutional orders typically incur significant market impact that retail orders don’t. A second key difference is that institutions typically employ sophisticated transaction cost analysis (TCA) that facilitates discussions with their brokers on how to execute their orders. An institution with a lot of flow will pay a lot of commissions to their broker, which in turn leads to the broker paying attention to the way that their institutional customers want to interact with the market – typically these instructions arise from a Best Execution process that utilizes TCA to provide data-driven recommendations on trading behaviour.

There are many aspects of trading behaviour to consider, such as relative volume constraints and how aggressively to engage with the market (i.e., crossing the spread or booking passive orders). Very often the details of routing are an important input. Here is where the problem lies, though. A conflict of interest arises between the broker’s fiduciary duty of Best Execution and the access fees the broker must pay when interacting with the various trading venues that collectively make up the market.

Why is this? Because there are so many different trading venues, each venue differentiates itself through specific feature that include order types, and fee structures. As a result, there is a spectrum of fee structures available from maker/taker where liquidity providers are paid for executing passive orders through to inverted venues where liquidity takers are paid rebates by the trading venue. A key point is that since only brokers can access these trading venues directly (with some exceptions), it is the brokers, not their clients, that pay fees to the venues or receive rebates from the venues.

So where is the conflict of interest? When brokers act as agent (for example, when a broker executes a large institutional order) the broker charges a commission, which in North America is usually based on the number of shares executed. But the bottom line for the broker is the commission less any trading fees (or rebates) incurred during the execution of their client’s order. So, a broker will be incented economically to execute orders on trading venues that minimize their fees (maximize their rebates). This behaviour may be at odds with the broker’s duty of Best Execution for their clients’ orders.

For example, when executing a VWAP order, a broker may post shares on maker/taker venues that pay a rebate for passive executions – and they may choose the venue that pays the highest rebate. They may route their active orders to inverted venues that pay to remove passive liquidity. While this activity can optimize a brokers fees paid to trading venues, it doesn’t help the client who may be better served by accessing venues with higher fees. In this way a conflict of interest is introduced between the client’s desire to access expensive liquidity and a broker’s desire to optimize fees.

There is an easy solution. To remove this conflict of interest (whether real or ‘optical’), the broker can simply pass back the incurred fees, along with, but separately from the commission. In this way, a client that wanted to be ‘paid’ for providing liquidity could do so by instructing their broker to post on maker/taker venues, and take from inverted venues, while a client that was more concerned about accessing liquidity more quickly could be willing to pay the trading fees to access liquidity. This approach provides more control to the client and removes the uncertainty of variable trading fees that a broker would typically have.

A similar type of practice has been in place for some time in Canada. When an institution is trading an inter-listed security, such as Blackberry, their broker can access liquidity in Canada and liquidity in the US (since inter-listed stocks are fungible). For a client that specifies a CAD fill for an order, the broker can often route to the US to tap into a deeper liquidity pool and often access better prices, and then, once the execution in the US venue is complete, convert the fill into CAD via an FX trade to convert from USD to CAD. The trade done in the US is reported back to the client in CAD. Because this activity is done at the fill level, no changes to back-end accounting systems need to be made. This is a key point that allows all unit holders of a fund to benefit.

In an analogous approach, by incorporating the trading fees into the execution price provided to the client, issues surrounding netting and changes in beneficial ownership of internal accounts at a buyside institution can be avoided. There is an added benefit. When a buyside firm is trading on a net basis across multiple internal strategies, some of which are more active than others, instead of strategies that incur low trading fees subsidizing trading strategies that pay high trading fess, the trading fees can be fairly allocated. So, we not only address the broker’s routing conflict of interest, but we also solve for the internal conflict of interest of properly allocating fees across internal strategies.

The other main benefit of this approach is that it provides a market driven solution rather than a regulatory response. Institutions can choose between brokers that provide the option to pass on fees and rebates and those that don’t offer the option. It also allows buyside institutions to monetize their liquidity by developing strategies that maximize rebates. Alternatively, it can allow buyside institutions to pay to access expensive liquidity that their broker may not otherwise be inclined to access. The brokers that do offer this option would be better able to manage their commissions and focus on executing based on client instructions thereby removing the routing conflicts of interest that exist today.

In conclusion, we have described a market-driven approach to remove the conflicts of interest that brokers have when executing institutional orders in a market that consists of trading venues offering different trading access fees – this is institutional payment for order flow (PFOF).

Brokers can remove a source of uncertainty with respect to profitability of their agency business and can execute on similar terms for clients that trade in very different ways, since the brokers now can focus on client trading instructions and ignore trading fees. The approach also addresses internal conflicts of interest at buy-side firms that trade on a net basis for strategies that incur different trading fees on average.

Ultimately, this approach leads to buy-side institutions being able to take more control of their flow and the way it interacts with the market allowing for them to better manage their Best Execution requirements. By analyzing results using TCA, the buy-side firms can ensure that they provide Best Execution to their stakeholders.