An alternative to PFOF: on-exchange trading incentives with inherent price competition

We discuss the PAX zero-fee + cash-back model and compare it to payment-for-order-flow (“PFOF”). PFOF has drawn criticism for potential conflicts of interest between brokers and their clients. PAX avoids this issue altogether by removing the middlemen and introducing price competition.

Share this post

PAX is a new exchange with a new fee structure: we sell instant order placement via the PAX λ API. All other trades, regardless of order type – both making and taking – cost nothing and get cash-back. Today, zero-cost trading is only available through retail brokerages, but PAX is changing that. Because PAX technology See our blog post: PAX λs, where we describe PAX "co-lo on a chip" technology. helps reduce market making risk, PAX can go to zero Industry jargon for offering zero-fee trading. Of course, at PAX, it is zero-fee plus cash-back. for all market participants – i.e. for retail and institutional, alike.

The rise of zero-cost trading

From the time of DTCC DTCC today comprises the DTC and NSCC, established in the late 1960s to replace paper stock certificates with electronic records. and NASDAQ to now, electronic trading has driven new market efficiencies. In 2013, Robinhood introduced zero-commission trading for retail investors, financing its business by selling order flow to systematic internalizers, Market participants that algorithmically execute trades (mostly) against their own inventory of assets, i.e. the trades are executed internally. a practice known as “payment for order flow” (PFOF). People Are Worried About Payment for Order Flow by Matt Levine.

Six years later in 2019, perhaps because of Robinhood, Schwab also went to zero. In the meantime, nascent innovators such as WeBull came to market. Today, retail traders expect to trade at zero-cost.

Never heard of PFOF? Let’s talk credit cards

Some readers may not be familiar with PFOF, much less the debate surrounding PFOF and its role in the retail brokerage industry. We think that most readers will be familiar with credit cards.

It’s likely that you have a credit card, probably one that gives you points or cash back. How does this work though? When you swipe your card, the merchant pays a transaction fee. The fee they pay is shared with the payment processing industry and also with you. That is how points and cash back are funded. Merchants may not like paying the fee, but they derive significant value nonetheless: evidently, merchants do more business by offering payment by credit card. One might wonder: without credit cards, would prices be lower and would we all be better off? Although impossible to settle this question directly, intuitively, in a world without credit cards, we would have less economic activity and that the elimination of transaction fees would be more than offset by that reduction in economic activity.

As a business, PAX is like a credit card for institutions and for individual traders alike: PAX reduces the friction to trading by offering zero-cost market access with cash-back on every trade. This, in turn, is good for markets through increased trading volumes and liquidity.

The PAX incentive model

At PAX, speed sensitive traders can opt to use the fee-based PAX λ API. The fees paid by our PAX λ customers are shared with all other participants at PAX. This is why we say PAX offers zero-fee + cash-back trading: anyone can trade at PAX using standard industry APIs or our web app for free and experience cash-back on every trade, regardless of whether their order is characterized as “making” or “taking.” Most exchanges offer a less attractive incentive using the so-called "maker/taker" fee structure. In this model, passive quotes (orders that "make" the market) are given a rebate (i.e. are paid) and marketable orders, that cross the spread, are charged a fee. PAX pays a rebate for both order types.

If anyone can trade at PAX, for free, why then when would market makers choose to pay to trade via the λ API? Market makers derive two significant benefits:

  1. Speed. Simply put, speed is essential to market making strategies. Speed provides front-of-line queue priority and speed enables market makers to cancel quotes the very instant price appears to be moving against their bid or offer. Importantly, the capability to cancel reduces the risk of adverse selection. We discuss adverse selection in detail below.
  2. Access to order flow. Market makers want to trade at venues with high trading volumes, and we expect PAX incentives to drive significant volume to the venue.

At PAX, market makers pay for speed – i.e. for access to market moving information as it happens PAX innovates on speed in two different dimensions. The PAX exchange offers the fastest possible tick-to-trade latency through its λ API and it offers the fastest publication of market data, of any exchange, anywhere. – and for access to a high volume venue.

The retail price competition SEC rule proposal

PFOF has been criticized because there appears to be a conflict of interest. This criticism asserts that the internalizers, i.e. the recipients of PFOF, do not provide the best prices on the orders that they fill.

This assertion – that retail traders receive sub-optimal prices based on PFOF – is significantly nuanced. The returns on seeking best-execution are asymptotically diminishing and internalization has a long history and precedent. In the U.S. regulated financial space, retail is gated from the backend system through middlemen. The fact that middlemen are, by law, required effectively means that retail traders can never access the "best price." Even though one could feasibly route retail orders directly to NMS exchanges or ATSs, the regulatory inertia makes radical change like this unlikely. This seems to be a point in favor of crypto, i.e. direct market access is open to everyone in crypto. The price slippage experienced by retail as a result of PFOF is exceedingly small, and alternatives exist, but those are not obviously significantly better What Does Payment for Order Flow Buy? by Matt Levine. and in certain cases worse. Some Free Brokers Are Cheaper Than Others by Matt Levine.

Nevertheless, as a result of these criticisms, the SEC under Gary Gensler This rule proposal faces stiff headwinds and may not get adopted, especially now that we have a new SEC chair coming on board. proposed a rule change requiring internalizers hold open auctions for retail orders, i.e. such that any market participant Technically, any broker dealer with direct access to an NMS exchange, so not actually any market participant, just any other industry incumbent. could compete to fill the order.

We do not endorse proscriptive remedies to the perceived deficiencies of PFOF in the U.S. retail brokerage space. On the other hand, we strongly advocate for market driven outcomes, and PAX zero-fee + cash-back offers a better alternative.

What about adverse selection?

The existence of a PFOF relationship between retail brokers and systematic internalizers and the apparent lack of such a relationship This is nuanced, see discussion below about payment for market access. between institutional traders and broker dealers is generally cited to be a result of adverse selection. Also People Are Worried About Payment for Order Flow by Matt Levine.

Adverse selection happens when a market maker transacts with a market participant that has a more informed view of the market that is unavailable to the market maker. Usually, this puts the market maker on the wrong side of the trade.

Adverse selection can occur in two major modes:

  1. Incremental, as HFT market takers transact with mis-priced quotes, and
  2. Significant and immediate price dislocation, also known as a market sweep.

When internalizers make markets for retail order flows, the direct risk of adverse selection is low. In general, retail traders are not considered to have more informed views of the market than market makers. On the public and open markets, market makers are directly exposed to adverse selection risk, but they still pay handsomely for access to those markets. By co-locating, and paying for market data feeds.

In the retail brokerage PFOF model, internalizers are fully insulated from latency arbitrage but remain exposed to large price dislocations. The PAX λ API creates a similar situation, in that market makers are insulated from latency arbitrage – in other words, at PAX, market makers can directly incentivize their counterparties because the PAX λ API helps mitigate their risk of adverse selection.

Zero-fee + cash-back: an improved market structure

Incentive systems can change entire industries. Today, retail traders expect zero fees and most people with credit cards expect points or cash-back. PAX combines the best of both worlds - we offer zero-cost + cash-back on every trade for all market participants, funded by opt-in use of our "instant" order placement λ API.

In comparison to retail PFOF, the PAX incentive model also benefits institutional trading and, importantly, features open price competition. At PAX, the problem of adverse selection is significantly mitigated because makers can instantly cancel mispriced quotes using our λ API. Thus, PAX offers an improved market structure that will lead to improved price competition and liquidity.