I am a firm believer in capitalism and free markets; but I also believe there is a role for regulators. I prefer to see regulators take a backseat to competition and natural market forces. But regulators can only do that when institutional structures are designed to reduce conflicts of interest and skewed incentives to the greatest extent possible. We cannot design or regulate a perfect world, but we can put structures and rules in place to allow the market and competition to pursue that naturally.
The demand on US regulators’ time is high, and their jobs are difficult. They are nervous about making changes to US equity market structure and demand conclusive evidence that demonstrates such changes are going to be effective and beneficial. Unfortunately, this demand is untenable. In a complex system such as the US equity markets, it is nearly impossible to come up with conclusive evidence without testing new rules extensively via pilot programs.
By looking at the predominant business model for US Exchanges, based on the maker-taker pricing scheme, the struggle between natural market forces and the impact of regulators is apparent. In this system, those who post orders on exchanges (declaring, for example, that they are willing to buy 100 shares of Google for $1,000) are called market makers. Generally, institutional or retail orders that come through and hit that posted order (that is, somebody that wants to trade immediately and sells 100 shares of Google for $1,000) are called takers. Under the maker-taker system, the taker pays a fee (which is capped by regulators), and the maker is issued a rebate (which is de-facto capped, as exchanges generally don’t offer a rebate higher than the take fee). Exchanges make their money on the differential between the rebate they pay out and the fee they receive.
It’s not difficult to see why, without regulatory intervention, this model becomes the predominant structure. With maker-taker in play, any exchange that does not offer significant rebates does not attract resting orders (i.e., liquidity). Therefore, when someone comes along who wants to trade, they are forced to pay higher taker fees because most of the orders are resting on venues with the highest rebates.
This structure is not dissimilar to something called Payment for Order Flow (PFOF), in which a wholesaler (such as Knight Getco or Citadel) will pay a retail broker for its order flow and then either internalize that flow or send it along to the broader market. In the case of PFOF, it is very profitable for the wholesalers to buy retail order flow and lucrative for brokers to sell this flow – so much so that it is generally referred to as “uninformed” or even “juicy” order flow. These terms describe the willingness of retail traders to pay the spread and the associated fees without a short-term view as to what will happen to the stock price over the next few milliseconds and seconds because of their long-term investment outlook or for swing/day-trading. For wholesalers, the profits from trading against this flow more than offset the fees that they pay to the retail brokers.
However, the PFOF model does a disservice to the market at large, as removing this order flow can theoretically inhibit price discovery, discouraging market makers from posting orders if they know that the only orders that make it to the exchange are the so-called “toxic” or “professional” flow. Internalizers and wholesalers are free-riding the public quote, and in essence “queue jumping” (a loaded term, to be sure) orders on lit exchanges that may have price-time priority from an absolute perspective.
A fascinating new study by the University of Notre Dame quantitatively demonstrates the problem inherent in the maker-taker model – an issue referred to as the principal-agent problem. In a healthy economic scenario, a broker should act as an agent of the investor and should pass along fees and rebates, which ensures that incentives for the broker and investor are aligned.
However, for both retail and institutional brokers, this is ordinarily not the case. They both generally charge the same fees, regardless of whether they capture rebates or pay fees to exchanges. It is, therefore, in their interests to route “cost-effectively” to maximize their profits. This often means trying to capture the highest rebates when routing limit orders, or internalizing/selling to wholesalers when routing marketable orders. Even if they are passing along fees/rebates, they may not be incentivized to route in their customers’ best interests in order to hit volume tiers at certain venues and reduce their firms’ cost of trading there.
The Notre Dame study offers “strong evidence that venues with high take fees (liquidity rebates) offer inferior limit order execution quality.” The authors present substantial evidence that order routing decisions are governed by the fee schedules rather than execution quality, making the following conclusions:
- “Limit orders resting on venues with high take fees require more time to fill than those on venues with lower take fees.”
- “For take fee differences exceeding $0.0001 per share, the lower take fee venue has higher measured limit order execution quality.”
- “The decision to route the bulk of one’s limit orders to a single venue offering the highest liquidity rebate is inconsistent with a broker’s fiduciary responsibility to obtain best execution.”
- “Inverted venues have shorter queues and are at least as likely to receive marketable orders as the traditional venues.”
- “Several large, national brokerage[s] are making order routing decisions that appear to be consistent with the goal of maximizing order flow rebates. … Proprietary data suggests this type of order routing results in lower fill rates and increased adverse selection costs.”
So why can’t we count on the market and competitive forces to fix these issues? Because maker-taker creates a fundamental conflict of interest and a puts exchanges between a rock and a hard place. It makes no difference to an exchange’s revenue what its rebates/fees are – as the exchange makes money on the difference between the two. However, exchanges are unable to lower their rebates and fees because doing so would chase away resting orders to those exchanges that do not lower their rebates.
This is why regulatory intervention is required, and the Notre Dame study quantitatively shows that. In order to move forward, this intervention should be in the form of a pilot study – as advocated by Royal Bank of Canada (RBC) and others – that eliminates rebates in a set of symbols, and requires the trade-at rule in those names, as Canada and Australia do. This rule simply states that in order to execute a trade of less than 5,000 shares off-exchange (in a dark pool or internalization system), there must be significant price improvement (at least a tick, or half a tick if the spread is a tick wide). High-frequency trading (HFT) proponents explain, quite reasonably, that maker rebates are necessary to compensate them for adverse selection (as explained earlier, only the “toxic” flow actually makes it to exchanges now). The trade-at rule will reduce adverse selection and improve price discovery on public venues, ultimately benefitting all investors and providing compensation to market makers.
Jeff Sprecher, the CEO of the Intercontinental Exchange and the new CEO of the NYSE, agrees and urges regulators not to “allow us to pay for order flow.” I completely agree!