TL;DR The National Best Bid and Offer is not sufficiently sensitive to time differences to be able to distinguish differences in price from lags in communication; this fundamentally limits the effectiveness of enforcing regulation designed to prevent ‘trade-through’ type market manipulation.

DeFi and crypto marketplaces, in virtue of being relatively nascent compared to their traditional and centralized counterparts, have considerable room to mature, and it's inspiring to think about the ways in which these markets might mature beyond the world of traditional finance. In the history of traditional financial markets, attempts at moving towards a more equitable, fair and transparent set of industry standards, have almost always resulted in something of an arms race between regulators and industry participants.

In 2005 the SEC passed legislation describing the National Market System (Reg NMS). The NMS sought to increase transparency in US equities markets and improve both market fairness and efficiency. The regulation includes a component sometimes referred to as “The Order protection rule”, or Rule 611. This is in addition to rules aimed at improving access to market data and demanding decimalization of price quotes. Traditional equities in the US, brokers are required to route orders to the exchange offering the best price, per Rule 611. 

Understanding the order protection rule and its enforceability, or lack thereof, requires an understanding of the National Best Bid and Offer (NBBO) and how it threads into defining a market manipulation practice known as a trade-through. Rule 611 is regulation designed to prevent trade-through type practices, where a trade is executed at a worse price than that available on some other exchange. A trade-through might occur where a brokerage pads an order in order to take a profit and in so doing offers the trader a worse price than they could have benefited from. Whether or not a trade-through has occurred is defined, from a regulatory perspective, by whether or not the execution price differs from the NBBO. Where NBBO tracks the highest bid price and the lowest ask price for a given security, taking data from all available markets and exchanges. The NBBO is the benchmark for deciding whether or not a brokerage has routed an order to an exchange that wasn’t offering the best available price. 

The challenge in enforcing proper behavior from brokers comes from the fact that small variations in price might be the result of lagging price signals from various exchanges, or from a brokerage manipulating the market. Without being able to reliably distinguish between the two, the markets remain a wild west in sufficiently small periods of elapsed time. 

A handful of uniquely well equipped actors effectively impose a 0.5 basis point tax on all trades executed by any other market participants. Extremely fast High Frequency Trading (HFT), with signals being communicated within 5-10 millionths of a second, known as “latency arbitrage” can account for approximately 20% of trading volume in traditional equity markets. Further, there is evidence that a handful of firms (~6) win ~80% of the races. FCA board members and the Chicago Booth school of business reckon there could be as much as $5 billion at stake here. 

As access to crypto markets continues to open up to an ever broader set of industry participants, and hard problems like private key management drive users into the comforting, password-recovery capable, arms of centralized solutions for exchange we can expect similar sorts of rent-seeking nonsense by privileged parties in crypto markets. Unless we design a system that prevents it…