Is the ongoing market data debate much ado about nothing?
Market data has become a battle ground between exchanges, who generate data, and investment managers and brokers, who claim they are forced to buy proprietary data to meet a requirement to demonstrate they are getting the best deals for their customers.
“It is clear that the fundamental processes of the buy-side and sell-side are becoming more data-driven and there is no question that the value and necessity of data is growing,” said Jack Miller, Head of Trading at Baird in New York. “But the market for such services is as much a function of the health of the consumers as it is for the value proposition of the suppliers. In this regard the exchanges have to walk a delicate balance insofar as they can leverage their value proposition without inflicting damage on a client base that is feeling the pinch on both the revenue and expense side of the equation.”
Under the current model, exchanges from the giant NYSE down to start up IEX, sell their own data products or share in the profits from the government-mandated Securities Information Processor (SIP) distribution of market data. And while the reasonableness of the revenue or profit generated from market data is in the eye of the beholder, the market framework, Miller explained, needs to achieve a balance where competitive forces can keep the prices of such services in check.
Nasdaq, which has cited its Global Information Services unit as a chief driver of growth, has weighed in on the discussion with an effort to combat misconceptions around its data business. In a bid to be more transparent, Nasdaq shared some revenue data not previously publicly disclosed in an article published in late September. The hope is that its candid disclosure will prompt other exchanges to step forward with similar transparency.
“We want to engage with stakeholders on market data, but we want to engage with the facts,” said Jeff Kimsey, Vice President and Head of Data Products for Nasdaq’s Global Information Services. ”Our markets are too important to base our discussions on hyperbole or misconceptions. We took the step of disclosing meaningful revenue data in the interest of transparency and to put proper perspective on this important issue.”
Kimsey noted that stock exchanges are one of the smallest expense factors in the financial ecosystem, as pointed out by Nasdaq Chief Economist Phil Mackintosh in his article, The Big Picture on the Data Debate. Mackintosh noted total exchange costs to the industry and investors is collectively about $1.5 billion per year, a tiny fraction compared to the $50 billion each year generated in investment advisory fees and management fees, and the $31.8 billion the top five investment banks make annually in equity trading revenues, according to a Trefis study highlighted in Forbes this summer. Overall, exchange fees amount to approximately 1.18% of total investor costs.
He added that the cost of market data fees available from Nasdaq have not risen as much as some would like to think. “For U.S. equity market data products offered by Nasdaq, the compound annual growth rate (CAGR) of revenue associated with price changes from 2009 through 2018 is 2.4%, or 1.8% when adjusted for inflation. That’s a far cry from 14% claimed by some.
Overall revenue growth (CAGR) across all of Nasdaq’s U.S. equity market data products, from 2009 through 2018 was 6.6%.”
Kimsey said equity exchange data fees paid to Nasdaq since 2009 post an average annual price change of only 2.4%. That, he noted, basically mirrors the rate of inflation during the same period.
He also sought to debunk the theory that brokers ae forced to buy proprietary data products. “Our data shows the following rates of traction on the part of Nasdaq proprietary data products: Only nine percent of our clients take data from Nasdaq directly; most recipients of Nasdaq data pay an intermediary to receive it. Just one third—33%—of our U.S. equity data clients take depth-of-book data. And the percentage of our equity data clients who use colocation services in the U.S. tops out at 27%. These are hardly majorities.”
Spencer Mindlin, capital markets analyst at Aite Group, said that data fees aren’t so much of an issue as many make would like it to be but rather just a good incendiary talking point.
“The reality is market data fees is a relatively small cost to an end investor,” Mindlin began. “Generally, the debate around market data is between exchanges and their customers, with the regulators stuck in the middle. For sure, larger firms can more easily carry costs; smaller firms will feel them more. But it’s important to consider the all-in costs to transact on an exchange in order to be intellectually honest about the impact of exchange fees.”
Mindlin said exchanges could move the debate if they were to provide more justification on their overhead and what informs their price setting decisions. “The fact that exchanges, notably international exchanges, have such dramatic differences in prices is telling about how opaque the market for market data is – and it’s rarely discussed,” Mindlin said. “Many of these exchanges seem to price data arbitrarily, meaning it is unclear what actual operational overhead informs these pricing decisions.”
Jim Angel, adjunct professor at the McDonough School of Business at Georgetown University the market data issue is extremely thorny and will likely be debated for some time.
“We have been arguing over market data for over a century, and we will continue to argue over it for the next century,” Angel said. “Data is the new oil. Good data are essential for fair and orderly markets. The problem is how we define property rights in data. Right now, we have nationalized the top of book and last sale into Stalinist collectives known as NMS plans. We clearly need to improve the governance there.”
There is no obvious good solution, Angel said. The data business is one with very high fixed costs and very low marginal costs, and that leads to what he termed, ”weird economics.” Furthermore, the exchanges produce what economists call “joint products.” For example, similar to how a sheep produces both meat and wool, and exchange produces both trade matching and data.
“Any attempt at cost-based pricing quickly runs into the problem of how you allocate the high fixed costs. For the sheep, how much does it cost to produce the wool? How much of the feed should be allocated to the wool versus the meat? For the exchanges, how much of the cost of the data centers and the software belongs to the trade matching versus the data? There is no perfect way to determine costs,” Angel said.