With buybacks, asking the right questions can reduce trade execution costs.
Imagine paying the cost of a company’s investor relations department with savings generated by executing the firm’s share repurchase program more effectively.
That claim was made during the NeuGroup’s recent Treasurers’ Group of Thirty (T30) meeting in a session focusing on corporate buybacks and the impact of today’s market structure on the large equity trades supporting them. And although it couldn’t be independently verified, the issue of share-execution quality has become ever more relevant as major companies prioritize buybacks. In fact, S&P 500 buyback volume in the second quarter was $134.4 billion, down a bit from the first quarter but still at a historically high level, and session leaders anticipated the trend continuing well into 2016.
Scrutinizing brokers’ trading strategies in today’s complex trading environment is necessity for corporates. Another claim made during the session was that more scrutiny and asking the right questions can save corporates up to a penny per share. That’s also difficult to verify, although pension funds and other financial institutional investors huge sums doing just such analysis.
Had corporates applied more scrutiny in the second quarter, they potentially could have saved $13.4 million—not insignificant—if they repurchased their stock at an average of $100 per share. And yet treasurers attending the session were unfamiliar with the “dark pools,” a key element of current equity market structure and factor in executing large trades. More than 35 such trading venues that have emerged in the last 15 years, and they now execute nearly a third of shares bought and sold. Their unfamiliarity is unsurprising given treasury executives have tended to rely their brokers to execute trades, freeing them to attend to their broad plate of responsibilities.
The treasurer of a global retailer in the Fortune 500 said in a phone interview that he had not considered the trade-execution issue until reading Michael Lewis’s Flash Boys: A Wall Street Revolt, which was published in 2014 and describes the rise of high-frequency trading (HFT).
“If you said HFT in 2006, no one would know what you were talking about,” the treasurer said, adding that “suddenly you had this new industry making all this revenue, which had to come from somewhere.”
Twenty years ago, block trades were mostly done by phone, and brokers took home a fat spread. Electronic trading picked up in the late 1990s with the emergence of a handful of new electronic-trading platforms competing against the exchanges, and then really revved up when regulators permitted quoting in pennies instead of eighths of a dollar. When Regulation National Market System (Reg NMS) went into effect in 2007, requiring brokers to find best execution for clients, even if it required routing orders to other execution venues, the equity market literally exploded into a multitude of trading platforms, each with its own rules and pricing.
Along the way, algorithm-driven trading firms have emerged that execute trades in microseconds ahead of competitors. The heightened competition has dramatically reduced equity trading costs.
“If you’re talking about Apple or some other highly liquid stock, you can trade instantaneously for a spread that’s next to nothing,” said Spencer Mindlin, an analyst at consultancy Aite Group.
Mr. Mindlin noted that spreads and trading costs widen with less liquid stocks, and there are arguments that high speed trading and a highly fragmented market may lead to market instability–the 2010 Flash Crash is the poster child event, but there have been many others. The growing complexity of equity market structure can also lead to unnecessary trading costs as brokers may seek to maximize their profits while still finding the best execution price as required by Reg NMS.
The treasurer acknowledged that execution costs today are indeed far less than even 15 years ago, but he added that the speed and complexity of today’s market may unnecessarily increase not only trade-execution costs but share-price volatility. “I’m tracking this more carefully today, and asking my brokers more questions,” he said.
Corporate treasuries tend to lack the resources of other institutional investors, but it nevertheless behooves staff to ascertain a base level of knowledge about how the market currently functions.
“Anyone executing a buyback program or working with a broker to do so needs to invest the time and resources to make sure they have the right information, the metrics and reports, and that they have an appreciation for how to gauge best execution,” Mr. Mindlin said.
That probably means working with a broker or vendor to come up with performance metrics to determine benchmarks to meet over different time periods. The more urgently large trades must be executed, the more likely some HFT firms will connect trading activity in a stock to a corporate’s pre-announced buyback program, enabling them adjust their orders around the bid/offer spread at high speeds, in ways that can increase costs for the corporate. HFTs pursuing latency-arbitrage strategy, for example, use sophisticated technology to detect a broker entering a large order in one trading venue that will likely end up seeking liquidity in another venue; by quickly buying up the liquidity in the second venue, it can sell it back to the broker at a higher price.
Varying the timing and size of trades can throw HFTs off the scent, but brokers using third-party trade-execution tools may not have the controls to mix it up, or the order-routing technology of those with proprietary order-routing technology may be too predictable.
“That’s the classic challenge of using a broker who doesn’t have its own proprietary trading technology, or has prioritized its own routing technology to reduce trading costs,” said Craig Viani, vp, market structure and technology, at Greenwich Associates.
The treasurer of the consumer-goods company said he started asking the company’s stable of banks about their order-routing technology a few quarters ago. “Some use best of breed algos from several banks, and some have only one, but at least it’s now a part of the discussion,” he said.
The treasurer said it was reassuring when one of his company’s brokers acknowledged routing trades to dark pool IEX, a rival of its own dark pool. IEX touts institutional investor-friendly technology that blunts the advantage of speed, and it was portrayed as a potential solution in the Lewis book. Brokers are obligated to provide customers with their trade-execution data, so their claims can be double-checked.
Although brokers are required to find best execution, generally defined as best price, they may post bids and offers in high-rebate execution venues or even their own dark pools to wait for a match, and while that may reduce the broker’s costs it can result in the large orders taking longer to fill and a lost-opportunity cost for the issuer.
To minimize that cost, treasury executives using several brokers can require those brokers to provide their hit and fill rates in the venues the customer’s orders were routed to. If a broker continually routes more orders to a venue with a 4 percent hit ratio over a venue with, say, a 40 percent hit ratio, then that broker may be playing the waiting game, overlooking opportunities elsewhere in order to trade more at venues with greater financial incentives.
In addition, treasury executives can ask their brokers for the overall average price at which the company’s shares were purchased across all the trading venues compared to the average price at the broker’s dark pool.
“If the overall average price was $10.05 but the broker’s pool prices averaged $10.35, the company can question the broker’s decision to post orders in that pool,” Mr. Viani said, adding, “Those kinds of questions, though, have been more likely to come the head trader at Fidelity Investments than a corporate CFO.”