Remarks at the North American Electronic Bond Trading Forum

As prepared for delivery

Electronic Trading in the Bond Markets

Good morning. Thank you [Mark Monahan] for that introduction. I’m pleased to have this opportunity to join today’s conversation on the increasing automation of fixed income markets.

Many of us have been in the securities business long enough to appreciate how the first electronic trading system—the NASDAQ quotation system that was launched in 1971—helped transform trading in the U.S. equities market. On the equities side, strong investor demand for faster executions, lower trading costs and more complete market information through limit order books helped spur the movement toward electronic trading. Today, nearly all stock markets both here in the U.S. and abroad are electronic.

On the fixed income side, however, trading and markets are a dramatically different story, and the bottom line is electronic trading platforms have not yet been as transformative. The reasons are likely varied. One may well be the inherent value of voice trades in maintaining discretion in thinly traded markets. The other related reason often cited is the large number of unique assets and infrequent trading of many corporate bond markets. To put it in perspective, in 2012, the average daily trade frequency for equities was nearly 3,800 times, while the 13 most liquid investment-grade corporate bonds traded an average of 85 times per day. And, as of the end of last year, 37 percent of corporate bonds did not trade at all.

But we know from recent research that electronic trading in the corporate and municipal bond markets is gaining traction, like a slow train gathering speed. Since 2008, the number of electronic bond trading platforms in operation has more than tripled. Today, there are 19 platforms operating, up from 5 in 2008. In fact, seven of the 19 platforms entered the market in the last two years, and four more platforms are expected to launch this year. The platforms have evolved in response to the differing demands of participants and to increase participation.

While I can’t predict whether or when electronic trading will indeed transform bond trading the way it has transformed the equity markets, I can tell you that markets will continue to evolve. As a regulator, FINRA’s goal is to be nimble and responsive to innovations that have transformed the fixed income market structure and the process of price discovery. So I want to spend my time with you this morning exploring what factors seem to me to be contributing to the increased use of electronic trading in the fixed income markets, and what electronic platforms means for investors, market structure and transparency.

Where We Are Today

To put it in perspective, I want to begin with a look at the segments of the fixed income market where electronic trading has gained a foothold. Research tells us that electronic trading is dominant in the most liquid of these markets, namely the fixed income futures and the U.S. Treasury markets. In the fixed income futures market, nearly 90 percent of transactions occur electronically. And in the U.S. Treasury market, the shift to electronic trading started in the early 1990s with bond market dealers calling for greater market transparency. By the end of the decade, order-driven central limit order book platforms were widely used.

We're also beginning to see a general uptick in electronic trading in the corporate market. The number of corporate bond trading protocols that have been structured to address the differing demands of market participants has more than doubled since 2013, from 14 to 37. In fact, Professor Larry Harris from the University of Southern California Marshall School of Business found that more than 1700 corporate bonds had two-sided quotes on electronic platforms almost every day in the first quarter of this year. And the percentage of market volume in investment-grade corporate bonds being transacted on electronic platforms has more than doubled since 2013, with odd-lots and round-lots-trades between $100,000 and one to onepoint-five million-being the most typical trade sizes for electronic platforms.

All-to-all trading venues, where multiple parties from the buy side and sell side come together and quotes can be requested from several different parties electronically, are proliferating, with 26 protocols on 14 platforms providing the service. All-to-all trading can enhance liquidity by enabling greater market connectivity and centralization; but from a regulatory perspective, there are concerns with all-to-all trading, which I will talk about later.

Factors Driving Increased Electronic Trading in Fixed Income Markets

What are the factors driving the creation of so many electronic platforms right now, and the increasing shift to electronic trading?

One is the significant shift in the balance sheets of traditional liquidity providers—banks and broker-dealers—versus that of non-traditional liquidity providers—asset managers and proprietary trading firms. Since the financial crisis, regulatory reforms have resulted in greater capital and liquidity requirements for banks. In turn, the traditional liquidity providers have offloaded a significant amount of inventory. This pattern has only been reinforced by the “long and low” interest rate environment that has existed for nearly a decade now, squeezing volatility out of the market. Dealer inventories have declined, as has the ability for dealers to act as effective market-makers on a principle basis. In addition to the new regulations, or maybe as a reaction to them, the growth of the corporate bond market, in record new issuances, has affected secondary liquidity—particularly for seasoned issues.

This new liquidity landscape has likely affected the demand for electronic trading in a couple of ways. First, there is likely a change in the demand for liquidity services, led by market participants who demand liquidity because they may need to adjust their portfolios quickly to meet inflows and outflows. And second, as traditional liquidity providers move back toward more agency-like models, it takes more time and more work to identify counterparties and more opportunities for asset managers to selectively provide liquidity.

So it’s not surprising that there is a boomlet in electronic trading right now. But what does it tell us about the provision of liquidity? Market analysts have a number of interesting theories, and I want to pose two observations that are worth considering.

First, the research suggests that there are important differences between dealer-to-dealer systems versus dealer-to-customer systems. Dealer-to-customer systems predominantly tend to be “request for quote style”—accounting for about 95 percent of transactions—meaning that customers have essentially replaced calling multiple dealers by phone with an electronic version of the same activity. It may permit more parties to participate, but it doesn’t really change the dynamic of price formation or information asymmetries. In some ways, it reminds me of the introduction of technology to the floor of the NYSE, gaining efficiencies without changing the business model.  

Where new platforms tend to compete on providing more pre- and post-trade information is in the space of dealer-to-dealer platforms. And in fact, these dealer-to-dealer systems account for most of the growth in trading in recent years—estimated at about 45 percent of electronic fixed income trading. While the number of all-to-all systems has increased, interestingly, the smallest increase in actual trading activities occurred on these platforms. In total, this suggests that some of the fundamentals of the way trading has worked in this market are changing slowly, even with electronification.

This brings us to an important point. Despite what seems to be more limited impact of technology, electronic markets have already fundamentally changed price formation. A recent report by Greenwich Associates indicates that most all dealers say that they look to electronic systems to validate prices. But despite the technological innovation, evidence shows that dealers still revert to voice trading to complete most transactions.  

In addition, there is also an open question about the impact of electronic trading on execution quality. Electronic venues have likely led to better executions because market participants can see where prices are, the new technology has the opportunity to “democratize” the provision of liquidity, benefitting traders. But these platforms have also introduced algorithmic trading, particularly on the dealer-to-dealer platforms. As we’ve seen in equities, this is associated with smaller trade sizes, and less block trading.

The customer now is likely to hold more of the execution risk. Further, initial evidence points to the conclusion that electronification tends to lead to greater liquidity in already liquid names, but less liquidity in assets that are already thinly traded. So it’s harder to assess whether these changes provide a net increase or decrease in execution quality.

A recent Bank of International Settlements study on the role of electronic platforms in the fixed income markets made an important point about the evolution of liquidity provision. The BIS argues that banks and dealers provide liquidity in fundamentally different ways than do the non-traditional liquidity providers. The paper suggested that traditional liquidity providers respond to price uncertainty by increasing spreads but maintaining depth of book, while non-traditional liquidity providers tend to maintain spreads but reduce available depth. This difference—at least anecdotally confirmed by some large players in the market—means that the ways regulators monitor the health of liquidity may not be as effective at best, and conflicted at worst. Over reliance on “bid-ask”-like measures may not provide as much of a view on current market conditions as it has historically.

Do the recent trends represent more permanent changes in the way we might expect to see corporate bonds trade? The answer here is yes. The automation of fixed income trading is here, gaining momentum and not going away. It is catching up to the revolution that began in the equities markets in the 1970s, in the FX market in the 1980s and the Treasury market in 1990s.

I want to return to where I started my remarks, and go back to oft-heard concerns that electronification can’t happen in the fixed income markets because of the inherent nature of corporate debt securities. I recently moderated a panel at a Federal Reserve Bank of Atlanta conference. The panel focused on the current state of market liquidity, primarily in fixed income. Reflecting on what it takes to make electronification of trading dominate in an asset, one of my co-panelists didn’t think that heterogeneity in assets or even infrequent trading was a particular hurdle. He noted, however, that central clearing and consolidated interest—in the form of a CLOB—were vitally important. Co-location of trading interest creates an externality, increasing trading interest even more, while central clearing reduces costs and limits counterparty risks. We haven’t yet reached a moment in corporate fixed income markets where these attributes yet exist as the number of new platforms competing for eyeballs and orders suggests.  

Regulatory Concerns

I want to shift to some of the regulatory concerns. Electronic trading, in particular automated and high-frequency trading, poses a number of challenges to regulators. One of the issues we’re dealing with as a regulator is the lack of transparency regarding alternative trading systems, or ATSs. At FINRA, we have taken steps to increase market transparency for ATSs, including all "dark pools." We currently make the volume and trade count information for equity securities executed in an ATS available on the FINRA website, and plan later this year to start providing additional data about the levels of block activity within ATSs. Last month, we began publishing the remaining equity volume executed over-the-counter by FINRA member firms, including the trading activity of non-ATS electronic trading systems and internalized trades. This information will be available free of charge to all users on FINRA's website, and will help investors better understand a firm's trading volume and market shares in the equity market.

On the fixed income side, we’re seeing increased trading of TRACE-eligible securities in dark pools, greater use of request-for-quote processes, and more executions of customer orders by firms that participate as both broker and as dealer. So, in July we will begin requiring firms to report additional information, including the unique MPID of the ATS to TRACE. The new rule allows for firms to apply for an exemption from reporting if they meet certain criteria. If they qualify for the exemption, the FINRA firms that transact on their platform will have to indicate on which ATS the trade took place and include the MPID of the ATS in that field. The rule doesn’t apply to all ATSs, as some will continue to make their own trade reports, and use their own MPID for that report. The rule, however, will provide FINRA greater insight about the frequency with which the ATS is involved.

We’re also concerned about some of the protocols within the platforms. For example, some protocols allow participants to filter out participants with whom they do not wish to trade. Because participants may not wish to trade with certain counterparties, this may create an uneven playing field.

The October 15, 2014, volatility in the U.S. Treasury market also highlighted a number of issues with automated trading. The U.S. Treasury Department has issued a report on the October 2014 events and is gathering industry feedback on market structure through a request for information issued earlier this year. It is imperative that regulators and the industry continue to work together to strengthen programs that minimize the impact of market volatility and to limit market disruption.

Close

We may be at a classic Betamax vs. VHS moment, or more recently, Blackberry vs. iPhone, waiting to see which style of electronic trading dominates. Or we may still be at a moment of great ideas before their time—like sixdegrees.com or pets.com; but the forces of technology, and to a lesser degree regulatory reform, are inexorable and formidable. The challenge is to continue to develop venues that provide an opportunity to trade fixed income at a fair price, in size and with immediacy, and provide a fair return to the service of liquidity provision.

I look forward to today’s conversations, and the future of this pursuit.

References

Bank for International Settlements – Markets Committee (2016): Electronic trading in fixed income markets, January

Bank for International Settlements Quarterly Review (2015): Shifting tides – market liquidity and market-making in fixed income instruments, March

Bank for International Settlements Quarterly Review (2016): Hanging up the phone – electronic trading in fixed income markets and its implications, March

Barclay, M; Hendershott, T; Kotz, K (2006): “Automation versus Intermediation: Evidence from Treasuries Going Off the Run”, The Journal of Finance, Vol. 61(5)

FRB St. Louis Review (2006): The Transition to Electronic Communications Networks in the Secondary Treasury Market, November

Greenwich Associates (2015) “Corporate Bond Trading Market Structure Update: Overview and Insights”, February

Greenwich Associates (2016): “Has Bond E-Trading in Europe Hit the Ceiling?”, February

Hendershott, T (2003): “Electronic Trading in Financial Markets”, IEEE, July

McKinsey & Company (2013): Corporate bond e-trading: same game, new playing field, August

Securities Industry and Financial Markets Association (2016): Electronic Bond Trading Report: US Corporate & Municipal Securities, February

Treasury Borrowing Advisory Committee (2013): Electronic trading in the secondary fixed-income markets