Good morning, it is a pleasure to be here.
Thank you, [ICI President] Paul, [Stevens] for that kind introduction.
It’s good to be with so many professionals whose job it is to help American families reach their financial goals. I know that we share a commitment to stable financial markets that provide investors the opportunity to achieve the economic security they desire.
Of course, as we know, markets are not always stable. And, one year ago today, we got a reminder of just how unstable they can be.
At 2:42 on the afternoon of May 6, 2010, stock prices began to fall with almost unprecedented speed: a Dow that had already lost almost 4 percent on the day plunged almost 600 points in about five minutes. That left the nation’s most prominent stock index down nearly 1,000 points from the previous day’s close.
And then, just as suddenly, the markets reversed themselves, recovering almost 600 points in the minutes that followed.
Between 3 and 4 p.m., 17 million trades were executed. In that time, investors suffered steep losses, and respected companies saw their reputations harmed as stock values evaporated almost instantly.
But the significance of May 6 is greater than the investor harm caused by these wild swings in prices — it lies in the significant blow to investor confidence this volatility delivered, as well. Because, while every investor accepts financial risk as a fact of life, they operate under the assumption that America’s markets are structurally sound — that the funds you represent and the investors you advise could confidently entrust their capital to the world’s most sophisticated financial markets.
When that confidence declines, the ramifications — in lost wealth and increased cost of capital — can be great.
So we responded swiftly. Within days, I summoned the heads of the exchanges and FINRA to the SEC to discuss ways to strengthen our markets and to address unnecessary volatility. Later that month, our staff issued, jointly with CFTC staff, a preliminary analysis. And, by early the following month, we had approved a proposal to put in place market-wide circuit breakers on individual stocks.
This rapid response reflected not only the significance of the event, but also our effort, begun months earlier, to engage in a comprehensive review of the structure of the U.S. equity markets.
Let me take a moment to make one thing clear before I go on: today’s financial markets are fast, accessible, efficient and inexpensive by historical standards. They are, in many ways, a marvel of capitalism and technology.
But our markets are extremely complex, as well — with 15 equities and options exchanges, more than 30 dark pools, 3 electronic communications networks, and more than 200 broker-dealers that execute orders internally. This complexity means that, when things go wrong, the consequences can be sudden and severe. It is important, then, that we identify and put in place systems that can quickly contain and minimize harm should a breakdown occur. We must also continue to scrutinize the markets in search of ways to strengthen the system overall, so that breakdowns can be prevented in the first instance.
As outlined in our January 2009 concept release, we have established and are executing on a broad market structure agenda. But addressing the unprecedented and unacceptable events of May 6 has, not surprisingly, been the focus of those efforts since that day.
On May 6, a combination of algorithmic trading systems and rational trading behavior drove a broad liquidity crisis that caused unprecedented gyrations in the equities markets.
The first domino to fall was the broad index level in the E-Mini S&P 500 futures product, which experienced a sudden liquidity crisis. The impact of this liquidity crisis was then transmitted almost instantaneously to the securities markets through cross-market trading strategies.
The next phase of the disruption was a liquidity crisis in hundreds of individual securities. At the worst end of the spectrum, more than 300 securities suffered declines of more than 60 percent from their 2:40 p.m. prices. Ultimately, more than 20,000 trades were broken, many of which were executed at prices of a penny per share.
As noted in the joint SEC-CFTC Staff Report released last September, this liquidity shortage and subsequent volatility were exacerbated when the automated trading systems of many large securities market participants temporarily paused in reaction to the sudden price declines. Some market participants widened their quoted spread, others offered reduced liquidity, and a significant number withdrew completely from the markets. All of these were rational responses by individual market participants. But, in aggregate, they were extremely disruptive at a moment when the markets desperately needed liquidity.
The resulting extreme price moves across hundreds of securities at once were unprecedented in modern U.S. equity market history.
The SEC’s response, launched in the immediate aftermath of the event, continued as we worked to bolster our markets and restore the confidence of investors:
The lengthy and laborious detailed investigation of May 6 also underscored the need for a better way to reconstruct trades. That is why the SEC proposed the creation of a consolidated audit trail — a system that would provide the SEC and other regulators direct, timely access to data on all orders in the national market system, from all participants across all markets.
This would allow us to rapidly reconstruct trading activity and quickly analyze both suspicious trading behavior and unusual market events.
We proposed creation of such a system last May, and we are considering the many comments received as the staff prepares its recommendations to the Commission.
In addition to these actions, the CFTC and SEC asked the recently-established expert Advisory Committee on Emerging Regulatory Issues to consider the market disruption and make recommendations related to market structure issues that may have contributed to the volatility of that day
In April, the exchanges and FINRA, working with Commission staff, proposed a major expansion and enhancement of the pilot circuit breaker program aimed at addressing extraordinary volatility.
This proposal, dubbed the Volatility Plan, which is available on the Commission’s website, offers many potentially important changes.
In thinking through our next steps, we need to consider several important questions:
These questions touch on fundamental public policy objectives for the U.S. equity markets and deserve a robust discussion.
And, in answering them, I believe that we must be guided by the objectives of promoting capital formation and — critically — protecting the interests of investors.
Regarding the first question, it is possible to make an argument that occasional instances of extreme volatility are simply a fact of life for the equity markets. News stories, earnings reports and other events that affect market valuations should get priced in quickly.
But the events of May 6 crossed an important threshold. First, tremendous harm was done that day. And second, rather than resulting from an irregular but unavoidable occurrence, this harm resulted primarily from a technical market structure problem — not a "real" change in valuation or other fundamentals of the issuers.
Approximately two-thirds of the intra-day decline on May 6 was reversed within minutes, as buyers rushed in to take advantage of the drop. Since that day, we have heard over and over again from fundamental investors that, given the opportunity, they would have been strong buyers during the decline.
But during the decline, sell orders simply outpaced buyers’ ability to trade. Indeed, the speed of the broad decline caused many market participants to fear that a cataclysmic event — of which they were not yet aware — had occurred. This further widened the gap between short term supply and demand as investors stayed on the sidelines, until their fears could be allayed.
While it’s difficult to determine the exact point at which short-term volatility becomes "excessive," it is nonetheless a critical question and one that demands our attention.
The answer to the second question — how excessive volatility affects different entities — is not difficult to discover. But it needs to be underscored because the answer points us toward a possible solution.
Excessive volatility harms a listed company by increasing the perceived risk of investing in its stock. And many individual investors were harmed when they relied on the integrity of the market prices, suffering losses when their orders were executed at temporarily disrupted prices.
On the other hand, many short-term traders, whose revenues are typically driven by volatility and volume, enjoyed a highly profitable day.
The importance of these traders’ actions can perhaps best be understood by comparing their actions to those of their predecessors during the "market crash" on May 28, 1962.
There are a number of similarities between 1962 and 2010. For example, neither of these severe price moves could be readily explained by a particular news event. On both days, some market data systems were overwhelmed by the heavy volume.
And, in both instances, the sudden declines struck at investor confidence, leading them to question the stability and integrity of the equity markets.
But the differences between those two events are even more striking.
First, the magnitude of the declines, both at the broad market index level and for worst-hit individual securities, was much more severe in 2010 than 1962. In ‘62, the Dow declined to intraday lows of 6.3 percent compared to 9.9 percent on May 6. And one of the worst-hit individual securities in 1962 dropped 9.3 percent in a 12-minute period. In 2010, many securities lost 100 percent of their value in a matter of seconds.
Perhaps the biggest difference — and one that may help explain the difference in the magnitudes of the declines — is the volume and trading behavior of the professional traders who were expected to be the primary liquidity providers.
In ’62, the specialists who were then the primary liquidity providers, represented approximately 17 percent of market volume and were net buyers in aggregate during the decline. In 2010, the high frequency traders who are today’s liquidity providers represented well more than 50 percent of market volume and were net aggressive sellers during the broad index price decline.
High frequency traders turned what was a very down day for many investors into a very profitable one for themselves by taking liquidity rather than providing it. I think their activity that day should cause us to thoroughly examine their current role.
The final question is more difficult. Do the Commission’s mandates to protect investors and promote capital formation justify the imposition of trading obligations or other responsibilities on high frequency traders?
Naturally, we are mindful of the unintended consequences that could result from imposing such obligations. For example, imposing significant obligations for market quality on some firms, but leaving other firms free to operate without those obligations, could create an unfair playing field and, in the end, do little to promote market quality.
In addition, we need to assess the entire regulatory structure surrounding high frequency trading firms and their algorithms. Do current regulations reflect their impact on trading? And, are their algorithms programmed to operate properly in stressed market conditions?
As we continue to examine the options of imposing enhanced obligations on high frequency traders, we are considering other changes, as well, that may offer investors a measure of needed protection in the near term.
The Volatility Plan proposed by FINRA and the exchanges would, for example, serve to limit algorithmic trading of the sort that added momentum to the price decline on May 6.
In particular, reducing the trigger to 5 percent is designed to be more effective in halting non-fundamental runs on individual stocks and even the broader market.
The proposed plan would have made a real difference on May 6. The current price parameters in the pilot circuit breaker program are set at a 10 percent price move within a five minute period. Yet at least 98 percent of the trades and 84 percent of the securities involved in the market disruption on May 6 would not have been affected by the current level of circuit breaker protection.
Under the proposed Volatility Plan, a price move that hits the percentage parameters would trigger a brief "limit state" in which a liquidity imbalance would have 15 seconds to correct itself "naturally." In a down market, for example, the limit state would provide 15 seconds for buyers to respond to a sell imbalance and stop a cascade of prices. If the plan is approved, this brief period could allow minor liquidity imbalances to be corrected without the need for the full trading pause.
Also, if the plan is approved, these pauses could provide a period in which market participants have an opportunity to assess the market and decide whether and at what prices they wish to buy or sell. The result should be trading driven less by momentum-seeking algorithms and more by rational trading based on fundamentals.
Setting appropriate price parameters is also important as we work with the CFTC and the securities and futures markets to update the existing index-wide circuit breakers that apply to both the securities and futures markets. These index circuit breakers have never been triggered at their current price parameters, which require a broad market price move of at least 10 percent from the previous day’s close. The index circuit breakers need updating to reflect current levels of volatility, as well as to work smoothly with the proposed Volatility Plan.
Another contributing factor to May 6 was the fragmentation of trading across a great many different venues, including many dark venues. On normal days, the volume of orders handled by these venues, which include dark pool ATSs and internalizing broker-dealers, exceeds 30 percent of total volume in U.S.-listed stocks.
During the worst of the disruption on May 6, their share of volume plummeted to 11 percent as liquidity disappeared in the dark venues and sell orders suddenly were diverted to the public markets. Our staff is considering whether the respective rules applicable to public and dark venues appropriately reflect their contribution to price discovery and stability.
Finally, we need to look beyond May 6, at other issues that could lead to structural breakdown. One imperative is ensuring that the technology systems of exchanges and other key market participants are sufficiently robust to handle today’s trading practices, particularly during volatile periods when volume and message traffic can skyrocket.
I believe, for example, that the SEC should consider making compliance with the SEC’s Automation Review Policies mandatory — and as such require market participants to meet adequate standards for the capacity, resiliency, and security of their automated systems. These rules could apply to exchanges, alternative trading systems handling appreciable volume, clearing agencies, depositories and securities information processors.
Our response to May 6 was rapid — and as a result of our actions, we have reduced the likelihood of a similar event occurring again.
But we will continue to consider additional ways to improve the structure of our markets.
We need to continue examining the effects of high speed trading on the markets and on buy-side and fundamental investors. The role of these traders, whose prominence in the markets seems only to increase, should be subject to further scrutiny. The possibility of imposing obligations during times of potential turmoil must remain on the table. And we need to pay attention to other potential flaws that could bring about equally disruptive events.
The stakes are sufficiently high — for markets, for firms seeking to raise capital at a reasonable cost, and of course for the investors investing through the funds you manage — high enough that further examination is not only desirable but required of conscientious regulators.
The events of May 6 remind us that despite the enormous benefits and efficiencies of today’s markets, there are risks as well. But these risks can be identified and addressed.
No market structure could ever entirely eliminate the possibility of major imbalances arising between sellers and buyers, or discover every potential imperfection in the system. The question is whether we have sufficient safeguards in place to ensure that the markets operate in as fair and orderly a fashion as possible on good days and bad days alike.
As we continue to address the structure of the financial markets we will do so in the spirit of the measures we have already taken: measures that are already strengthening our market structure and making it more resilient; measures that align precisely with our mission to protect American investors; and measures that seek to promote the capital formation American enterprises need.
Thank you.