"Danger, Will Robinson: Advanced Analytics and Regulation of Markets"
Keynote Address of Commissioner Bart Chilton to the Institutional Investor TraderForum, New York, NY
January 26, 2011
Good afternoon, it is good to be with you. A special thanks to Lew Knox for the kind invitation to be here to talk with the members of TraderForum about a few things going on in markets and in Washington today.
“Lost in Space”
First, let us start with a little television trivia. This will date me and some of you. By a show of hands, who remembers the 1960s TV series “Lost in Space?” Young William Robinson and the Jupiter 2—does anyone remember that? Of course, the line that caught on and we still hear once-in-a-while today is “Danger, Will Robinson, danger.” Well, as Cliff Clavin, the postman from Cheers would say, here’s a little known fact: despite the millions of times that phrase has been uttered all over the world in the decades since the show, the robot only used that phrase exactly once in the entirety of the series. There were 83 episodes of “Lost in Space,” which by the way ran three seasons. Only that once, in episode 11 of season three, was “Danger, Will Robinson, danger” used. A movie released in 1998 also used those words. It is remarkable in these days where repetition of a message is key in all of the media clutter to having folks remember something, yet “Danger, Will Robinson, danger” was only used that once on the tube.
Well, today we are going to discuss some potential dangers out there with regard to algorithmic, high frequency trading (HFT) and the advanced analytics that are being used in financial markets today. We will also talk about some of the dangers of regulation—too much and too little. As part of that discussion, I will let you in on a hush-hush strategy that some are in the middle of as a way to get out of regulation. What I can say is that if we are not cognizant of the dangers out there, on these topics and on others, we certainly have the great possibility of ourselves being lost in space.
As we all know, our markets have changed dramatically in the last decade. Open outcry in trading pits is quickly becoming a thing of the past. Instead of traders screaming at each other in the pits, computers are screaming all day long, not once raising their voices or taking a coffee or smoke break. Algorithmic programs are cranking away like journeymen and HFT computers are trying to scoop up micro-dollars in nanoseconds. It is an amazing thing how quickly and vastly these markets morphed.
Technology in markets and in trading is valuable for many reasons. It adds liquidity. Although, it may just add liquidity at first to other HTFs, but liquidity it does add to markets. It adds access. The third largest trader by volume on the CME is based in Prague. Technology also provides an electronic data trail, which is great for regulators. This is a vast improvement over sorting and taping together pieces of papers from trading room floors. As with most things, however, there are potential dangers that compel us to consider the myriad ramifications of technology.
One such ramification gave us a wake-up call on May 6—the Flash Crash. Talk about “Danger, Will Robinson!” Jeez, was there danger. Financial markets came unwound that afternoon. You have heard the horror stories.
Last summer, the joint SEC-CFTC Advisory Committee on Emerging Markets and the CFTC’s Technology Advisory Committee (TAC), met. We spoke about elements of the Flash Crash that may have been caused by so many orders going into the order book at such a high-speed that it overwhelmed the New York Stock Exchange computers. One member suggested something he called “algorithmic terrorism.” I also asked about algo programs that work to instigate other algo programs into action. I called it algo-price piracy, where one algorithm “invades” another and trades on that action. It does not appear this took place on May 6, but many experts agree it could happen. It may be innocent. It may be a pre-meditated market move to gain a trading advantage, or it could be outright financial terrorism.
It may surprise you, but mini-flash crashes occur all the time, too. They don’t cause as much of a disruption as that of May 6, but more than once last year in futures markets and several times in stocks, runaway robotic programs disrupted markets and cost people money. Last February, for example, one company lost a million dollars in the oil market in less than a second. That company lost its own money, but sometimes, whole markets are affected and innocent people are hurt.
As you know, these advanced analytic computers are intuitive. They can track market moves and make adjustments with lightning speed. They are sensitive and receptive, reactive and responsive. They play off one another. There was a recent newspaper article that made the point that one new strategy being used these days is to have mega-powered computers read news reports—even Twitter feeds—and then trade on the information. Can you believe that? The computer scans Twitter, and news stories, or blogs, whatever, and makes a determination about activity in this or that and trades on the information. Of course, humans programmed the computer, but the actions are taking place by the computer once it is off and running.
What does all this mean? To me as a regulator, it means we need to be careful of the dangers. We certainly should not just assume all will be well. We have evidence that there not only could be problems, but there have been problems. There are real dangers.
Things can happen so fast that, all too often, regulators are like a fire department that comes in and cleans up the charred remains, or CSI and comes in to do a post-mortem and find out whodunit. Give it six or nine months and we will give you a report, thank you very much. That shouldn’t be acceptable in today’s environment. These markets are too important, too tied to the economic engine of our democracy and to the livelihoods of all Americans—not to mention the global ramifications.
In practice, regulators need to be more aggressive and pre-emptive—more like the police department trying to stop bad things before they take place. We have to be nimble and quick, and looking around the corner to see what is going on next. On these advanced analytics, remember the CFTC began operations in 1975 for human-to-human trades, not for computer-to-computer trades. Now, regulators require supercomputer-like systems to monitor the kind of trading that goes on. We essentially need to get into the mind of the mainframe. Even then, however, even if we get a better handle on how to regulate this breakneck speed trading, the methods, the machines and the markets will continue to evolve. For sure, all that means we need people power, but like a spaceship trying to exit the Earth’s gravitational pull, we also have a dire technology need for speed in order to keep up.
So, let us do a little of that right now. Think ahead about what could or should be done—if anything. Given our experience with the Flash Crash and mini-flash crashes, I’d argue we need greater risk controls.
In addition to things like better harmonization between futures and equities and in including coordinated circuit breakers (not to mention greater global harmonization), we should consider whether there should be limits on advanced analytic trading. Should there be specific position limits for HFTs? Let us say that we, the CFTC, allow 10 percent of open interest in a futures market to be held by one trader. Should high frequency traders be allowed to trade 10 percent, repeatedly, in a very short time period? What about five HFTs, each trading 10 percent of the market in ten minutes, in concert, and it moves a market? Is that acceptable? What are the boundaries of play? Right now, there really are none. It is an open universe. You want to develop a program that does this or that? Fine, go right ahead. You want to sell the name of a star to somebody? Go right ahead.
The issue of HFT trading, though, raises a broader question about this type of trading, in general. Don’t get me wrong, HFT trading is part of our world today, but should there be different rules for these market participants? Is this type of trading in any way outside the boundaries of the fundamental purposes of capital formation and risk management in financial markets?
I have heard from a number of commercial firms trying to hedge their risks who complain about the inability to get into the markets as they have in the past due to the sheer number and speed of HFTs. By the time his company is ready to click their mouse and make a trade, there are so many HFT orders jumping in front of the commercial that they cannot get in at their price point. They get in eventually, but not at the price they wanted. I understand there are arguments on both sides of this issue, but the danger is that if we lose the commercials, we will not have the types of markets we need to ensure price discovery and appropriate risk management.
Another way to address some of these potential dangers is to impose legal responsibility on high frequency and algo robot trading that roils markets and causes concerns and obstacles for traditional traders. I think those who instigate runaway high frequency or algo robot trades should be held accountable when they hurt other market participants or injure consumers. We should set limits on certain conduct that is specific to algo robots and high frequency trading. Taking into account what happened on May 6, this just seems reasonable.
I also believe there should be some standard definition of what high frequency trading is and maybe even a kind of “Good Housekeeping Seal of Approval.” Should these trading programs be demonstrated in a market-testing environment to see what and how they do? Should the exchanges be required to, at least, monitor some of the potentially dangerous trades? Shouldn’t we be able to get information about all of these types of trades—be able to slice, dice and chop up data to get a better handle on what is and could go on in markets?
Wall Street Reform
Advanced analytics and HFTs are all “gee whiz” stuff and they pose these interesting and challenging policy issues, but there are, however, many other fascinating changes taking place in markets and in our economy today—and some other potential dangers, one of which we can spend some time on now: regulation.
You are all smart folks, so I won’t give a long description of what I think took place leading up to the economic meltdown. Here is my summary though: it started 1999 with the repeal of a key banking law and followed by other deregulatory changes. If you had to point to one thing, I suppose most folks would say the subprime mortgage fiasco. That was compounded by credit default swaps (CDSs), which many times were comprised of some of those same crappy subprimes—all taking place beyond the view of regulators in dark over-the-counter (OTC) markets. Excessive leveraging was also a key culprit. Lehman Brothers, for example, was leveraged at 40 to 1. The economy went into the gutter under that regulatory regime. Hence, Congress saw the need to re-write financial laws. Last July, Congress passed and the President signed the most sweeping financial reform bill in history: the Dodd-Frank Wall Street Reform and Consumer Protection Act.
If there is one chief concept behind the new law, it is transparency. In the futures world, the major change will be bringing light and regulatory oversight to the OTC derivatives market, which some have equated to a black hole. To give you an idea of how this increases the regulatory universe, the currently CFTC-regulated exchanges account for roughly $5 trillion in annualized trading. The OTC market is roughly $600 trillion. To say the least, we have our work cut out for us.
Funding: A Clear and Present Danger
As I said, however, there are dangers. First, we need to ensure that we don’t over-react, that we don’t go too far in our efforts to regulate. That could be disastrous for business and for our economy. But there is a more urgent danger, and to me this is—as they say in the world of foreign affairs—a clear and present danger. Congress gave regulators the responsibility to implement the law, but guess what? There may not be funds to do the job, to have the people power and the computer capability that I have talked about to actually institute the changes Congress itself sought. Without additional resources, we could be right back where we were when the economy tanked.
Right now, we are operating under what is called a continuing resolution or CR, which means we are operating under last year’s budget. We can write the rules and continue to do the work we have been doing for now. However, six or nine months from now, as the new rules are being implemented, more resources to enforce the law will be needed. We cannot increase our responsibilities without increased resources—period.
Nobody wants to hear somebody from Washington talking about needing more money, but remember that we are taking on a large part of that $600 trillion in trading that has been totally unregulated. Therefore, we can write great rules that will be good for business, and for consumers who rely on these markets for price discovery of almost everything they buy. We can collect all the data from that trading. Without resources, however, we will not have the ability to analyze what it means. As the robot from the “Lost in Space” also said, “That does not compute.”
Bottom line: if we don’t find a way out of this dilemma soon, we will have a monstrous moment where many folks say: “Danger, Will Robinson.”
Here is the hush-hush secret. What many of us perceive as a clear and present danger because of a lack of funding has put some folks’ tail in a wag. Opponents of reform could not stop the Dodd/Frank Act from becoming law. Opponents of reform have not been successful in getting the agency to grant cavalier exemptions or cater to their every desire. They are smart enough to know that Congress will not pass a wholesale reversal of the new financial reform law, although one Member of Congress has proposed a bill to do so. Therefore, now they have another idea—a double secret strategy. They seek to deny resources to regulators—starving us on the vine if you will—and thereby denying us the ability to enforce the new law and oversee these markets. And, they think it just might work.
For 25 years in government, I have been opposed to user fees of various types and stripes. Ronald Reagan proposed many of them in the 1980s. I have opposed user fees in financial markets, because markets are a public good and therefore the government has a responsibility to ensure they are efficient and effective. Use of public tax dollars for this purpose is appropriate. I still believe that. If we are faced, however, with the draconian option of no funding to implement the reform bill, putting us directly back where we were in 2007 and 2008 when the economic mess began to show its ugly head, then perhaps some type of user fee is the least onerous remedy. Not having necessary and important market oversight certainly is not the way to go.
Approximately a billion-and-a-half contracts are traded on regulated commodity exchanges in the U. S. every six months. If market participants are charged even a third-of-a-cent transaction fee on those trades—assuming we were given the authority to assess such fees—this could provide the funds to do our jobs. But I wouldn’t suggest we just have a fee only on futures. We certainly should include swaps transactions as well—the new area of regulation that is actually going to require us to increase our workload. In fact, I wouldn't support this draconian option only on futures transactions—if it were considered—and again, I hope we don't ever have to get there—it would have to be done on a level playing field with the newly regulated swaps market, and ensure that those transactions had appropriately calibrated fees as well, so that neither the futures nor the swaps market was advantaged or disadvantaged.
By the way, last year we operated at $168 million. With the passage of the new law, we requested $261million. That leaves a funding gap this year of $93 million. The President will submit his budget request to Congress within the next several weeks. Whatever numbers are requested and proposed, it is clear that we will require tens of millions more than we did before the new law was passed.
The CFTC already has some authority to impose fees. We collect some fees currently and we could change that system. We also have authority to propose a plan to collect additional fees and seek the approval of our authorizing committees of jurisdiction. Finally, we can seek legislative authority to impose additional fees as an off-set to the agency’s appropriations from Congress.
Again, I do not like having to take this route. Traders and exchanges will not like it. If we don’t get funding from our traditional appropriations, however, we risk the kind of lax oversight that got us into such a fine economic mess starting just a few short years ago. So yes, here too, I’m saying “Danger Will Robinson.” We must have effective and efficient markets free of fraud, abuse and manipulation for the benefit of the consumers who are impacted by them every time they pay for a gallon of gas or orange juice or milk.
If we don't receive the needed funding this year, I hope we use our existing authority to impose sort type of reasonable fee, and or seek approvals for additional fees and request from Congress additional authority to impose supplementary fees.
Sometimes memories in Washington are too short. If some of us can remember a TV series from more than 40 years ago, however, surely we can remember what happened in markets just a few years ago. Congress didn’t forget when it wrote the Wall Street Reform Act. It has only been since it passed that the amnesia set in. The regulation of our markets needed to be reformed and it still does today.
It has been a pleasure to be with you today. In looking through all of the “Lost in Space” stuff, I found one final quote attributed to the robot: “My micro-mechanism thanks you, my computer tapes thank you, and I thank you.” With that, I will say thanks myself. I appreciate your attention.
Last Updated: January 27, 2011