The Gateway to Algorithmic and Automated Trading

Praising the Modern Markets

Published in Automated Trader Magazine Issue 36 Spring 2015

Quantlab set up shop as a long-short statistical arbitrage hedge fund in 1998. Regulatory changes such as decimalisation in the early 2000s adversely impacted the firm's implementation costs and new execution strategies had to be found. That exploration led to Quantlab becoming one of the top HFT firms in the US.

Cameron Smith, President, Quantlab

Automated Trader talks to Bruce Eames, co-founder and CEO, and Cameron Smith, president, about the maturing HFT industry in an era of regulatory scrutiny.

Automated Trader: What kind of operation is Quantlab?

Cameron Smith: Definitions are always hard but we turn our portfolio over a little bit each day, so we probably would be in most people's definition of high frequency trading even if that is not a term I like. I believe we are unique among so-called HFT firms because we came into this business trying to solve our own execution problems and did so in the relative isolation of Houston, Texas, without any preconceived notions of what to try. Given that, we had to base our approach solely on quantitative research. In comparison, many of our competitors came off the floors and developed trading strategies based on their experience from the pits.

AT: What asset classes do you trade, and in which markets?

CS: Equities and futures, FX, treasuries and equity options. Not all of those in every jurisdiction however. We trade essentially globally, where there is the potential to trade. So firms like ours are going to be active in North America, and then Europe, and then Asia. A bit in South America, Japan and Australia.

AT: Are you in other emerging markets?

CS: We generally are not. I assume several of our competitors are in a lot of those places. But those markets, like Hong Kong, India, South Korea, have huge transaction taxes or stamp duties on equities. So, even though some of those markets are quite large, they don't lend themselves to what I would call professional trading.

AT: Any new directions for Quantlab? Where do you see growth?

CS: We spent the last several years just improving what we do. What you see is a maturing of the market and a reduction in trading costs for investors, and a reduction in trading profits for professional traders. Rather than new directions, we have been focusing on getting better at what we already do.

There is a mythology that speed is central to success but the reality is that the quantitative predictive aspect is the key to success. Just trading fast doesn't make you any money if it is the wrong trade. You need to know when to trade and then execute that order as quickly as you can. When we look to the future, we are keeping an eye on Asia and at Dodd Frank which, through regulation, may create opportunities to trade things that have been traditionally traded on bank only venues.

AT: Like FX and fixed income?

CS: Yeah. It's ironic that Switzerland is saying we need key instruments to be traded on an electronic exchange with automated trading to avoid scandals like Libor. So, while we have all this criticism of automated trading in the US, the reality is automation creates an incredibly fair and efficient market.

When you start comparing asset classes, everyone is wringing their hands over equity markets and Michael Lewis writes a book about it, but there is a lot more notional value being traded in FX and fixed income and they are way behind in terms of fairness, best execution and execution costs.

So we look forward to having an opportunity to trade something like fixed income in Europe.

AT: Do you find barriers to entry for trading firms to be getting better or worse?

CS: Competition seems to be intensifying, that is why we spend a lot of money and a lot of time always tuning our technology and our quantitative part. Our trading systems from four years ago probably wouldn't do much today, so there is a constant need to improve and come up with new strategies and diversify. At the same time, it is more expensive for new firms to enter the market, at least on a global basis.

AT: How do the technological barriers to entry compare to the regulatory ones?

CS: For technology, there are many service providers that allow a trader to basically show up with an algo in a 'lunch pail' to plug into an existing infrastructure that could allow that trader to roll that strategy out on a potentially global basis. But while the technology is there, the compliance and regulatory risks are increasing.

If you are a global bank and you have clients engaged in high speed trading, you have to be really careful who you associate with. Even the most innocent of mistakes is really played up in the media and with the regulators.

Let's say you mismarked something in the US and it was purely a technical issue. No investor was harmed but you had a system that wasn't programmed correctly and nobody caught it. An honest error. Maybe five to ten years ago, that fine is $50,000. I am not even sure what the fine would be now for the same thing, but I think it would be in the millions, and so that has an impact on people's willingness to do this business.

Bruce Eames, Co-Founder and CEO, Quantlab

Bruce Eames: On one hand, the technology barriers to entry have gone down. On the other hand the costs of trading have gone up significantly - like exchange fees, data fees, carrying charges - that firms never faced before. Especially after the exchanges became public entities and they started looking at every aspect of their business that they could monetise.

Books like Flash Boys did a disservice in the name of entertaining people really. I think it set things back. It distracted the attention of regulators, legislators, and the public in general, from the real issues. I don't think any high frequency trading firm argues that there aren't some aspects of the market microstructure that need attention and can be improved.

CS: Ultimately, when you read people's books and testimony, and ignore the first two sentences or so about HFT, it turns out we want the same things changed in market structure.

AT: What are some of those things?

CS: For example, I have long advocated eliminating the ban on locked markets. The problems it creates contributes to inefficiency in the market and fragmentation. It facilitates internalisation and dark pool activity. In theory, we don't want to make it easy for people trading away from the exchange.

In the US, there are thirteen exchanges and no exchange has more than 22% of the volume. There are 40 dark pools. If you are an investor who wants to get the best price in IBM and price discovery is happening in 53 places, that's a lot of fragmentation and complexity.

BE: It is all the professional intermediaries that are trading and arbitraging between these fragmented markets that is driving the prices to the equilibrium price. That's what high frequency traders do.

CS: But we think there is excessive fragmentation. We don't need nor want forty markets or 13 exchanges.

AT: In 2013, Quantlab became one of the founding firms of the Modern Markets Initiative, to broadcast some of these issues. What is the message MMI is trying to get across?

BE: What we are trying to do is get the facts out there because unfortunately the narrative is kind of one-sided at this point.

We are aiming our message towards the industry and the public, because all of the major firms in the industry, they have their lobbyists talking to regulators and the legislators for policy, but nobody is trying to put out that counterpoint to the message that uninformed writers and reporters are putting out there in the media. When you look at the data, that is when you start seeing the reality.

CS: To give you an example, Fidelity's Bill Baxter (head of global programme trading) said that his trading costs have gone from 120 basis points to 32. Automated markets are really valuable and business has never been better. I don't think any media outlet reported that. MMI hopes to change that.

BE: For years the firms in high frequency trading kept a very low profile because in many respects they were learning the industry, and there is a lot of uncertainty. So it is very difficult to get our colleagues to pull together to really answer some of the criticism and try to inform the public about how the markets work, and how high frequency traders interact. In our frustration, we finally said we will just start the organisation ourselves. Very quickly some of the other firms said, OK we will join.

AT: The other three founding firms of MMI - Tower Research Capital, Hudson River Trading and Global Trading Systems - were named as clients of Barclays dark pool in the NY Attorney General's complaint alleging predatory trading in the venue. Has that damaged MMI's credibility?

I don't want to comment specifically but it should not be a surprise that some of the largest professional trading firms were trading in one of the largest dark pools. MMI has consistently stated that markets need professional traders. Without professional traders there just isn't enough liquidity which just increases the cost of trading for all investors. This is true for exchanges or dark pools that want to best serve their clients with a low cost, liquid market.

AT: What would you say to critics who believe you are just finding the data that happens to fit within a "pro-HFT lobby"?

CS: First of all, I would say to such critics 'tell me what is wrong with the data, methodology or market quality metrics'.

By analogy, for health we have measures such as cholesterol, triglycerides and blood pressure. The assessment of person's health is based, in part, on these metrics.

Market experts have developed similar metrics for markets. These metrics are things like volatility, spreads, price efficiency, and implementation shortfall. And by all these measures the US markets are the healthiest and most efficient and are much better now than before automated trading. The critics claim that authors of these studies we often don't even know are all pro-HFT but what they have yet to do is question the measures of market health or otherwise prove that their standard calculations are wrong. They should be asked to provide their statistical measure of market quality and not simply tell some story about how they tried to execute a trade last Friday and it didn't go well. The data driven firms like Fidelity and Vanguard unequivocally say the market has never been better.

AT: MMI recently signed on former CFTC regulator Bart Chilton, who used to call HFT firms "cheetahs". Cameron, you are a former SEC regulator yourself. What would you say to concerns that the HFT industry and regulators are too cosy?

CS: If they are cosy with HFT, they have been very good at hiding it from our perspective (laughs). The reality is that we live in a world where it is sort of conventional wisdom now that Silicon Valley, maybe Google, should have opened an office in Washington a long time ago, or Microsoft before that. If folks are going to mispresent our business in Washington or elsewhere then we are going to need to correct the record.

Further, "cosy" suggests that we would want less regulation in the market. But when it comes to market regulation, professional traders have a huge stake in the market being well surveilled. For example, the Justice department recently announced sanctions against a firm that was placing bids and pulling them over time periods of like 28 milliseconds. A human eye can't even see that, so who do you think they were trying to ensnare? Professional traders.

And judging by the escalating fines I see of all types of trading firms, it doesn't look very cosy to me. I think it is almost a zero tolerance environment right now. You have to be totally on your game and really good with your risk procedures, otherwise there is a big risk for you.

AT: Despite all this regulatory scrutiny in the US, quite a few of the emerging markets are actively welcoming HFT activity. Any advice?

BE: Learn from the mistakes that you see in the US and Europe and be very responsive to bad actors in your markets because that increases the integrity of the markets.

CS: Trading activity that (regulators) are announcing the fine for in 2013 dates back to 2009. Regulators need to move faster in interdicting and sanctioning bad actors. The deterrent value drops significantly if the fine is four or five years later.

BE: And the other additional point I would make to emerging countries is that having an open free efficient market is beneficial to the economies and you don't want to hamper that by viewing that exchange as a source of revenue for the government. That is where things like transaction taxes come in, there is nothing they can do to hold down the growth and efficiency of their markets more than imposing heavy transaction taxes on the markets and expecting them to be a source of revenue stream.

Note: This Q&A has been edited and condensed.