Joe Saluzzi: HFT Parasites are Killing the Market Host
Joe Saluzzi, expert on algorithmic trading -- also known as high-frequency trading, or HFT -- returns as a guest this week to explain how the players behind this machine-driven process act as parasites that are destroying our financial markets (and, increasingly, even themselves).
Since Joe first spoke with us last year, HFT firms have only increased in size and share of market activity. Here are some staggering statistics on how influential they have become:
- HTFs make up between 50-70% of the volume seen across market exchanges today.
- 2% of the traders on many exchanges (HFTs, specifically) represent 80% of the volume.
- A single large HFT firm (referred to as a Direct Market Maker) can account for 10%+ of a market's volume on a given day
- Large HFT firms make between $8 to $21 billion a year.
- HFT trades occur in milliseconds (i.e., a small fraction of the time it takes your eye to blink).
With such scale, speed, and profitability, HFTs have turned the market away from being an efficient price-setting mechanism and perverted it into a casino where the clientele of human investors gets fleeced.
And our regulators are so outmatched by the scope, complexity, and funding of these titanic HFT players that at moment, there are pretty much zero consequences for bad actors.
Interestingly, these HFT parasites, which live by generating fractions of pennies in millisecond-timed trades, may be sowing the seeds of their own demise through their blind gluttony and hyper-competitiveness. As their quest for incremental advantage begins to bump up against the limits of physics (such as the speed of light), the marginal cost of the next increment of advantage increases exponentially. Profitability is being squeezed out and will disappear entirely some day.
Sounds good to the rest of us investors, right? Not so fast. A key question to ask, should these parasites experience a self-induced mass-extinction effect, is:
What will happen to asset prices when all that volume suddenly disappears?
HFT's Bloodsucking Role in the Financial Markets
There is a host-parasite relationship. The host is the traditional order or the retail or institutional order. They will always lose. There is no doubt about it. The parasites are circling around that host all day long trying to find where they are going to take advantage of them – whether it is a VWAP order (Volume-Weighted Average Price) or something like that. If there are no hosts or the hosts are starting to decrease – because they are based on the mutual fund outflows that we talked about before – the parasites find it hard to make money. There is no more to feed off of. So they start to feed off each other, which means that their margins by definition are going to have to start shrinking until it becomes unprofitable. And it will become unprofitable when all of a sudden they have to invest hundreds of millions of dollars to gain an extra microsecond, yet they are not getting their returns back.
So the real fear that we have is that when it becomes an unprofitable opportunity or venture for them, what will they do? Will they walk away? Who will be left holding the bag? Where did all that “liquidity” go? Who is left now? Hopefully what will happen was the market will find its own solution at that point. But you would have some scary days, I can bet, between now and then.
The Bastardization of "Investing"
If you are an investor and you want to diversify -- which everybody should be doing, right? -- you want to pick different asset classes. You want to get things that are inversely correlated, because that is how you can prevent yourself from taking a large loss, especially if you are a conservative investor.
There was actually a report – I think it was a couple of months ago and it was produced by the U.N. -- they studied the correlation between oil and stocks. And they found it at record levels over the last five years. It just shot up off the charts where oil would normally be a negative correlation with stocks. And they were scratching their heads. And one of the things they pointed to was the correlation effect of the high frequency traders trading multiple asset classes.
So this has been documented now. It is not just us kind of guessing, saying, “Well, I bet it was the HFT’s correlating asset classes.” Everything trades together. That does not make for a healthy market. That does not make me feel comfortable that I can hedge my position right now unless I was just trading around a zero position all day, like most of these guys do in the high frequency trading world.
So what do you do as an investor? How do you diversify yourself? It is very, very troubling and at this point there really is not an answer to it.
The Impotence of Our Regulators
When you are dealing with this type of computing power and this heavy amount of quote traffic as well as trade traffic, the quote traffic is enormous. Every time an exchange tries to update their capacity, it immediately jumps up to the capacity level. It is a constant amount of quotes, trades, and cancellations. They do not have the systems available to track this type of behavior.
So you have got one of two options if you are a regulator. Either a) get up to speed quickly so that you can track this behavior so the investing public could feel confident again, or b) you have to start limiting this type of behavior. There is no other option. You cannot allow this to continue to go on. And by limiting it, something that we would suggest is maybe a real cancellation fee, not the ones that have kind of been suggested. Maybe a minimum order timelife. If I said to you, “Hey, we want a 50 millisecond minimum order timelife, would that be a problem?” And I would think it would not be a problem. Because, guess what? I just blink my eyes and it took me 200 milliseconds to do that. So 50 milliseconds really shouldn't be that big of an issue. And there have actually been reports, studies by academics that have said, “Any order less than 50 milliseconds really does not contribute to any liquidity.” So do not give me that you are going to be hurting liquidity.
So the bottom line is the regulators are overmatched and they need to do something now. And they have really one of two options. And the option of getting up to speed probably is not in their budget right now.
Click the play button below to listen to Chris' interview with Joe Saluzzi (47m:53s):
Chris Martenson: Welcome to another ChrisMartenson.com podcast. I am your host, of course, Chris Martenson, and today I am very happy to be welcoming back Joe Saluzzi to our program. Joe is partner, co-founder, and co-head of Equity Trading at Themis Trading, LLC, a leading independent agency brokerage firm that trades equities for institutional money managers and hedge funds. Joe is one of the foremost experts on algorithmic trading, often referred to as high frequency trading or HFT, and the dangerous risks it has introduced into our financial markets – something very important today. He has a really deep resume in equities trading, and in particular the electronic side of the equation, having headed a team responsible for sales and training hedge fund accounts at Instanet for more than nine years. He has just come out with an excellent book on the subject, Broken Markets: How High Frequency Trading and Predatory Practices on Wall Street Are Destroying Investor Confidence and Your Portfolio.
Joe, it’s a real pleasure to have you back.
Joe Saluzzi: Great. Thanks for having me again. Appreciate it.
Chris Martenson: You know what? I want to start with the dedication in your book. I love this. It says here: “We dedicate this book to the executives at the major U.S. Stock Exchanges, to high frequency trading firms, to lobbyists, and to numerous other conflicted parties in Washington, D.C. and Wall Street. Without your actions we never would have become outraged enough to write this book.
Joe Saluzzi: I think that is pretty much – if you pick that up, you know where we are going to go with this book right away. And we did it – actually my co-author, Sal, who we did not mention yet – Sal Arnuk. We both wrote it together. And we wanted the readers to know right away that this is not going to be a sugar-coated book. This is not going to be the typical thing that you see out there from an industry insider. This is going to be a book that is going to come straight at it and hit them hard and pretty much talk about a lot of the things that we have spoken about over the last three or four years now. And I think that dedication pretty much summed it up.
Chris Martenson: I thought so, too. I liked it a lot. And before we dive into the outrage that is sort of embedded in that quote, let’s start at the beginning and insure that everybody is on the same page. First, I am sure some of our listeners have read the news account of a $350 million fiber-optic connection that is being laid between London and New York for the sole reason that it can shave a few milliseconds off of an equity trade. I want to get to an understanding of how it is even remotely possible that a few milliseconds can justify a third of a billion dollar investment. And that is what we will do today. Plus, I hope, understand some of the benefits, but also what never gets talked about – the risks that our markets currently operate under because of high frequency trading. So before we discuss that investment, right at the beginning, what is it that our equity markets are supposed to be?
Joe Saluzzi: Well, they are supposed to be helping investors, first of all, to allocate capital, helping companies raise capital so that they can grow, right? We want our small companies to get bigger, and in order to do that, they need access to funds, so they usually come to the equity market and investors will, if they like the company, invest in them. That is the point of the whole stock market, is to have these companies grow so that they can increase jobs, so they can help the economy. Well, somewhere along the line I think we kind of lost that. It seems like over the last 15 years we have turned it into no longer a capital raising function and an economics function, and now more of a, “Hey, I can beat you faster than the other guy” type of trading-casino mentality, which really does not help anybody. Right now we are looking for investors to come back into the market, and all we see is week after week after week, money coming out of domestic equity mutual funds. And obviously part of that is the financial crisis ongoing for years now in Europe as well as our own country, but a lot of it is also due to the fact that people do not trust the equity markets any more. Investors are frustrated. They are tired of watching the games and the flash crashes and IPO’s failing and constant movements in these stocks with no explanation. And they are wondering what is really going on in this equity market.
Chris Martenson: Well, it used to be that fundamental analysis counted for something, and maybe it does over the long haul, but over the short haul it does not seem to count for much anymore. You note in the book that – in the book it says, “Some 232 billion has been withdrawn from domestic equity markets mutual funds between May 2010 and January 2012.” So a quarter of a trillion dollars up and out and gone, and yet the market still seems to power higher and go places and do things. How is that?
Joe Saluzzi: Yeah, it is an amazing calculation, right? You have money coming out, yet the market still goes higher. I think part of that is – which is not the subject of our topic, but it is maybe another day – that part of that is obviously the free money mentality of the Federal Reserve when they are constantly keeping rates at 0% trying to force people into a higher, riskier asset. But as we know, money is coming out, so the true investors are not buying that argument. So what is going on here? There is this giant correlated game as well. One stock, if you noticed – I am sure you listeners trade multiple stocks – nowadays everything seems to move together, whether it is stock, currencies, commodities. Everything is one giant correlated kind of asset class and you have to wonder, well, how does that happen? Well, obviously when humans were trading that could not happen, because we could not react that much. But now that computers pretty much have taken over, when there is a move in the e-mini contract, well, everything moves alongside of it, all the offers – a bid comes in the e-minis and all the offers get cancelled, and the underlying stocks. And the next thing you know you get those flickering quotes constantly all over the place.
So you do not really have that fundamental feel anymore until a news story breaks. And then, all of a sudden that correlation will break, particularly on one name or a news item. So maybe you have a stock that was trading at – I do not know – maybe $15 or whatever for many days, and then all of a sudden you are wondering, why is this stock even up here in the first place? And then the news story breaks and it gaps down five bucks. And everyone is like, “Well, what happened? What is going on here?” Well, it is kind of being held up in that correlated ETF kind of one giant market that we have right now.
Chris Martenson: Exactly. I want to get back to that if we have time, because that correlation is extremely important, I think, to some of the risks that are embedded in this. Before we get there – high frequency trading, what is it?
Joe Saluzzi: Well, that is the topic of an entire group now that actually I am a part of, on this CFTC committee where they are trying to define high frequency trading, and it is a difficult thing to define. What is it? Well, obviously it uses computers, and co-located is a term that people often use, and that just means that they take their servers and basically place them in the same room as the exchange server. So if you wanted to get an edge, physical location is extremely important. You need to cut down the amount of distance, which is kind of what you were alluding to before with the cable. And we will talk about that in a second, I guess.
But you need to be as close as possible to that exchange server. So what the exchanges have figured out was, “Hey, we can rent space in these giant co-location facilities and charge whoever wants to buy it.” So, in other words, charge for premium access. That is just one thing that they do.
High frequency traders are also – they will put in lots of orders. They will place lots of cancellations. It is all about speed and there is usually no human intervention. There is no human intervention. It is a machine based on someone that is created an algorithm for that machine, but it is all about speed of access to the markets, whether it is reading the data, and then entering the orders into the equity system or into whatever system you are trading on. And then also cancelling them when moves start that you do not like.
Chris Martenson: So the essence of high frequency trading is that you are trying to move into and then back out of existing positions. I assume that the high frequency traders are not there to accumulate positions.
Joe Saluzzi: Well, that was actually one of the dissents that we had. Being part of this committee, there was actually an open forum last week. The definition was read and we actually had a public dissent on that part, because there was nothing in the definition that talked about what we call inventory management or flipping of positions. The definition pretty much reads as if only an institutional money manager could fall into that definition right now. But we wanted to include more of – well, if you are constantly flipping from back and forth, similar to what a market maker used to do to manage inventory, that is more of a high frequency trader, as opposed to the traditional long-only or even a retail trader or something like that. But it is a constant inventory flipping and it is not just stocks. Here is one thing that a lot of people kind of miss is there are multiple asset classes that they are trading. So they can be long e-minis and short ETF or vice versa, or they could be something in the currency world. It is a giant – it is not just one class. So the problem that the regulators have is, “How do we identify what is going on?” They cannot even identify one asset class, let alone cross asset classes.
Chris Martenson: I am sure other people who actively trade have had this experience and I am wondering if this is connected. Up through about 2007 I had a system that worked really well. I was trading futures – gold and silver primarily, but other commodities. And then starting around 2007 I would notice things that really drove me nuts where I would have a position in and I would have a limit sell or a stop sell way under my position and I would watch this finger come in and just reach down and touch my position and take it and go away again. And I assume a human could do that. But it would happen so fast and so regularly that I realized that I was playing against – it felt to me like the rules had changed and I was playing against something inhuman at least. And I know that can happen with humans, too, but the speed with which it would happen was just startling to me. And I realized I could not compete.
Joe Saluzzi: Yeah, I think a lot of investors or a lot of traders, let’s say – because people do make – there is a big distinction if you are a long-term grandma’s mutual fund, they are not trading that particularly. But if you are a short-term trader, you are going to feel that. And I think that is what a lot of professionals have felt is that, “Hey, I cannot compete anymore.” Even to the point of not just regular professional investors like ourselves or retail or institutional, but even the Quants. And there is a new book – and I will plug Scott’s book as well -- Dark Pools by Scott Patterson, which I think is a fantastic read. But he talks about in the first few chapters – he is talking about a Quant who basically said the exact same thing that you just said, “I could not believe how my systems were – it was not working anymore. I was not making the P&L I used to get.” Until he finally figured out that he was basically being beaten to the punch every time by high frequency traders who had access to different types of orders, access to data feeds – all this premium type of service that he was not doing and essentially he basically got put out of business.
Chris Martenson: Well, one of the things that I was taught in business school with the efficient market hypothesis is that there are no asymmetries of information, that those would get competed away. But what you have just described where I get to see a tiny corner of the market – like I am peering through a pinhole – but somebody else gets to see the whole room. They have the entire order flow. They get to see the entire book. They get to see where all the position limits are set. They are not playing with the same information that I am or even this Quant is, are they?
Joe Saluzzi: No, not at all. And you can – the other side would argue that the old-time specialists also had similar information. We would kind of come back with the argument by saying, “But, yes, the specialists also had affirmative and negative obligations. They could not run ahead of a client. They could not do certain things.” So with those benefits came obligations, as opposed to the market now, where they are getting all these benefits but no obligations or very little. I mean, the obligation for a market maker and stock is they have to be within 8% of the NBBO, which his ridiculous. So a hundred-dollar stock, they need to be a $92 bid to offer at 108. I mean, come on. What kind of obligation is that?
But now the other side again – and I will try to preempt some of their arguments – they will say that, “Well, anybody can get access to this, Joe. You are just complaining because you do not want to buy the co-location. You do not want to hire the sophisticated programmers and you do not want to buy those proprietary data feeds that the exchanges will offer.” And will say, “Yeah, you know what? You are right. I do not want to spend millions and millions of dollars trying to shave a microsecond off the trading process when I am really a long-term investor just trying to get the best possible price for my client.” It makes no sense that it is becoming an arms race and whoever – it is an arms race -- and whoever spends the most money will win. But the problem that they are having now is they forgot about one important factor, and that is the speed of light. It is actually bumped up against the speed of light.
It is now going to get interesting. Who wants to spend all those millions now for an extra microsecond? We will see how they go with this one.
Chris Martenson: Well, so maybe the bulk of the arms race is over, I guess, because the speed of light is hitting there. Let’s do some numbers then. High frequency trading accounts for how much of the volume on the exchanges each day?
Joe Saluzzi: Well, estimates range between 50 and 70%, and we have seen those estimates for many years now and most people will agree with that. That is in the U.S. I just saw an estimate today. Canada is very similar. So it is not just a U.S. kind of situation. But 50 to 70% of the trading volume that you see is that of HFT machines. Now, sometimes they are trading with themselves. Sometimes they are trading with investors like ourselves as well.
Chris Martenson: Okay. And how many companies are out there – I assume these are individual entities we could point to and say, “That is an HFT company right there.” I assume. Is that right?
Joe Saluzzi: Well, yeah, if you could see their one-page websites, which most of them have, they do not tell you much. They are proprietary trading firms, so they are not market-makers in the sense of traditional, or they are not traditionally-regulated market makers. They could pretty much do what they want. And if they want to put up a one-page website, that is what they do. But the numbers that we have seen [show that] 2% of the actual traders are representing 80% of the volume on these exchanges. So there is a giant amount of business being occupied by only a few of the bigger – of the high frequency trading firms.
Chris Martenson: Oh, okay. You have one statistic in here, then, which makes sense now. You say, “A few of the larger HFT’s account for more than 10% of any given day’s trading volume.”
Joe Saluzzi: That is right. So you can take one of these large firms, whatever – there are a whole bunch of big ones. They call themselves DMM’s now, or Direct Market Makers. They are actually sitting – they are on the floor of the exchanges. But they will – those bigger ones will represent at least 10% of the volume, if not more on certain days.
Chris Martenson: And so the volume is something I used to track very closely and then once I understood what HFT was up to, I started to follow it less closely. In that volume, so if you are an HFT and I am one and for some reason we are flipping back and forth trying to best some poor institution out there that is trying to get a few shares off, will our trades with each other – will those count?
Joe Saluzzi: Oh, yeah. They will go to the tape and there is actually – they are talking recently – people are talking about wash trades as opposed to – it is possible that some HFT’s are trading with themselves, not only another one. So if that constituting a wash trade? So maybe they have one strategy trading one way and another strategy trading another way. They could actually be trading with themselves. This is something that the FCC is looking into and this is something that has been in the press over the last few days. But you would think – and this is very basic – cannot the FCC just say, “Yes, we see this trade”? They problem is that the end user is not tagged. You need to tag them. Every trader needs to be tagged. It is a very simple process. And if you are not doing anything manipulative, you really should not have a problem with it. I should have an ID number and when I trade in the system, they should be able to know what I am doing. And very easily I should be able to run a report and say, “Well, this guy just traded with himself and he has done it a million times today. So, guess what? There is probably a problem.” But that is not available right now
Chris Martenson: What do you mean? You mean – I did not know this. The trades are not tagged.
Joe Saluzzi: Well, no, you go through a sponsoring broker. But your broker – let’s just say you are an HFT proprietary firm. Well, you cannot access the market directly because you are not a broker. You have to go through a sponsoring broker. So let’s just say your broker is XYZ broker. The trade shows up on the tape or in the clearing field as XYZ. It does not show up as the end user so XYZ is the buyer and XYZ is the seller. Well, that happens all the time because you can have multiple clients where they are on different sides of it. But when it is the same counter party, that is where there is a problem. That is where they are talking about these wash trades. That needs to be investigated. That is something that the FCC has talked about a consolidated audit trail. That needs to happen. That should have happened years ago. Two years ago they proposed this, which would basically be this kind of eye in the sky to watch everybody and they proposed it unfortunately as a dynamic system. In other words, constantly being monitored, they can kind of get it on the fly. Well, that would have cost them $4 billion. And as soon as everybody heard $4 billion, they ran away and they said, “Oh, no, this is too expensive.” Well, you do not need to do it on the fly. You can have a one-day lag as long as you are investigating it. And the bad guys – let’s just say there are bad guys – know that the police are watching. Guess what? That behavior will probably stop by itself. But when the bad guys know that there are no police, well, their bad behavior is kind of incentivized.
Chris Martenson: Well, I can think of one thing right off the top which actually did impact me when I was – one of the things I did was I rode certain mortgage insurers and home builders down during the crisis and the crash and all that. And as a short seller I would watch the tape very carefully and when all of a sudden I would see volume flooding in and the price going up, I would bail often on my positions. So if you have the ability to suddenly create the appearance of volume by washing with yourself that would have been what I would have considered market manipulative behavior.
Joe Saluzzi: Absolutely. I mean, that is – there is no doubt that is manipulation. But if no one is catching it, that needs to be identified. Just like with, say, the news that broke with the LIBOR price fixing.
Chris Martenson: Yeah.
Joe Saluzzi: With Barclay’s. Well, that is market manipulation and the behavior that we are talking about right now is market manipulation, if it is going on. Again, we do not know. This is speculation. But if it is, you would think that our regulators would know it and would crack down on it heavily.
Chris Martenson: Listen, you can make money by manipulating volume flows and front running what you create in terms of market behavior next. So if you can do it and make money at it, it is being done, that is my un-initial hypothesis. It always seems to be true.
Joe Saluzzi: Yep.
Chris Martenson: So one last statistic then. You say here in the book that HFT’s earn anywhere from $8 to as much as $21 billion a year.
Joe Saluzzi: Those are also based on some industry surveys. The initial survey was done on – I think it was Tab Group that we were quoting from. We have seen surveys afterwards – it could be three to four billion. It is hard to predict. It is certainly a volume-based number. So just kind of take the number of shares that are outstanding and apply a margin to that or a profit margin to it. Now their margins have come in. There is no doubt about it. The HFT’s are competing against each other and they are knocking down their own margins so their profitability may be suffering lately, but that does not mean that their activity is suffering. They are still out there more than ever. Because now if you are only making, let’s just say, a tenth of a penny as opposed to two-tenths of a penny, well, you are going to have to trade twice as much in order to make the same amount of money you were making you were making before.
Well, there is not enough – the way we look at it – and we made this comparison before. There is a host-parasite relationship. The host is the traditional order or the retail or institutional order. They will always lose. There is no doubt about it. The parasites are kind of circling around that host all day long trying to find where they are going to take advantage of them – whether it is a VWAP order – a Volume-Weighted Average Price – or something like that. If there are no hosts or the hosts are starting to decrease – because they are based on the mutual fund outflows that we talked about before – the parasites find it hard to make money. There is no more to feed off of. So they start to feed off of each other, which means that their margins by definition are going to have to start shrinking until it becomes unprofitable. And it will become unprofitable when all of a sudden they have to invest hundreds of millions of dollars to gain an extra microsecond, yet they are not getting their returns back.
So the real fear that we have is that when it becomes an unprofitable opportunity or venture for them, what will they do? Will they walk away? Who will be left holding the bag? Where did all that “liquidity” go? Who is left now? Hopefully what will happen was the market will find its own solution at that point. But you would have some scary days, I can bet, between now and then.
Chris Martenson: So what occurs to me is that once upon a time we had to contend with the bid and the ask and that was the spread we had to deal with and now it looks like we have a HFT vig to add to that. So there is the bid to ask and the vig. And so we have got these high frequency algorithms, they are all circling around; they are doing what they do. And I want to talk now a little bit about maybe some of the risks that are embedded with that. So let’s go back to our favorite teachable moment, which was the flash crash. In your mind, what really happened there?
Joe Saluzzi: The flash crash – they will claim – the HFT proponents will claim that it was caused by a Midwest mutual fund manager who entered a large e-mini contract, which is absurd.
Chris Martenson: Yes.
Joe Saluzzi: It certainly was not caused by that. Now that order came into the system and obviously that was the host – okay, going back to the host-parasite. The host came in and the parasites had a feeding frenzy on it. But something went crazy along the way and they kind of got caught in a feedback loop. And this has been documented by the CFTC and there is an economist there named Andre Kirilenko who talks about this” hot potato effect.” That is exactly what happened. They started going back and forth real quick and, obviously, there was news out that day and the Greek riots, which nowadays we have become accustomed to almost every day. There is always news in Europe. But this kind of activity sent them into a free fall. It created a vacuum. And this is something that we have always predicted would happen. When they started to see this, the data started getting corrupted. And if you ask any high frequency trader, “What happens when data is corrupted?” They will say, “Well, immediately I have to manage my risk which many times means that I have got to pull out of the market.” Because if they do not trust the data, their models do not work. And once the models do not work, they have to exit because of fear of – who knows? They may actually lose money that day, which is unheard of in their world.
So they all pulled out and there are a couple of them that have documented themselves – I should not say they all – there are a few of them that said, “I exited the market.” One guy actually said, “I typed in the code “HF Stop,” which basically meant pull out. So they pull out. Those are some of the market makers. And then the next problem was the internalizers were people that basically – these are broker-dealers who will basically take the retail order that is on its way to the market and intercept it most of the time. That is when you get these sub-one penny price approvements when you are trading in your Ameritrade account. Say the stocks offer ten cents and you get a fill of 0.9999. That is an internalizer trying to take advantage of your order.
Well, the internalizers which would normally make up a large percentage of volume, decided also to walk away because they did not know what was going on. So they shipped all the fall that was coming in off to the lit market, which was the exchange-driven market, which normally is considered exhaust as they call it. They could not handle it. There was nothing there. So if you ship a market order in when there is no bid, guess what? The market order just goes and looks for the next best bid. And in some cases like Excentra, which was the poster boy – it traded at a penny a share because that was what they called a stub quote back then.
So all of these orders basically got flooded into the market. All the “liquidity providers” ran for the hills and the next thing you know we are down a thousand points. Until someone says, “Wait a second, this does not make any sense.” And there was a CME stop logic function that supposedly came in and stopped things, and that is debatable. But anyway, the next thing you know, the market zips right back up. And to us the biggest problem with May 6 was not so much that it went down, but then it came right back up at almost the exact same speed. That tells you there is a massive, massive structural issue here which still has not been addressed, which basically tells us that you can still get one of these – even though there are types of circuit breakers in there. And it is the speed of the drop, right? The speed of the decline and the speed of the ascent is really what shook investor confidence.
Chris Martenson: Well, it sure did for me and now I am hearing you years after the fact talking through a post-mortem of what may or may not have happened and it sounds like there are still disagreement. And yet we find that the degree of regulation is thin enough that we still cannot accurately say who is posting orders and whether wash cycling is happening in some of these trade flows. So is it fair to say that this is still, regulatorily speaking, a fairly opaque market with low amounts of oversight?
Joe Saluzzi: They are – let’s just say that we would think that the regulators are overmatched in this situation. But when you are dealing with this type of computing power and this heavy amount of quote traffic as well as trade traffic, the quote traffic is enormous. Every time an exchange tries to update their capacity, it immediately jumps up to the capacity level. It is a constant amount of quotes, trades, and cancellations. There is – they do not have the systems available to track this type of behavior. So you have got one of two options if you are a regulator. Either, a) get up to speed quickly so that you can track this behavior so the investing public could feel confident again. Or, b) you have to start limiting this type of behavior. There is no other option. You cannot allow this to continue to go on. And by limiting it, something that we would suggest is maybe a real cancellation fee, not the ones that have kind of been suggested. Maybe a minimum order timelife. If I said to you, “Hey, we want a 50 millisecond minimum order timelife, would that be a problem?” And I would think it would not be a problem. Because, guess what? I just blink my eyes and it took me 200 milliseconds to do that. So 50 milliseconds really should be that big of an issue. And there have actually been reports – studies by academics that have said, “Any order less than 50 milliseconds really does not contribute to any liquidity.” So do not give me that you are going to be hurting liquidity.
So the bottom line is the regulators – they are overmatched and they need to do something now. And they have really one of two options. And the option of getting up to speed probably is not in their budget right now.
Chris Martenson: Well, it is also, I think, a question, too, that it is a high complex system, meaning it is nonlinear. There are feedback groups embedded, there are multiple exchanges, there are lots of computers, there are lots of competing algorithms, there are tons of firms, there is this enormous flow. It is like trying to calculate what happens when a wall of water goes over a cliff and how is it going to break? It will do it maybe a little bit differently every time. And so I see a system that is extraordinarily complex. And if the regulators are overmatched, I do not know what you could do in another flash crash besides pull the plug. Like that guy typed in that command, which basically pulled the plug on his own algo’s.
So we had the flash crash. I know there have been some meetings. You have been involved in some of them. There are some proposals. Has anything really been fixed? I know there were some new circuit breakers put in. Are those providing us any sort of a bulwark against what might happen if we got to another flash crash sort of set of conditions? Where are we?
Joe Saluzzi: Well, there have been some, what we call, band-aids put in place. One was they removed the stub quotes. Okay, so that is illegal. So, in other words, the country will not be able to trade at a penny a share now. Okay. That is a little bit of a start. Then they put in what they call single stock circuit breakers which basically meant that if a stock were to drop 10%, within five minutes the stock would be halted at that point. But then market participants were complaining about halts that were being caused by erroneous trades. So a new proposal which just got approved is called Limit Up Limit Down so halts will now work more similar to the way the futures markets are, which makes sense. I mean, that is okay. But it still does not prevent a flash crash event. Because as we talked about before, the problem with the flash crash was the speed. So if I told you that we could drop 10% now before it gets halted and that 10% could happen in 30 seconds. Well, that is going to scare the heck out of everybody. You have done nothing to prevent that 10% drop. Yeah, okay, maybe you cannot drop 20% because everything will kick in and pretty much shut down. But you can drop that quickly in the blink of an eye. Look at the Facebook IPO, right? There are situations out there where this can happen.
So they have not really addressed the major problem and it is kind of what you were getting to before about the system. The word that we use all the time is “fragmentation.” The equity market is this fragmented web of different market centers, whether they are dark – being the dark pools – over 50 of them out there – or the lit pools – over 13 exchanges plus ECN’s on top of that. They all kind of interact with each other. And any failure at one point in the system can cause problems anywhere else. So you really are relying a lot on these computers and technology and when there is a failure, things break down quickly. So you have to say, have we created too much of a complicated situation where you took the very, very simple process of buying and selling a stock – right? It is very simple. You want to buy them; I want to sell them. Let’s get together and put up a price and be done with it. No. Now it is this entire web where it has got to get routed through smart order routers, sending with indications of interest to various dark pools who could either execute it or decide to pass it along to somebody else, maybe internalizer the steps in between. By the time it gets over to its final destination it has been touched, sniffed, and processed more than a TSA person at the airport. It is ridiculous. We really do not need that in an equity market.
Chris Martenson: Well, then let’s talk about what happens if this liquidity is both there and then can evaporate like water on hot steel. What happens then? So you mentioned something that raised my antenna. You said that sometimes there is just the normal flow and somebody drops – like this so-called fat finger trader out of the Midwest dropping all those e-minis on May 6, 2010. But there was also news in that cycle. So what happens if we get some really big news cycle that really scares these or is outside of the boundaries of these algo’s? I do not know. Let’s pick something. Israel attacks Iran tomorrow. The Strait of Hormuz gets closed. Something big, right? Really captures everybody’s attention. Is there a risk that this 70% of the volume of the market could basically just up and flee again?
Joe Saluzzi: Absolutely, because there is no obligation. Now, again, the other side will tell you, “Well, it is nothing different from the old days when the specialists were there because they could have walked away.” Not quite. It was a bit different. Nowadays they can pretty much press a button and everything evaporates at the instant. At least when you had various specialists and market makers, there were different timing spots, there were different locations where stocks traded, they all did not pretty much act as one. You can press a button pretty much right now and things would start to evaporate before you can blink. And then those market orders, if they are out there, will start to chase prices down the ladder. It could easily happen in an event like that.
And, by the way, the flash crash – that Midwestern mutual fund – it was not a fat finger. It came in there as designed. It came in to say, “I want to be X% of the market. I am in the E-mini’s, which is the most liquid contract in the world and I want to sell a certain percentage.” And they only sold 35,000 contracts on the way down in the first 15 minutes. Nowadays you can see that in seconds. So it was not that big of a contract. It was that it kind of spooked a few people and the next thing you know you got caught in that feedback loop that we were talking about and the hot potato effect started. That could easily happen again. There is nothing to prevent that.
Chris Martenson: Oh, yeah, I have seen orders bigger than that hit the tape since. Now to tie this back in, you mentioned something else that really caught my attention and I have talked with people who have been either in the equities business for decades or, in one case, a gentleman who ran a very successful hedge fund for a long time. And they sort of said, “Well, the world has changed. What used to work, does not work.” And one of the things that is really war on this hedge fund individual particularly was this degree of asset correlation – that there used to be some alpha in being able to ride across the asset classes and there was some diversification that was offered by that. And now it seems like everything is sort of trading in the blink of an eye, risk on, risk off, across all markets globally, all asset classes until something sort of goes out of whack. So is that just the way the world has gone? Is there anything wrong with that? And if so, what is the risk?
Joe Saluzzi: Yeah, it is extremely troubling. Because if you are an investor and you want to diversify, which everybody should be doing, right? You want to pick different asset classes. You want to get things that are inversely correlated because that is how you can prevent yourself from taking a large loss, especially if you are a conservative investor. There was actually a report – I think it was a couple of months ago – and it was produced by the U.N. I think it was called UNTAG – I forget exactly the group. But they studied the correlation between oil and stocks. And they found it at record levels over the last five years. It just shot up off the charts where oil would normally be a negative correlation with stocks. And they were scratching their heads. And one of the things they pointed to was the effect – the correlation effect of the high frequency traders trading multiple asset classes. So this has been documented now. It is not just us kind of guessing, saying, “Well, I bet it was the HFT’s correlating asset classes.” Everything trades together. That does not make for a healthy market. That does not make me feel comfortable that I can hedge my position right now unless I was just trading around a zero position all day, like most of these guys do in the high frequency trading world, and they just kind of long-short-long-short-long-short. So what do you do with an investor? How do you diversify yourself? It is very, very troubling and at this point there really is not an answer to it.
Chris Martenson: Well, if that diversification is not there then let’s talk for a second then what any of us can do. I guess there are a couple of things. One is to try and change the system so that we can have – oh, I do not know – a quarter of an eye blink delay in the order flow. Or so we can work towards those regulatory changes. But until those happen to protect the market, how do people protect themselves in this world?
Joe Saluzzi: Well, you have to be careful. Obviously you have to take control of your orders. There are different ways of looking at it. As a retail investor, certain things that we mentioned actually in the book – in the back of the book – we talk about routing strategies that some of the online brokers use. They will usually send them through a smart order router. And let’s just say you put in a stock that has a nickel – say the stock is offered at $10 and you say, “You know what? I will pay up to $10.05, send it through the smart order router.” Well, that order is kind of getting passed around like I was mentioning before, amongst different destinations – sniffed, and tagged, and everything else – and you are probably not going to get the best execution you can. We would rather go through the particular – some of the exchanges or some of the regional brokers will allow you the option of dropping down and going direct to a particular exchange. We think that will get you at least a better feel of getting to the destination you want to be at.
Another thing you do not want to do, which sounds obvious, is do not put in market orders. When you put in market orders, you are just looking to get killed. Because what will happen is by the time that market order comes through the tape, they are going to see it coming through, the guys who are on the offering really are not on the offering, they are going to disappear, they are going to walk away, they are going to take the stock that is ahead of you. This is all happening in a few milliseconds, by the way. And what stock that was originally offered at $10 is now going to be offered at $10.05 and you are going to get your fill 10.05 and then the stock is going to go right back down to $10. Likely – not every time, but that happens. So do not put a lot in a market. Put in limits.
From the institutional side you really need to know where these smart order routers are going. You need to know. What happens is a lot of institutions trade with algorithms themselves and these algorithms use smart order routers that basically ping and sniff different destinations as well. You need to direct that to the destination that you feel comfortable with. But, again, that is very difficult, too, because not only is it the destination that is the problem, it is the types of orders that certain exchanges are using to get to these destinations. I mean, it really is a very, very complicated web and the plumbing, as we call it, of the stock market – it takes a lot to understand. We try to get to it in the book as much as we can when we talk about smart order routers, we talk about dark pools and how some are good and some are bad. But it is very difficult to kind of know where your orders are going to end up and how to protect it.
And with that said, we are starting to see a few signs that regulators are starting to act – unfortunately not in this country. Okay?
Chris Martenson: Um-hmm.
Joe Saluzzi: The Canadian market has already put through something called the Trade Route Rule which will be taking effect in October, which basically says that a Dark Pool cannot trade ahead of a Lit Pool, which makes total sense. Because if you go out first and you put a thousand shares on the offering at $10 and a buyer wants to come in and pay that $10, well, right now you may not get that offer. You may not get the trade. You display the price liquidity, but an internalizer could come in and sub-penny you and give that other guy a meaningless price approvement. But you were the guy who set price. How is that fair to you not to get the trade? So things like a Trade Route Rule with a minimal price approvement is something that we think is probably in the right direction.
Even a currency market just yesterday decided – one of the brokers in the currency world called EBS said that the fifth decimal place on currency trades is no longer going to be allowed and it is only going to be allowed in the mid integer. So 1.234555. So that is gotten to the point – that is how crazy this is. We are talking about fifth decimal points in currencies. People are starting to react, I guess is my point and the reaction is because the public is starting to complain. The members of these exchanges are starting to complain. The institutions, the corporate issuers have had it up to here with the flash crashes in their stocks. Everyone wants to know why is the market so screwed up and what are you doing to fix it? So I think what is happening now is the exchanges are realizing this and even the guys like Dunkin Niederauer of the New York Stock Exchange appear to be making a little bit of a pivot here. Appear to be saying, “Okay, we hear you and we are going to start to make some changes.” And that is how we think this will be fixed. A market-based solution, not a regulatory solution, because we know that the regulators will come in and screw this up just like they did for the last 15 years. A market-based solution is the right way to go.
Chris Martenson: You have done a just a remarkable job explaining the complexity, but it should not be that complex. And even still you are hopping and skipping over the ways that an order might be routed and tagged and intervened and sniffed and all kinds of things that can happen between here and there, which is just “I am trying to sell a stock and buy a stock. That used to be so simple. A guy in a blue suit. I do not know.” And so the old saw is if you are at a card game in Texas and you do not know who the sucker is at the table, it is you. And that was why I walked away from the market a couple of years back – just gave up trading. I felt like I did not understand what was happening anymore and it was true. That was a good decision on my part. I did not know what was happening. Yet there are a lot of people who still have to participate in markets. You have got pension funds, mutual funds. There are a lot of reasons why they are in there. Are they just hostage to this now? How does a company or a big institutional fund participate now?
Joe Saluzzi: Well, there are things – I mean, there are certainly combative ways to fight. And, obviously, our firm, what we do for a living is we trade for institutions. The way we earn our living is to trade for institutions and to make sure that they get a better execution than if they were just blindly putting it into the market and hoping that it got done the best way. So you have to kind of know some of the tricks. You have to pay attention unfortunately and we think it requires a lot of manual work. You have to really look at stocks, feel them, talk to you clients. The old fashioned telephone actually does work still and we could pick it up and talk to our clients and we think that adds a lot of value. But for the regular investors it is really taking control of your orders more. And unfortunately the manpower unfortunately on a lot of these trading desks has been reduced because of cost cuts over the years and maybe a trading desk which used to have ten traders now has one or two. Well, the workflow for one or two guys is obviously way too much for a person to handle, so maybe they outsource it to firms like ours or maybe they put it into some of these algorithms. But that is where you have to be careful. That is where the plumbing of the algo really needs to be investigated. You need to do the transaction cost analysis afterwards that identifies if there was a pool out there that was getting you an adverse selection.
In other words, say every time you traded with one Dark Pool, the next thing you know the stock went up by a nickel or a dime or something like that. Well, you know that that Dark Pool is probably leaking information on you and it is very difficult to find out how they are doing it. But at least if you know that there is one particular pool that is considered toxic, we would extract that from our routing table. And we are constantly watching as to who we trade with. You have to be aware of the other side and really kind of dynamically change it all the time.
Chris Martenson: Okay, so another flash crash is possible, right?
Joe Saluzzi: Possible and likely in our opinion.
Chris Martenson: And likely. Describe what if we had one -- presumably some rules got put in place, what would this next one look like if it happened?
Joe Saluzzi: I think it would happen faster and the question would be what point of the day did it happen? If it happened when it was towards the close, well, that would be even scarier. Because if you stayed down there, then you would have a global market turmoil overnight. But it will probably happen very similar to the last one when there will be news in the markets, whether it is European news or something very unexpected. Things will gap quickly. Buyers will disappear like they always do because they do not need to be there until you reach the limit up-limit down type of mechanisms which are not in place by the way until next year. So let’s just say you are operating under the current single stock circuit breakers, stocks would not stop halting until they are around 10% and that would have to be one by one. Right? So each stock would have to plug until you finally got to that point.
It would be fast, but it would be fast and furious, if you want to use that term, and it would be another confidence-shattering event. Because there is nothing in there right now – there is nothing that the FCC has done that tells us that this cannot happen again. A single stock circuit breaker is not going to prevent what you saw on May 6 of 2010, which basically tells us that you have a house that was built on a poor, poor foundation and if you drive by that house on a sunny day and you look at it and you say, “Oh, okay, this house looks nice. Maybe I will take a peak.” But if the wind came through and a storm came through, that house is going to fall pretty quickly because the foundation was not fixed. And that is exactly what the equity market is – it is a house with a poor foundation that was never fixed.
Chris Martenson: That is a great description. Now I am going to ask you a question and if you cannot answer it, I will just cut this part out. But a bit of speculation – so this so-called fat finger trade out of the Midwest where a pretty ordinary sized bunch of e-mini seller orders came in and drove the market down. It sounds like the opposite could happen, too, that if we have these parasites circling sniffing for any hint of liquidity that let’s imagine there is an interested party that does not like the stock market’s direction, wants it to go up, it sounds like they have got a perfect mechanism now to just go in, do a fat-fingered upward trade, just put in a big old market order for a bunch of e-minis, and the market would go higher. Is that true?
Joe Saluzzi: It could happen. I mean, people will refer to that as momentum ignition type of trade, which is illegal under the FCC’s definition of momentum ignition. It is kind of sparking the market for your own intentions. It is actually making a very good point. Markets go up, markets go down. For some reasons when markets go up, it is not considered manipulation.
Chris Martenson: Right.
Joe Saluzzi: But when they go down, it is manipulation. And if you are a long-short HFT, it does not matter which way the market goes. All that matters is that it goes in the direction you want it to go. So absolutely things can happen during the day where a lot of times – we are friends with the fellows over at NANEX, which I highly recommend their blog. They do a lot of good work and they have the real data to crunch and they put out some pretty interesting charts. They pretty much will see when these quote spikes come through and a lot of times quote spikes are often at the same time of a price spike. So whether or not someone is intentionally jamming a channel, which is part of the theory – in other words, I can jam through all of these quotes. I will be able to slow down the quote for everybody else. I can take advantage of that myself. Let’s just say I was a nefarious trader. Buy up everything before you even see it and the next thing you know everyone starts to chase it. That happens or at least that is the theory with these quote spikes and they refer to it as quote stuffing. That needs to be investigated. This is something that is very basic again, but if you cannot identify who the end user is, you will never figure it out.
Chris Martenson: Well, I saw these a lot and they happened at key support moments. I remember one famous moment when the DOW was approaching 10,000 again and it stopped at 10,000.48, but there was this extraordinary premium shift that happened on the S&P Futures and there was just an extraordinary buy order that just sort of flooded in. I looked at it and I was a little suspicious. I was like, “Well, that was a bold move right there.” But it certainly seemed – it did not seem like the collective actions of a whole bunch of individual investors – I put quotes around that word. Somebody stepped in and did that.
Joe Saluzzi: Well, the invisible hand, right? It is something that we cannot say so I cannot –
Chris Martenson: Understood.
Joe Saluzzi: Right. But if it is going on and there is not a regulator to look at that because they do not have the proper tools, well then we all have a problem, whether it is going to the upside or to the downside.
Chris Martenson: Right. I am not going to speculate as to who or what or the motives, but if we can jam the market down through a sell order, we can probably jam it higher. It is just that the structure of the market now is that we have a whole lot of very aggressive, very fast moving, liquidity-chasing machines and any removable or dumpage of liquidity is going to cause them to do what they do.
Joe Saluzzi: That is right. They are always looking for signals. It is all a signal-based thing. If everyone is kind of reading the same signals, they kind of react and they have said it themselves that many follow the same strategies. So they are all kind of reacting to each other which is what causes these feedback loops.
Chris Martenson: Right, right. Yeah, we have seen that. That is part of our market now. And that is the increased volatility, so –
Joe Saluzzi: Right. And that is not what an equity market is, right? Going back to the core question again.
Chris Martenson: Right.
Joe Saluzzi: What are we doing? What are we doing? How do we feel comfortable now? We are trying to make small companies grow to help the economy, to help our jobs picture. That is what an equity market is. How did this become a giant momentum, ignition-type, “stuffing” ridiculous game? So it is going to take – and unfortunately guys like us are called Luddites and turtles and buggy whip manufactures – I have heard them all, okay. I am sure you have as well.
Chris Martenson: Yeah.
Joe Saluzzi: But sometimes it takes someone to step in and say, “Hold on. Think about what you are doing here. Let’s just think about the goal of what we are trying to do.” We do think the biggest missing voice in this argument is the corporate issuer. Those are the folks who benefit from the equity market. Those are the folks who are raising capital, who really need to kind of come out now and say, “Wait a second. I do not like what is going on with my stock anymore. I do not trust the participants who are trading in my stock. I do not even know who is trading my stock.” Back in the day of the specialist at least they had some color. There is no color. Go ask an IR person at a major company, “Do you know what happened to your stock today?” And nine out of ten times they are really not going to know. So those are the folks we want to see back in this debate. Hopefully we can stimulate them a little with the book, as well as some of the other guys that are out there, like I mentioned Scott’s book and so on. But we need to have a more of a debate here. And, like I said, you have got to kind of step back and say, “Wait, what is going on?”
Chris Martenson: Fantastic. Well, thank you so much for everything you are doing to illuminate this fairly seemingly arcane, but extremely important area. And the book is Broken Markets. It is by Sal Arnuk and Joseph Saluzzi we are talking with. Joe, obviously, and it is just fascinating to me and I think everybody should focus on it. It is just – it is that important. Otherwise we lose our capital markets, we will lose our edge. Already retail money is fleeing the markets, maybe for a variety of reasons. But one of those, in my case at least, is broken trust.
Joe Saluzzi: Chris, thanks for the time and really I do appreciate it and I do hope the message gets out. I certainly urge your listeners to get involved in the debate, send letters to the FCC, write your common letters, write letters to the editor. That is how things will get changed. But we need everyone to kind of get involved in this.
Chris Martenson: Fantastic. Thanks again.
Joe Saluzzi: Thanks, Chris. Take care.