The Dirty Little Secrets of Electronic Options Trading
Derivatives Strategy
When it comes to options trading, many electronic platforms are little more than overblown credit facilities, with voice brokers crossing upstairs trades on the screen.
By Barclay T. Leib
“You could tell something traded just by the price action and the broker scuttlebutt, but the only people who knew exactly what went through at what price are the original customer, one broker and one market-maker.”
— John Maher, ED&F Man
“Electronic trading has become an over-the-counter market connected by an electronic system, as opposed to a true futures exchange where things are brought to the floor.”
— Stan Jonas, Fimat USA
According to the latest prognosticators, human vocal chords—and the brokers attached to them—are doomed as instruments of trading. At least 88 different electronic trading systems now exist to trade global bonds, and the plethora of equity portals is even greater. The ultimate idea behind these electronic platforms is to execute trades without human interaction, the same way factories make frozen cupcakes untouched by the human hand. People will serve the computers by diligently restocking the paper tray and sweeping up the floor at night.
But is electronic trading really that easy when applied to complex derivatives? Do electronic trading systems work for all products? And when markets are said to go electronic, do they really function as well without voice intervention?
The answers to these questions are crucial to the Chicago Mercantile Exchange and Chicago Board of Trade as they advance their new electronic trading efforts. The CBOT’s electronic a/c/e initiative, for instance, already represents about 12 percent of the exchange’s daily Treasury bond futures volume, and the first few options on futures have already been posted electronically. It’s important for these exchanges to set rules—and to do so fairly—now.
Eurex and LiffeConnect are the exchanges with the longest history of putting modern technology to work in the service of the markets—and exchanges on this side of the pond are set to mimic some of their ways. But according to traders, these ostensibly electronic platforms, especially when it comes to options trading, are sometimes little more than overblown credit facilities with voice brokers still doing most of the work behind the scenes. Institutional players like the platforms because their rules are flexible, but they note that prearranging huge volumes of business off the floor and merely crossing it on-screen can reduce transparency. Deals still print in terms of volume but sometimes do so without an associated price. As a result, small players may not be dealing on an even playing field. Conflict-ridden pay-for-flow practices also abound between market-makers and brokers.
These dirty little secrets of electronic trading may not surprise daily market participants, and they are not heinous in nature. But together they lay out some issues for U.S. exchanges to be wary of. They also suggest that the role of the traditional voice broker may be far from dead.
Complexity issues
When it comes to trading something generic—like U.S. dollar-euro, December Treasury bonds or a given stock-–few would deny that electronic trading is a natural choice, especially if order flow can be aggregated into a unified platform or portal while maintaining market-making competition.
But what happens with something like a 22-day option struck 2.36 percent out-of-the-money to sell the euro-yen cross? Or when a participant wants to trade an option spread or purchase an option straddle?
Electronic platforms can, of course, be programmed to handle such nitty-gritty details. (See “Eating Internet Volatility for Breakfast,” July 2000.) Sophisticated electronic market-making tools now make it possible for market-makers to define a volatility matrix complete with odd idiosyncratic skews, and to quote a plethora of options at one time. By using such systems, market-makers can instantaneously highlight orders on an electronic communications network that may be executable or off-market, and pounce like electronic pythons finding prey amidst a jungle of thick grass.
This works reasonably well in markets such as U.S. equity options, where volatility bid-offered spreads are quite wide and the volumes traded are mostly retail. On the Eurex and LiffeConnect platforms, however, the trade flow is mostly institutional, the average trade size is larger, and the volatility bid-offered spreads are generally tighter. In such an environment, one London broker explains, electronic trading has a different effect.
“No market-maker is going to leave tight option quotes up on the screens,” says Alex McLeod, an active Eurex and LiffeConnect options participant at Electronic Brokers in London. “Prices posted on these markets are relatively wide and for small size. It’s an indicative market. No one wants to get picked off on a whole bunch of options at the same time. The real business is still going through the voice brokers.”
And therein lies the rub. According to Eurex market-makers, investment banks, brokers and hedge funds, approximately 80 percent of the options business done on Eurex, and to a slightly lesser extent on LiffeConnect, is actually pre-negotiated by human beings over the telephone. Once agreed upon, the trades are then routinely crossed electronically as a simple clearing mechanism.
“This way of doing business has been around ever since Eurex got started,” says Stan Jonas, managing director for Fimat USA in New York. “It’s become an over-the-counter market connected by an electronic system, as opposed to a true futures exchange where things are brought to the floor.”
Better than before
It’s important to note that this methodology appeals to most market participants and is now well-accepted. Eurex volumes in recent years have flourished, while the traditional Liffe physical trading floor fizzled. Eurex’s system of executing options business was in large part a direct response to a Liffe floor that few admired. “Liffe was a corrupt and venal place,” remembers Jonas. “Everyone was happy to see that floor disappear since it was so difficult getting business done down there.”
Matt Joyce, an options market-maker in London for Stafford Trading and a former Liffe member, agrees. “There was a general attitude on the Liffe to try and read the paper flow,” he says. “The locals would see someone coming in for a call price in a sharply rising market, guess that the interest was a buyer, and then everyone would kind of grimace and make a really high spoof price. If they had read the guy wrong, and he tried to hit their bid, they’d back off that bid in a flash. It was not uncommon to see options that were initially quoted 30–34 suddenly print outside that initial quote at 26. It was really bad.”
Crossing deals on the old Liffe floor between two agreeable upstairs counterparts was also often difficult. “Liffe required you to show your interest on the floor at every stage of the dealing process,” says McLeod of Electronic Brokers. “Sometimes the floor traders would not be very helpful developing a reasonable market in the first place, but they’d be entitled to jump in and steal one side of a deal once they could determine what price was about to trade between two clients. The rules pandered to the locals on the floor, and could leave an upstairs client who had stepped up to the plate with the best price shut out of a deal altogether.”
From the get-go, Eurex’s rules were much more generous on cross trades, allowing crosses to transpire even without publicly exposing the business on the electronic platform. This difference, along with the advent of a more unified Europe and euro currency, caused Liffe volume to dwindle. Between 1997 and 1999, Liffe lost its dominance in German Bund futures and options on the futures. Meanwhile, the Eurex boomed. Eurex options on futures volume ballooned from an annual 2.4 million contracts in 1997 to 29.1 million by the end of 1999.
All is fair in love and war, right? If Liffe was corrupt and venal before finally being forced to go electronic itself, it certainly deserves to have lost significant market share. But electronic trading on Eurex appears to have developed at a price.
The private trade
The first critique of Eurex procedures heard from market participants is that its very flexibility is also its Achilles’ heel. Because deals can be pre-negotiated outside of the electronic platform, bigger players can do their deals privately, without letting the little guys in on the action.
To understand this problem, it’s necessary to understand exactly how a typical Eurex options trade of any size and complexity (spread, straddle and so on) takes place in today’s market. Eurex itself has little functionality to handle complicated orders like a calendar spread on the actual electronic terminal, and while LiffeConnect does handle such orders, this functionality is thinly used.
Instead, a broker such as Fimat’s Jonas will receive an order or an interest for a potential order from a client. He then calls an upstairs broker in London that specializes in Eurex options to develop an inside market for his particular interest. That broker in turn calls a small group of upstairs market-makers dominated by two firms—Helios Group and Stafford Trading—to piece together a price. After a bit of hunting, the broker informs Jonas of the best market, and if the price is agreeable to the ultimate client, the broker will cross the trade on Eurex as a block trade for Fimat’s and the market-maker’s respective accounts.
No one working an order on the actual Eurex electronic system gets to see this interest except the upstairs market-makers Jonas’s broker called. Moreover, if the block trade is greater than 1,000 contracts, it typically does not even appear on the Eurex system with a price. Only the volume of the trade is disclosed, and the trade may not actually show up on the Eurex system for hours. This is not a transparent and even-handed market for all players.
“There was reputedly a trade that went through for 12,000 contracts this morning,” explains John Maher, a vice president at ED&F Man in London, referring to the Eurex Bund contract. “You could tell something traded just by the price action and the broker scuttlebutt, but the only people who know exactly what went through at what price are the original customer, one broker, and one market-maker. It’s been five hours now and nothing has printed, and when it does finally print, I can guarantee you that there won’t be a price associated with it.”
So for a given option where the inside market for small size might have been 20–22, that 12,000 lot could have traded up at 25 or down at 18—or anywhere else for that matter—without average market participants knowing.
“That’s probably not fair,” says Maher. Bond spread trader Tom Cooper of London-based Buttonwood Group expands: “It affords the market-maker an extra edge and extra time to scoop back his position at better prices.”
Indeed, Eurex recently reduced its minimum block size from 250 lots to just 50 lots, and while most small crosses still print with a price, there is no requirement that they must. It is therefore theoretically possible on Eurex that market transparency could completely disappear.
Empty screens
A second and related problem is that with so much electronic trading taking place via voice brokers, overall options volumes may be high, but intraday bid-offered prices shown on data vendors such as Bloomberg or Reuters aren’t great. Near-dated options almost always have a reasonable bid and an offer, but move a few maturities out and prices can simply disappear.
“I’m never really sure what I’m looking at on these screens,” says a London-based fixed-income broker. “Sometimes it could be last night’s close, other times it might be the last trade that took place two hours ago, or maybe it’s the current bid-offered market. I often just can’t tell. These screens are not well-populated, particularly in the longer-dated options.”
If this broker is confused, think of the poor hedge fund manager or retail investor trying to calculate where a given option spread would be tradable. “Particularly with spread interests, you can’t necessarily figure out where the price should be from the screens,” says Ron Mandell, a market-maker for London-based Goldenberg, Heymeyer & Co. “The retail investor who wants to trade 10 lots of a spread interest is basically disadvantaged,” says another London fixed-income broker. “There’s no facility to work a spread interest electronically on Eurex, and for small size most upstairs brokers would hardly waste the time canvassing market-makers for a good price.”
Paying for flow
A third problem with electronic trading as it now exists involves the way brokers are paid. In order to attract order flow, it has become common practice for market-makers on Eurex and LiffeConnect to pay fees to brokers for bringing option orders their way. Given that some market-makers may pay more than others, brokers face a natural conflict of interest as to whom they call, and there is no guarantee customers will always get the very best price available.
“If a tic is worth 25 euros, a market-maker has an incentive to pay up to 24 euros per trade to get an attractive piece of business with an extra tic in it,” says Buttonwood’s Cooper. Typical rebate arrangements tend to run well south of that, however, with standard rates of one-half to one euro per contract.
“I wasn’t here at the beginning of this market,” says broker Maher, “but I understand that pay-for-flow was an immediate phenomenon, and it is certainly now a standard in the industry. If a market-maker doesn’t pay, he won’t see any business.”
The big question: How much extra padding do market-makers add to their quotes to compensate for this added cost burden? If everyone adds the same amount, perhaps this is not a major issue, but some market participants think otherwise.
“It’s a very, very bad system that was around even before electronic trading,” says Chicago-based market-maker Donald Wilson of DRW Holdings. “I was in London between 1994 and 1996 as the Eurex was just getting going. At the time, I didn’t mind paying brokerage to attract business. That was until I figured out one day that I had made a certain amount of money taking real risk, but had paid out a third of what I’d made in broker kickbacks. That was ridiculous.” So as the market transited more and more to its electronic platform, but with the same type of broker-rebate arrangement, Wilson decided to stop participating. Instead, he returned to his home base in Chicago, where he now hopes the new Merc and CBOT electronic platforms will not become so tainted.
More recently, London market-maker Helios, one of the most aggressive pay-for-flow shops in the early years, tried to reduce the brokerage kickback levels it was paying. What happened? All the brokers stopped calling Helios, and Helios was forced to return to its old ways.
Some market-makers have also attempted to cut brokers out of the equation altogether, by encouraging customers to call them directly. “If 80 percent of Eurex business is pre-negotiated, approximately 80 percent of that goes through the brokers,” says Maher of ED&F Man. “The other 20 percent gets stolen off directly by the market-makers.” Another London dealer adds, “Not too many brokers are trying to move into the market-making business, but there are some market-makers increasingly trying to be brokers.” Create too much broker wrath, however, and, this fellow explains, “You can get completely cut out of the game rather quickly.”
Cuffed prices
Another transparency and market fragmentation issue revolves around the order-routing procedures at large investment banks.
While investment banks could theoretically route customers’ orders to brokers and independent market-makers, many large option orders are actually filled directly off of the banks’ internal market-making desks without always checking prices away. According to a range of Eurex participants, Merrill Lynch and Goldman Sachs fill all large customer Eurex options off of their own respective market-making books. These firms guarantee “best execution” to their customers, but it is very difficult to prove they truly deliver it.
“In the pit, you at least see when someone does something dirty,” says Maher. “But in this market, you only see what’s on the screen or what someone tells you about.” In other words, there’s really no discerning how much big firms like Goldman and Merrill could be skimming off their clients’ orders, and there are few checks and balances (other than normal market competition) to prevent them from doing this.
The power of speed
Lastly, according to some, in those instances where pure electronic trading actually takes place on Eurex or LiffeConnect (as opposed to the pre-negotiation process), trading success is increasingly driven more by systems speed than anything else. Several market-makers have autoquote engines hooked up to the exchange feeds to automatically initiate trades when a given bid or offer is deemed out of line. As a result, many less sophisticated participants have been scared away from providing market liquidity.
“These market-makers spend enormous amounts of money overlaying their bid-offered parameters—not necessarily showing prices, but capable of trading, reacting to any prices there,” says a London fixed-income strategist. “Anything that is off-market will trade before you even see it on the screen.”
Apparently, there are only three or so firms in this speed ball game, but the way their engines are set up, there may not be much incentive for others to show two-way prices. Show the wrong firm price or cancel a given bid or offer too slowly, and you’ll likely be picked off by an autoquote engine. Show the right price, and the autoquote engine may already be in front of you.
The power of machines sometimes scares away mere mortals.
Eyes on Chicago
In spite of these problems, Eurex still functions far better than some other electronic option platforms. Comparatively, brokers show even greater disdain for the Italian IDEM equity options market. But questions remain: Is there something just complicated enough about options trading to have resulted in this odd half-voice, half-electronic way of doing business?
And what monster of unfairness and suspect transparency has now been created? Even more importantly, perhaps: Are the Board of Trade and Merc about to embark in a similar direction?
“Try to move options to an electronic platform, and then impose too many rules about what you can and cannot do, and you’ll stultify the market,” predicts one Chicago executing broker. “Don’t have enough rules, and it won’t be long until I’m just calling my own upstairs govvie desk for prices and lobbing up the fills as crosses.”
From all appearances, the CBOT currently plans to be a bit stricter than the free-for-all on Eurex, and to follow the slightly more rules-heavy model of LiffeConnect. Recently, the CBOT requested a no-action letter from the Commodity Futures Trading Commission regarding the ability of upstairs brokers to engage in pre-negotiation. After receiving a favorable response, the exchange issued a notice on October 26 allowing for pre-execution discussions on any orders entered on the a/c/e platform. All that’s required under this notice is that when attempting to cross two orders, one must wait five seconds before entering an offsetting futures order and 15 seconds before entering the second side of a pre-negotiated option order. This is relatively similar to the current rules on LiffeConnect.
Mayer of ED&F Man, like many other Eurex brokers and market makers, originally came from Chicago. He claims to have recently executed the first CBOT a/c/e option trade following these guidelines, and is gearing up for a whole new upstairs brokerage business back in his hometown. Several other market-making firms are preparing to do the same.
But as Eurex market-makers slowly return to Chicago, how will the CFTC react if these firms bring with them the rebate system so commonplace for directing order flow on the European option exchanges? What if upstairs investment banks start making their own prices, show the unattractive side of a pre-negotiated cross deal on an electronic screen for 15 seconds, and then lift their own offer with the customer order when the clock allows? Is this going to be electronic efficiency, or could the traditional pits of Chicago quickly deteriorate into an upstairs market suitable for the big guys, but less fair and transparent for the little ones?
That risk is certainly there, but it will likely to be outweighed by other cost considerations. The inexorable push toward electronic trading (even if brokers are still madly piecing together deals behind the scenes) will likely continue regardless of the market consequences. “The cost savings to the large institutions not to have to support all those locals driving around Chicago in Mercedes is just too high,” says Fimat’s Jonas.
“The CBOT floor will shut down and the Merc will follow,” predicts Mandell of Goldenberg, Heymeyer. “While the two market systems could coexist for awhile, you can’t afford to have both a floor and an electronic market longer term, and the floor being more expensive will go.”
Jonas thinks the result will be great for his business, but a bit bizarre and worrisome overall. “It will be a return to the past,” he says. “Everything the exchanges did to bring liquidity to one place, we’re now going to have to replicate. If Fimat is doing business for a customer, we’re going to have to call up to eight different upstairs market-makers to determine where the market really is in certain products. It could look and feel something like the old OTC days. Nobody will know where prices are.”
Button fear
Even if Jonas’s worst fears don’t come true, trading options electronically may suffer from another foible: simple trader reticence. Even though some platforms such as LiffeConnect can handle complicated option orders such as ratio spreads, call butterflies, straddles, strangles and collars, many market participants don’t actually use this functionality.
“Customer desks are busy, increasingly trading a variety of products in a variety of markets,” says Joyce of Stafford. “LiffeConnect is a good platform, but even trading a one-by-two can be confusing. When the price comes back at
‘-2’ on a given spread, does that mean I pay that or receive that? People still need someone to hold their hand, and don’t have time to focus on all these different terminals with all their different rules.”
“Some end-users—even among the big swap and hedge fund traders—just don’t want to press the button,” says Wilson of DRW Holdings. “They know that if they somehow screw up and press the wrong key, it could cost them millions of dollars. They could get themselves fired. If they call their broker, who they have been dealing with for 10 years, it’s a better risk-reward situation.”
Overall, electronic trading is likely to flourish in fungible and easily tradable futures contracts. But the resilience and adaptability of different electronic platforms will face a tough fight on the options front. The electronic Eurex may crow about their burgeoning options volumes, but don’t be fooled. In many ways, the Eurex options market is now more a voice market than an electronic one—and there are numerous technical and subconscious reasons for the way it has evolved.
And, Chicago, take note. It may not be long until Treasury bond futures will be successfully flying around electronically on the a/c/e platform, but options trading may take longer and stand at more risk of being harmed. “I never thought I’d say this,” concludes DRW’s Wilson, “but compared with the direction we seem to be headed, I actually think the floor could end up looking like a pretty fair and open place.” God forbid.
Voice Brokers Behind the Black Boxes
The failure of electronic platforms to succeed without voice intervention in options trading is not limited to the Eurex. Anecdotal evidence seems to be popping up elsewhere in the derivatives world that a voice broker’s touch may be something of a necessity.
Despite last year’s hoopla about a new electronic brokerage model without high-cost brokers, electronic foreign-currency options consortium Volbroker.com recently hired seven former brokers and is now looking to add more. “Our platform started with great liquidity,” says CEO Dirk Ward, “but there is still a great deal to be desired in terms of actual transactions. We found, after a short time, that the bigger the trade and the more unusual the business, the more a voice broker has to be involved in the transaction process to actually get it done. This was something that we absolutely underestimated before.”
According to industry sources, Volbroker.com has also been entertaining lavishly to help kick-start support for its platform. Maybe this is understandable pump-priming to reach critical mass, but perhaps not. In the electricity market, the voice-assisted AllTrade platform (an alliance between Amerex, Ultra and Prebon) has recently been stealing market share from Bloomberg’s previously more powerful all-electronic system. “By being able to draw peoples’ attention to certain markets and provide them with true color, people tend to trade more actively,” says Robert Fischetti, vice president of Prebon Energy.
In the credit derivatives space, meanwhile, one desk head at a major money-center bank acknowledges that although he is extremely comfortable with new technology, he still likes to use voice brokers. “The market is a bid and an offer, and depending upon the product, there can be a lot of space in between,” he says. “The voice broker fills that gap, moving information around and stroking relative egos on both sides of the trade to get something to print. Without that function, the prices might just stay there for a long, long time.”
This trader has a creditex terminal on his desk, but uses it sparingly. “When trading, I can’t be bothered to even dial the phone, let alone figure out how to put an order into an electronic platform,” he admits.
John McEvoy, president and founder of creditex, confirms that the company now also uses “trade facilitators” behind the screens to help piece together electronic bids and offers that are close but haven’t quite resulted in a deal. In addition, the large majority of traders using CreditTrade, creditex’s largest electronic competitor, still call in orders by phone. The brokers then key in the orders electronically on the customer’s behalf—a manner of trading somewhat akin to the old Cantor green-screens.
Even Blackbird, the stealth derivatives trading platform, has what it calls “activity monitors” in each of its office locations. “Anyone who thinks you can put a bunch of machines in a room and create an all-powerful Borg from StarTrek just doesn’t appreciate how business really gets done,” says Sean Dorsch, the platform’s president and CEO. “Of course you need some people highlighting attractive bids and offers, and exposing your platform to market participants who might not even have a Blackbird terminal on their desk yet.”
Mike Gooch, chairman and CEO of broker GFI Group, best sums up the need for a voice-assisted platform with two thoughts.
The first relates to finesse. “A lot of customers want to put a generic order into a machine, but want to let the broker know, ‘Even though I’m just showing 100, I really have 500 lots to do,’” he explains. “They say not to tell anybody, but to just keep this in our back pocket. Sometimes the client indicates he has a little room to improve the bid or offer. The customer trusts the broker, because if the broker screws it up, everyone knows he won’t get that order again.”
Gooch’s second thought relates to the plethora of potential trades a dealer must watch. “For some of the more complicated and esoteric products,” he says, “the guy trading may be looking at any one of 200 different possible trades at any given moment. Without a voice broker, how are you going to get his attention that you have this trade popping out of the curve?”
Make no mistake: Within the options world and in the minds of many, the voice broker is far from dead.
—B.L. |
