Print this
Eating Volatility For Breakfast

One day in the life of Mike Riley, Amex options specialist

By Barclay T. Leib

8:45 A.M.
Mike Riley, senior New York partner for Letco Specialists LP, is sitting in his Trinity Place offices chatting about the day to come. "Don't let anyone tell you that this job doesn't come with a great deal of stress,” says Riley, a big burly guy with the freckled face of a redhead. "When I get down to the floor, I am totally focused. But if I lose my focus, I can lose a lot of money. This business has also not been without a physical toll.”

Although not quite in his 50s, Riley has suffered two heart attacks, one of which also involved a mild stroke, and has been carried off the trading floor on a stretcher on two occasions. Put a stethoscope to his chest and you'll hear the steady tick of a pacemaker.

Letco makes markets in 105 stocks assigned to it by the American Stock Exchange. Under AMEX rules, Riley is obligated to offer a price in every option in all market environments, making sure that customer orders are filled at prices competitive with those available on other exchanges. Somehow, in the midst of these efforts, he must make money for himself, while ensuring that the market-makers who compete with him have access to the same customer orders.

Riley is by nature a skeptical bear making money off of a bull market. Today, he will be directly watching seven stocks, and will actively trade in four high-flying tech stocks with annual volatility in excess of 50 percent—Applied Materials, Nextel, Xilinx and Foundry Networks. "The way these stocks trade,” he says, "for all I know, they're all worth nothing.” But Riley is less concerned with the absolute direction of prices than he is with the pace of their movement.

Earlier this morning (it's a balmy late May day), Riley spent 20 minutes in a meeting with his colleagues, where they labeled many of the firm's stocks "stable” or "unstable.” Letco is holding positions on several thousand option contracts in total—most of which were taken on after a customer dealt on a price the firm was obligated to show. Riley and his colleagues are always discussing how to structure the firm's options book—what option positions they'd like to have—knowing full well that there's often an unbridgeable chasm between what they want and what the market may actually give them.

9:15 A.M.
We march through the Members Only entrance to the American Stock Exchange. Stocks have been plunging for two months straight, but in yesterday's session the Dow jumped 240 points and the Nasdaq 100 was up over 8 percent. Will this turn into a significant reversal? Riley has no particular view, but says today could be an important test of the market's ongoing ability to rally. Riley stands at Post 21 in a corner of the AMEX Red Room—a cave-like basement adjacent to the exchange's main building. The low stucco ceilings are crowded with computer screens, and the air-conditioning system blasts away incessantly in the background.

He slides behind a long countertop. There is no stool or bench. A series of computer screens flash information upward from under a panel of Plexiglas, and a row of flat-paneled screens labeled XTOPS and AODB sit on top. Across the counter, a line of market-makers are standing in front of him in a similar row, each with a portable electronic trading slate. One fellow is playing Solitaire on his machine, another is intently checking rows of numbers. Tucked behind Riley is a Bloomberg terminal that his firm uses to keep abreast of news, relative trading volume in options and an intraday chart that he occasionally glances at.

"So where are we?” Riley barks to his assistant, Sean.

"Nextel's still up $7,” replies Sean. "Looks like Xilinx is coming in firmer. AMAT is soft at $85. Foundry's unchanged.”

The volatility of Riley's four active stocks would give the most experienced trader a bad case of the willies. Over the past year, Nextel Communications has ranged from $36 to $160 before descending into the $80s. (It has since split two for one.) The stock moves almost twice as much as the Dow, sporting an implied volatility of around 85 percent. Xilinx's volatility is just marginally lower at approximately 74 percent, the stock having ranged from $22 to $88 a share in the past year. Meanwhile, Applied Materials, with the symbol AMAT, has become one of the hottest of the hot semiconductor stocks in the past year, rising from $30 to $115 before suffering a significant reversal to $85.

Riley will continue massaging his XTOPS system right up to the day's close, jiggling his volatility curves higher and lower as he sees the flow. If he guesses wrong on a volatility level or skew, he could find himself obligated to trade at a disadvantageous price, or potentially arbitraged against a different exchange.

None of these stocks, however, even comes close to Foundry Networks, which has dropped from a high of $208 in March to $75 today, with a front-month volatility of 110 percent. "That one is a real piece of work,” explains Riley. "It flies around on very little volume, often trading on less than a third of the shares the others do.”

This kind of volatility is commonplace in his trading neighborhood. The Letco partner directly to his right is making option markets in Priceline.com. The next one down is covering Yahoo.

Riley quickly checks the flat-paneled XTOPS screen, the heart and soul of his market-making. The XTOPS system allows him to build and assign an appropriate options-pricing volatility curve and enter other variables, such as the cost of money and stock short-selling rebates, in order to price thousands of different strike prices.

Riley will be tapping volatility levels into this machine all day: 95 percent for the front-month Xilinx options, 80 percent for the next series out, 65 percent for the longer-dated options. He will also be adjusting out-of-the-money options higher and lower to reflect the volatility skew he expects. Once his volatility parameters are set, the machine calculates the actual dollar prices of the options in real time as the underlying stock moves.

But Riley's market-making parameters are not set in stone, of course. A little like a pinball junkie, Riley will continue massaging his XTOPS system right up to the day's close, jiggling his volatility curves higher and lower as he sees the flow. If he guesses wrong on a volatility level or skew, he could find himself obligated to trade at a disadvantageous price, or potentially arbitraged against a different exchange. If the system freezes—even for a few seconds—his trading will have to stop.

XTOPS is also much more than just a simple pricing tool. The system helps him manage the second-most-important machine on his desk—the AODB (Automatic Order Data Base), the actual electronic order book for each option series. After XTOPS generates all of the specialist's bid-offered prices and transmits them electronically to the world, it automatically matches these prices against the bids and offers that appear on the AODB terminal.

From the AODB terminal orders stream directly into the XTOPS machine. If a bid and offer are close enough to warrant attention, the stock turns from green to yellow. "Yellow means pay attention,” explains Riley. "Something is close to being tradable.”

When a stock flashes to yellow, Riley will quickly move to a full-screen view of that stock's options, find the specific strike and maturity that is potentially executable (also now color-coded yellow) and tell Sean, his assistant, what he wants to do. Generally, he will also query the trading crowd in front of him as to whether they want to participate in the trade.

It's now the opening at 9:30 A.M., and his first yellow-coded trigger of the day pops up on the screen. One hundred and thirty June AMAT 85 puts are offered at 4 1/8. His theoretical valuation curve within XTOPS tells him that he should buy some of these options at that price.

"Spec buys 50 June 85 AMAT puts at 4 1/8,” he announces. "There's another 80 offered there, do you guys want 'em?”

Out in the crowd, four 20-something market-makers—one each from Timber Hill, Susquehanna and GPZ, plus an independent trader—look up from their terminals. They all raise their hands like schoolchildren with the correct answer to a math question.

"Fine, 80 more trade 4 1/8—20, 20, 20, 20,” Riley barks, as assistant Sean goes into the electronic order book to effect the execution.

A few short keystrokes later, and Sean announces, "AMAT waterfall.” The trades print on a large electronic screen in back of him, allowing the individual market-makers to verify what they each have done. There will be refrains of "waterfall” from the specialist clerks all day as trades are completed. The trades also migrate electronically to the market-makers' hand-held screens and to the specialist's XTOPS machine.

The XTOPS screen also amalgamates small trades of one and two contracts that have been automatically executed by the AMEX's Auto-Ex System. Small pink slips of paper representing these trades are available to be examined just behind Riley, but he will not do so. The trades simply filter into his book without human intervention.

"There's another 80 offered there, do you guys want 'em?” Out in the crowd, four 20-something market makers—one each from Timber Hill, Susquehanna and GPZ, plus an independent trader, look up from their terminals. They all raise their hands like schoolchildren with the correct answer to a math question.

Every option position held by the specialist, whether he knows about it explicitly or not, is then multiplied behind the scenes by its real-time delta and combined into a "net synthetic stock position” that appears just under the name of the stock on the screen. If this total position starts to become too negative or positive, either from new trades or a change in the price of the underlying stock, Riley will look to buy or sell stock to adjust it. His goal is not only to buy options on his bid and sell others on his offer, but to trade the underlying stock adroitly to rebalance his position constantly.

Notably, XTOPS' synthetic calculation does not include Riley's underlying equity positions, which he watches independently. Every few minutes he will ask Sean for "a new stock count” in a given equity. So while XTOPS shows Riley that he is long 86,000 shares of Applied Materials, he is actually close to flat once he factors in his short equity position. Riley does this calculation in his head all day long.

9:31 A.M.
Riley must now pay attention to all of the other option strikes and maturities for AMAT as well as each of the six other stocks that he is directly responsible for. What other option bids or offers coming through from the AODB machine need to be filled?

Within the first minute of trading he will clear all of the other yellow-coded situations. He announces seven or eight further trades in five different stocks, then turns to hover over Sean's shoulder to look at all the incoming new orders.

"Book, book, book, spec, allocate everybody,” he reels off, looking down the list and comparing the new orders to the current market for each option already on the screen.

"Book” means the option is not currently executable and is to join the rest of the orders on the electronic order book. "Spec” means that he is taking the other side of the new order—filling it from his own position. "Allocate everybody” means that each of the market-makers in the crowd is getting part of an executable order, whether or not they know it yet. Riley is like a maestro making the trading flow just as the orchestral band leader helps make the music come forth.

To an outsider, some of his decisions may seem arbitrary, but for every snap decision he makes, he must keep upstairs customers happy by making sure his prices are competitive with those at other exchanges. At the same time, he has to make sure that the crowd of market-makers standing in front are not unhappy. It is a fine balancing act that he has clearly mastered.

In the first few minutes of trading, AMAT has jumped from $85 to $86 1/2. It is not clear whether Riley ever had time to buy a stock hedge against the AMAT puts he bought on the opening.

A Merrill broker in a striped jacket approaches for a price. "AMAT June 95 puts?” he queries. A few toggles of the XTOPS system, and Riley announces, "10 5/8–11.”

"I have 500 to go, 10 5/8.”

"You sell 200 to me, PRO 809M [Riley's floor badge number], you guys want the balance?” Riley queries. The market-makers' hands all go up again. "Fine. 500 trade total, 300 out there, how much each? 50? 80? 120. 40.” The market-makers exchange further details with the Merrill broker, while Riley turns to yet a third terminal to his right—an Instinet DOT machine—to check the underlying stock. AMAT is suddenly falling, now trading down to $84.

"That guy came in just at the wrong time,” says Riley with a bit of glee. "Look at where the stock is now. I'd pay 11 5/8 for those options now.” He is quickly entering buy orders for his delta hedge against both of his first put deals. Once he has finished clicking away at the DOT machine, he turns to Sean and says, "Get me a stock count on AMAT when you have a chance.” Riley wants to confirm his total exposure to AMAT and he can only do that if he knows exactly how much stock he has positioned.

10:15 A.M.
"Holy shit,” says Riley. "Xilinx is flying. AMAT is collapsing. Foundry is up huge. These stocks are all over the place. That's how I know they're all worth nothing. Sean, I need a stock count, Foundries. Does anyone know what that Pru broker is trying to do? I think he may be sniffing around to sell those Foundry calls he bought yesterday. Can we call their booth?”

"Holy shit,” says Riley. "Xilinx is flying. AMAT is collapsing. Foundry is up huge. These stocks are all over the place. That's how I know they're all worth nothing. Sean, I need a stock count, Foundries. Does anyone know what that Pru broker is trying to do? I think he may be sniffing around to sell those Foundry calls he bought yesterday. Can we call their booth?”

Riley is thinking primarily about Foundry for the moment, but must shift gears instantaneously when his assistant spots a large market order coming into Xilinx on the electronic book.

"Mike, I've got 100 Sep 75 Xilinx puts for sale here,” Sean announces, looking down at a market order on the AODB screen.

"I pay $6,” snaps Riley.

The trade prints, and the XTOPS machine immediately flashes back a new quote, 6 3/8–6/5/8. Riley has snagged a nice purchase: the stock is now dropping.

A few hours later, this trade will come back to haunt Riley when a floor official comes by to query his fill. At the time Riley bought these options at $6, the Pacific Coast Exchange was apparently showing a 6 1/8 bid and the upstairs customer wants a fill no worse than 6 1/8th. Riley will protest a bit, but eventually agrees to adjust the fill to the higher level. In his haste, he may have missed checking the bid-offered prices of the other exchanges.

This situation highlights one awkward aspect of multiple-listed option contracts. Although other exchanges' bid and offered prices always show up on the consolidated pricing screens, if Riley actually wants to send an order to another exchange for execution, or trade against that exchange for his own account, he must telephone in the order. There is no direct electronic connection—a small, time-consuming kink in an otherwise automated world.

11:00 A.M.
A broker-dealer comes in and sells Riley 100 December '01 calls on Nextel. "He may know something is going on, and back-dated options are dangerous,” says Riley. "I'm going to push my back-dated volatilities down a bit.” His fingers madly push at various toggle adjustments on the XTOPS system. December '01 Nextel option prices are suddenly all 1/8 to 1/4 lower than they were a few seconds ago.

Relaxing for just a moment between trades, Riley explains that longer-dated options are the most difficult for specialists on the floor to price. The principal risk every specialist worries about is a takeover, and the longer an option's maturity, the greater this risk is magnified.

"Nextel's been a takeover target off and on,” explains Riley. "If it ever gets taken over for cash, all the time-value we might pay for longer-dated options can simply go up in smoke. But because of its large capitalization, it's more likely a takeover candidate for stock.”

What Riley is really pricing, therefore, are the odds of Nextel being taken over in the next two years by another stock that has listed options on it. "If a takeover did take place, we'd likely end up with options on the new stock, but there is a tremendous amount of difference if this new stock is a Vodafone or a Bell South. Vodafone can trade at a volatility in the 80s, but the other is almost always in the 50s,” Riley explains.

So how does Riley make the price on a two-year option on Nextel? He walks me through his thought process: "Front-month volatility on the stock itself is trading around 90 percent, and its long-term mean volatility is probably worth 65 percent. But with maybe a 10 percent chance of a takeover for cash, and a 40 percent chance of a takeover for stock only worth 50 percent volatility, two-year Nextel options that currently trade at 57 percent volatility really aren't that cheap.”

Riley admits there is no definitive answer here. He thinks of his job as more of an art than a science, so he tends to keep his back-dated positions as modest as possible. "If I take too big of a position in longer dates, it's easy to drop a few hundred thousand dollars simply by getting unlucky,” he says. "There is a lot of risk we take all day long that is not particularly quantifiable. You can still get screwed even with good risk management.”

11:30 A.M.
From its 10:15 A.M. low at $84, AMAT has now vaulted up yet again to $87. Riley has been paid on calls on the way up, but strategically over-hedges himself by buying extra stock. He then gets synthetically short when—with the stock trading $86—someone sells him puts that he chooses not to hedge.

When the stock scoots up to $87 1/2, he curses lightly and says he wants the stock back down. "If it keeps going up, my long gamma will take me out of my short position,” he offers, "but it's still a real loss, not just an opportunity cost.” In other words, Riley is long enough underlying options on AMAT that they will eventually act as a stop-loss to his decision not to hedge his newly purchased puts. But it will be the equivalent of hedging at a less advantageous price.

Riley explains that other things being equal, he likes trading from a long gamma, short vega position when possible. The long gamma helps him do an able job market-making with less risk to sudden fluctuations in the underlying stock prices, while the short vega in longer-dated options helps avoid the takeover risk syndrome he worries about.

"Right now, as a shop, we're running with a total negative time decay of about $230,000 a day,” he explains. "AMAT by itself is costing about $15,000 a day, Nextel $10,000, Xilinx $5,000. I've easily covered all that theta already today in our delta hedges, even though I haven't been trading the market particularly well.”

"Are you always positioned this way—long of front-end options?” I ask. Some traders and clerks nearby giggle at my question.

"Not that long ago, I was short 70,000 shares of gamma per point movement in the underlying of AMAT,” says Riley. "I was collecting $600,000 in theta a day at the peak. People on the desk thought I was nuts. But it was right into AMAT's earnings announcement and I was just getting paid in volatility on everything at pretty lofty levels.”

Then Riley recalls that AMAT's earnings came in about as expected, but the stock dropped anyway, from around $83 to $75 in the after-market. "We were selling all the way down—we had to,” says Riley. "But we still ended up pretty synthetically long down at that lower level—even after all of our sales. It was not a night when I went home and slept particularly well. But by the opening the next morning, the stock was back up to unchanged and vols collapsed. We actually ended up doing quite well.”

12:30 P.M.
A noticeable lunch-time lull has arrived together with a few sandwiches and sodas. Per Riley's wish, AMAT has lurched lower from its $87 1/2 high back to $83 1/4, but has now found a midday equilibrium toward $85 1/2. Nextel has dropped from up $7 on the day to down $3 before recovering to unchanged. In the last half-hour, Riley has profitably legged into more delta hedges in both stocks.

But the incoming orders on the AODB machine suddenly drop off to a trickle. The yellow-coded signals on XTOPS stop arriving, and are replaced instead by a steady green.

Nature calls. Riley heads for his single trip of the day to the men's room.

2:15 P.M.
AMAT is back to $87 1/2, and rumors are swinging through the floor that somebody is buying Intel calls like crazy.

He takes a long breath and then announces: "2,000 July 115 calls trade at 2 3/8...500 me PRO, the rest out there...400, 40, 400, 160, 500.” He somehow magically knows what each market-maker is good for. Is this competition? Or is this a group of people following the direction of the specialist acting as their pied piper?

Suddenly, a broker comes in looking at the bid in July 115 AMAT calls for size—2,000 contracts. Two hundred contracts is considered a large-sized order; 2,000 contracts is clearly so big that it's being generated by an institution many times the size of Letco Specialists. To make matters more difficult, the broker announces that his customer is also checking the market for these options "away” at the Chicago Board Options Exchange.

Riley shows a 2 1/4 bid, a few of the other market-makers grunting in assent. The broker offers them at 2 3/8. There is a long pause as Riley contemplates whether this option trade is doable or not. If he doesn't step up to the plate, he may lose the business to Chicago. If he fills the client on the option, but then misses getting a good fill on his stock hedge, he can easily turn this trade into a money-loser for his firm. He quietly clicks out a sell of stock on the Instinet DOT. The market-makers are all clearly going to defer to his lead—his orchestration as to whether this option prints or not.

He takes a long breath and then announces: "2,000 July 115 calls trade at 2 3/8...500 me PRO, the rest out there...400, 40, 400, 160, 500.” He somehow magically knows what each market-maker is good for. Is this competition? Or is this a group of people following the direction of the specialist acting as their pied piper? It is hard to tell, but the trade gets done and all are happy. The business has not been lost to another trading floor.

3:30 P.M.
Riley takes a break to discuss problems Letco has had dealing with the Internet frenzy and the trading patterns he has witnessed.

"Back in 1999 the mere announcement of a stock split would bring in huge call option buy orders,” explains Riley. "It was stupid but meaningful at the same time. Once we had 500 orders piled up on that order book machine all looking to buy. Vol went from 25 to 50 and remained there for a number of days. We had to sell a lot of options we were not immediately able to cover. After that, we started pricing for possible stock splits adding skew to the upside...”

Suddenly Riley is interrupted by his Letco colleague to the left, who trades Texaco and wants some quick advice.

"Hey Mike, Texaco is up $3. Stevie here from Timber Hill wants a price on the July $60 calls. What do you wanna make?”

"4 5/8—7/8” says Riley studying his XTOPS machine.

"I could sell 400 there,” Steve, the Timber Hill market-maker, responds.

"Done,” says Riley as both he and his fellow partner inadvertently duplicate a stock hedge on their respective Instinet DOT machines. There is momentary pause as they figure out what they have done. Then the Timber Hill guy pipes up and says he could sell another 200.

"Done,” the Letco Specialists offer in relief. Their previous over-hedge has now proved fortuitous.

Soon afterward, a headline hits Bloomberg: "NYSE order imbalance to sell Texaco at the close.” That means that for some reason there are market-on-close sell orders building up over on the Big Board. Through its announcement, the NYSE is making people aware of this fact and trying to elicit some interest from potential buyers. The stock immediately drops a point and Riley and his partner sport big smiles. They now have a bit more gamma from their new option purchases to trade against.

The Internet stock circus plays on. AMAT is now getting hit, trading down to $84 1/4, then $83 3/8. In a heartbeat, Foundry networks drops from $87 to $83. Xilinx is falling into the close as well. Riley starts madly buying stock in all three.

3:58 P.M.
I've been standing all day. My legs are tired. I look over at Rob, the lone independent market-maker. The close is just two minutes away, but he doesn't look busy.

"Do you have a particular style of trading?” I ask.

"I used to be a short option seller,” he answers, "But I saw the light. Look at this goddamn stock Texaco,” he points out. "Up $4 a few minutes ago, now only up a buck and change. It took 1.6 million shares to get it up there; and now we're down 2 1/2 points from the high on just 250,000 shares. That specialist should be in jail.”

"Easy, Rob,” says Riley in defense of the Texaco specialist on the NYSE who he likely does not even know. "I think the market has more to do with it than the specialist. Don't badmouth our brethren. We love you too, Rob.”

"You're hiding it well,” Rob retorts.

Has Rob lost money for the day? Or is this just gentle banter between options market-making friends? Again, it's hard to tell. Turning back to address me, Rob chastises himself. "No brains up here,” he says pointing to his cranium. "The only other job I'm qualified for—or ever held—is delivering Chinese food.”

4:01 P.M.
Somewhere in the next room, a bell is ringing: the end of the trading day. Riley freezes his trading system before any after-market stock quotes start to wreak havoc with his carefully designed option-pricing grid. AMAT's range for the day has been $83 1/16 to $87 3/4, mild by current-day standards. Nextel has traded between $89 1/8 and $96 3/4, a tad more volatile but nothing outside the norms for Riley and associates.

Riley and all his teammates exchange a ceremonial fist touch to congratulate their hard work together. Some of Riley's partners head home. Within minutes, the floor empties far faster than would seem humanly possible.

About 90 percent of Riley's business that day has been completed electronically. Riley's job as a specialist dates back to the days of ledgers and paper order books, but there are none of these anymore. Instead, Riley has stood captain over his XTOPS pricing tool and his Instinet DOT terminal, while the AODB machine has quietly interfaced with the former. In time, it is easy to imagine Riley interfacing with a fourth screen connected to the International Securities Exchange, the new on-line options exchange. For the moment, though, AMEX rules allow this only to occur upstairs, off the exchange floor.

Riley is painfully aware that many people think specialists on the floor of an exchange have a "license to steal”—and that his job is viewed as antiquated, noncompetitive and unnecessary. Some even suggest that his function won't exist in a few years, completely replaced by new technology. But on this day at least, Riley appears to have worked quite hard—using modern technology as opposed to resisting it.

"How'd you do today?” I query.

"Traded like a baboon,” says Riley, "But all considering, we did just fine.”

Dirty Tricks in Internet Options

Are options used to manipulate the price of Internet stocks? From his position on the front lines, Mike Riley sees a lot of suspicious trading patterns. Some might be worthy of fueling an SEC investigation—or two.

In one situation, Riley was trading options on Yahoo during a particularly volatile expiration day in late 1999. In the morning, Yahoo had quickly moved up $7 dollars to trade around $183. By 3:30 P.M., the stock had reached $186. But with 15 minutes to go to the close, the stock was still well under $190—a level where a strike price existed with considerable open interest.

"Then, with just a few minutes to trade, a big retail customer came in looking for the 190 calls,” says Riley. Examining the AODB machine, Riley saw that Chicago was offering these options at just 7/8th while he was actually bidding $1, an inverted market—but who knew what this option was worth, if anything, with just a few minutes to trade? Riley called Chicago to try to lift their offer, and, just to be competitive with the rival exchange, agreed to sell the customer 200 of the options at 7/8th—a trade he was not exactly thrilled about but felt obligated to transact given his specialist duties.

"We finally got Chicago up to $1 bid,” Riley retells, "and then the customer plows in and buys 2,000 up to $2, then another 1,000 up to $2 1/2, and then the last 1,000 on both exchanges at $5.” Everyone was scrambling to buy stock at the same time, and the stock shot up to 190 1/2 by the close.

Of course, Riley could have made money on the trade—if he had been able to buy a sufficient number of Yahoo shares at some reasonable price. But, he explains, "We couldn't buy any stock—hardly a single share. The stock traded up to $194 1/2 in the after-market on no size and opened $199 on Monday, before vaulting up to $206. We got run over, as I'm sure the CBOE did too. That two-minute bet ended up costing us over $1 million.”

Was the retail customer trying to manipulate the stock using options? In that specific instance, Riley isn't sure what that trade represented. But it did make for a pretty wild expiration day with some fairly odd trading behavior.

Another questionable incident involved trading in Priceline options. Exchanges cannot begin trading options until 90 days after a stock's initial public offering. In the spring of 1999, the AMEX was preparing to start options trading on Priceline on the 91st day, and had assigned Letco to manage it.

"I should have seen it coming quicker,” Riley looks back. "We had requests for Flex customized options before we even started trading the standardized options. That should have been the tip-off that insiders were selling. The stock fell $15 before I realized what was going on, but it's clear now that insiders were huge buyers of puts and sellers of calls.”

Riley does not know what Priceline.com insiders were legally allowed to do, given lock-up covenants, but he is adamant that "they somehow found an outlet to hedge their stock holdings using options.”

"Within a month's time, I'd estimate that 10 million shares were synthetically sold using options,” he says. "Of course the skew to puts was incredible, and it was next to impossible to borrow the stock. The whole thing continued that way for the longest time until the secondary stock offering at $67. Then the stock became easier to borrow and the vols normalized somewhat.”

Riley has even stronger questions about potential insider trading he witnessed in Summit Technology shortly before a May takeover announcement. "Two days before that takeover hit the news services, one broker came plowing in wanting to do a most unusual spread trade—selling the June 12 puts and buying the in-the-money June 10 calls. The entire package had something like a 160 percent delta. Then, boom, the takeover is announced two days later. Now tell me that customer didn't know something.”

Letco reported the odd trading to the authorities, but Riley doubts if the SEC will do anything about it. "It's just another complaint from a group of people—the specialists—whose existence the government is not sure they even want to acknowledge. No one really cares when we get picked off.”