Monday, May 22, 2006

Cuban's Collar -- Anatomy of a Famous Trade

Mark Cuban's collar trade is one of the most famous trade using derivatives during the internet bubble time. Here I will analyze in detail how options make this ingenious trade possible and how both sides of the trade work.

Costless Collar

Back in 1995, Mark Cuban and partner Todd Wagner created Broadcast.com, a multimedia behemoth providing streaming audio and video of live radio, TV and sports. They sold the company to Yahoo! for $5.7 billion (yes, "B"). Yahoo's market cap at the time was about 130 billion -- Today it is 44 Billion, in between Yahoo's market cap reached almost 300 billion and crashed down to about 10. The ingenious option trade Mark Cuban made allowed him to protect his billions, without paying any insurance premium. At the time he had 14.6 million shares of Yahoo traded at $95. Thus he had a market value of $1.4 billion (I'm not sure about the rest of the total $5.7 billion, but that's not the concern of this article). Probably because there was a lock-in period for him to sell the Yahoo share, or he used these Yahoo shares as collateral for a bank loan, or he didn't want to miss the opportunity if Yahoo continued to rally, he chose to enter a collar to lock in his share value without selling the shares. Here is the detail of the option trade:

1) For each 100 shares of Yahoo stock, 1 contract of put (strike 85) was bought and 1 contract of call (strike 205) was sold. In total there were 146,000 contracts of calls and 146,000 contracts of puts traded.

2) The premium of the put exactly offset the premium of the call, thus there was zero cost for this trade (not sure about the whether Cuban had to pay the commission but it is likely the commission cost was rolled into the premium as part of the total deal between Cuban and the counterparty bank).

3) All options expired in 3 years.

Such option structure is called collar. A collar consists of two legs: buying a downside put and selling an upside call. A costless collar has the premium of the put offsetting the call's exactly so that there is no cost to enter the collar trade.

After Cuban's collar trade was entered, Yahoo's share price reached the stratosphere of $237 in January 2000. Such a collar trade seemed not so smart. Then the internet bubble burst, and the Yahoo reached the abysmal of $13 in late 2002, the collar turned out to be a stroke of genius, a great risk management trade that cost nothing to have. Cuban was able to pin down on the value of the share (or 90% of it), no matter how the subsequent movement of the stocks.

Several interesting observations of this trade. Even though the Call price equal to the Put price, the Call was $110 out-of-money, while the Put was only $10 out-of-money, that seemed to be a huge disparity. Was that the Call way too much underprice or the Put was well too much overpriced?

Actually the pricing is not so outrageous as it appears. There are two factors in working here: 1) Skew; 2) Interests. The first factor, skew, means the market always prices downside protection (puts) more expensive than upside (calls), thus puts are usually more expensive than the calls in the same out-of-money amount. Such market phenomenon is called skew, which can be measured in several ways (one way is to describe the skew by the pricing of a collar).

Another factor is more important in this case. Calls can save the holder interests than holding stock (since you don't need to shell out cash to buy stock until exercise time), while puts cost you interests that you should get if you short stock instead of buying put. Thus the combination of the two (sell call and buy put) - both costs interests -- means collar is costly in interests. Supposed Cuban didn't do the collar trade but sold the stock instead. The proceeds from the stock sales could earn him interests for the subsequent 3 years, no matter whether the market was up and down. While with the collar he would lock in the stock value (just as cashing out), but without the interests. Thus collar would be unfavorable to holders, unless such interests consideration is priced in. Thus 3 years of interests would discount call price and increase put price: the same effect as skew.

Did He Miss Something?

Even though in retrospect this collar trade was smart in protecting Cuban's original share value, he might have missed out on two things: 1) he might have overpaid for skew which typically is overpriced in the market place due to psychological factors of investors; 2) He missed out on the interests he deserved. If he really wanted to lock in the share price rather than speculating on the upside potential of the stock price, he could do 85/95 collar (buy 85 put, sell 95 call) which should give him quite substantial amount of premium income (since 95 call should have much higher premium than 85 put). Such income should be roughly equal to the interest income he should received if he liquidated his Yahoo shares into cash. The banks that marketed such costless collar to big investors like Cuban would like them to forget about the interest aspect and focus on the "costless" feature of collar. To construct such costless collar, it ends up with a collar with two far apart strikes (such as 85 put and 205 call), containing a huge speculative upside potential (which is attractive to the speculative mind). Basically it is a way of encourage the investors to forgo their interest income to exchange for some speculative opportunity. In Cuban's case, the banker might present the "benefit" that even when Yahoo should rise to 200, his share would not be called away so that he would still participate in the rally.

In the end, that collar worked more like a put (while the upside call financed the cost of the put), and the "only" thing Cuban lost was potential interests he would have received if he just sold the stock. How much? The interests for 3 years would amount to $200 million or more.

The Counterparties of the Trade

Cuban's collar helped him protect his billion. So did the party on the opposite side of the trade lose billion when Yahoo tanked? Let's use this trade to look at how the option market works for both sides of the players.

On one side, Cuban now owned puts and short calls, with his existing holding on Yahoo stock. His counterparty of this trade, the market maker or the bank, would short puts and long calls. So what would happen to these huge positions on the opposite end of the trade?

Obviously if the market maker did not do anything and just hold the short puts and long calls, the short puts would wipe a billion or more off him when Yahoo tanked. However, contrary to common belief that options are zero-sum game, the market makers do not gamble against customers, namely the customers and the MM may not typically on the opposite ends of a bet on the speculative part of a trade (see my earlier article on this). Instead, MM would hedge. In this case, to hedge the options, they should short almost the same amount of Yahoo stocks as Cuban was holding. Thus it almost looks like the counterparty sold (as short) all the shares Cuban would like to sell but couldn't (either because of regulations, or taxes or loan requirements, etc.). Thus when Yahoo stock dropped to $13, Cuban was safe, so was the counterparty.

Actually there is more to that.

In real life though, since the trade is so huge (292,000 contracts of calls and puts) which entailed many aspects of risks no matter how you hedge it with stock, the counterparty would typically lay off the risks to many other parties, which included 1) other market makers; 2) proprietary traders, who might take on part of the trade for reasons other than speculating on direction (for example, the short put leg might seems very juicy in volatility that some volatility traders might want to assume portion of the risk of the short put trade); 3) Speculators/investors (for example, the long 205 calls may be transferred to the hands of investors who were extremely bullish about Yahoo stock).

So how could the counterparties of this trade make profit (or loss), since we all know Cuban had lock in his money even when Yahoo share price plunged to $13 in 2002? The profits of the counterparties came from

1) Bid/Ask spread.

2) Interests.

3) Volatility.

4) Speculation.

The first one is obvious. The market makers provide a service for liquidity (they have the obligation to take the other sides no matter whether the investors want to buy or sell), so they deserve to profit from the bid/offer spread. The second one, interests, is less obvious. Suppose there was a single party holding exactly the opposite positions as Mark Cuban's: long 146K contracts of calls, short 146K contracts of puts, and short 14.6 million shares of Yahoo stock, how much he would make by the time when the option expired? He would make roughly $200 million in interests, while Cuban got to protect his Yahoo stock value. That was a pretty cool trade for both parties -- a win-win.

The interests show the glacier power of money. Many investors look at the few percent of interests with distain and focus their eyes on the quick money that might come from the fast movement of stocks. Actually I have not heard anyone commenting on the potential "loss" of $200 million for this famous trade, and everyone extols on the virtues of protective power of the costless collar.

The third point, volatility is more technical and less certain, and just like the fourth point, both are speculative and the profits and losses depend on the movement of stocks. It is difficult to estimate the total gains or losses of these two types, but it is unlikely they would lose billion.

So supposed the counterparties of this trade had only one type: the market makers who hedged. What would be the total value of this trade when we add the two sides' profit and loss together? Was it zero? Not at all. The net should be at least $200 million, not counting the instant profits for MM on the bid/ask spread, the commission, and the overpricedness of downside volatility, all these were in the profit side of the equation for the counter parties (while they did not add to the loss side of Cuban). So magically, a collar trade could create net positive value in substantial amount.

That's the way how the option market has the potential to create net profits and distribute risks: On one side, Cuban had a particular interest to create this huge collar trade for protection (with a tack of speculation on upside); on the opposite end of the stick, this huge trade would be broken down into many smaller pieces and the risks would be transferred to many parties with many different goals, each one had their own way of making or losing money. But in the end, the net sum is not zero, but likely a positive number. It shows the power of risk transfer function of options in a very complex space, which is not a zero-sum game as many people believe.

(The author can be contacted at huangxinw@gmail.com.)