Whenever I find myself at a roulette wheel I typically play the outside, trying to hedge myself by betting on two of the 50% (ok, it’s 47.4% with the green “0” and “00”) such as even and red. My wishful thinking is not only should one of the two come up, but I could potentially win on both. Of course I could also lose on both. Likewise with paired trades, which can give the impression of a hedged position but actually give you two ways lose as surely as two ways win.

A paired position is typically constructed by going long a specific security and simultaneously shorting a related issue. Some examples of classic paired trades would be Nike (NKE) v. Under Armour (UAA), Lowe’s (LOW) v. Home Depot (HD) or even Gold (GLD) v. Silver (SLV). The notion is one is trying to identify a mispricing based on valuation, upcoming catalysts or technical analysis.

The goal is to arbitrage this seeming price differential while assuming any macro move for the sector will impact both securities in much the same way and therefore minimize losses. If housing continues to recover both Lowe’s and Home Depot should move higher.

But as noted above, these pairs can also lose two ways. And unlike the roulette wheel, which is a purely mechanical game of chance whose probabilities should mean revert over time, the stock picking business is path dependent on numerous variables. This presents several challenges in setting up the pair; are the reasons based on valuation, product pipeline, management, technical or a combination of all and some other such as balance sheet issues.

This means one needs to weight the inputs, the difference in stock price and any upcoming events in trying construct a balanced paired position. As always, options can make that job a lot easier.

Pairing with Options

By using options to create a paired position, compared to going long and short the underlying securities, one can lower the cost of the initial outlay and therefore the total risk, and it’s easier to create a balance or tilt the trade with a minor adjustment.

For starters I would suggest simply buying calls on the name you like (bullish) and buying puts on the name you are bearish (or think is overpriced relative to the former.) In doing so you have a defined cost or risk equal to the amount of premium paid, and you can tailor the number of contracts and their strike prices to create a customized pair rather than the more binomial long/short stock. Here are the paired positions I’m looking at:

A bullish position in Advanced Micro Devices (AMD) vs. a bearish position in Nvidia (NVDA)

Both companies are market leaders in developing and manufacturing chips for graphic interface and visual recognition, to be used in virtual reality and self-driving cars. They also both have legacy PC semi-chip lines.

Both stocks have been stellar performers; AMD up some 225% and NVDA up some 175% over the past 52 weeks. And while both have pulled back from recent highs, NVDA’s decline comes after it went parabolic at the end of 2016, then hit a double top at the $120 level and quickly retreated back below $100. This seems to indicate it has put in an intermediate top.

Given paired trades are something of a thematic call based on management, product pipeline or cyclical themes, I’d suggest using options with at least 3 to even 9 months remaining until expiration.

Remember when pairing positions to keep it in balance and risk in check. In this case since NVDA share price is about 6x that of AMD. we should keep a similar ratio in our position size.

Hence:

In AMD I’m looking to buy 6 contracts of the October 13 Calls for approximately $2.30 per contract.

In NVDA I’m looking to buy 1 contract of the September 100 Put for approximately $10.50 per contract

The position obviously enjoys maximum benefit and profit if shares of AMD rally while shares of NVDA decline.