Sunday, August 20, 2006

Pairs Trading and Phelps Dodge

Finally! Here's my pairs trading post. I believe this strategy will soon make significant sense. It hasn't unfolded yet but sometimes markets don't oblige exactly when you want them to. First let's take a cursory look at pairs trading or more commonly called spread trading. Spread trading seems to have found its beginning in the futures market from all of my reading. I don't know who first developed such a strategy but Morgan Stanley takes some credit for birthing it. The concept involves finding two commodities or investments which are highly correlated. They can be positively correlated or negatively correlated. Doesn't matter. They can be seasonally correlated like heating oil and gasoline or they can be inflation oriented like copper and gold or they can be business cycle correlated like semiconductors and oil companies.

Spread trading is a tremendous opportunity to reduce risk while taking very aggressive positions. For equity investors, I guess something like covered calls or strangles or straddles or some other risk reducing strategy would likely resonate. Using stocks as opposed to futures contracts or options contracts is alot easier and cleaner. You don't have to decide which months to invest in, you don't need to hit the trade just perfectly, you don't have a time decay issue and a host full of other issues.

So, the concept is something like this. I don't care if prices go up or down. All I care about is the spread between the prices. So, both investments could go down, up or a combination. The way it works is to take equal dollar positions in two investments when the spread has reached historical extremes, conditions merit a possible reversal of trend or seasonality takes hold. Now, there is alot of very sophisticated work out there around spread trading so I'm really boiling this down the introduction of a concept. Joe Ross has written extensively of this for futures and there are a handful of other good books on the topic over at Amazon. An example of this approach may have been to short semiconductors in 2000 and go long the OIH ETF. Or, go long Exxon which pays a higher dividend than the ETF. (Or in the future use the actual Oil ETF and a semi ETF.)

Now, you might ask yourself why you wouldn't just short SMH. That's a valid point. But, it is an unhedged position. Hedging has the potential to reduce risk and you don't need to know which way the market is going to move. All you care about is the spread. There are specific benefits in the futures market you aren't going to get in the equities market such as reduced margin compared to owning the two individual futures contracts as well as some other benefits. Without thinking too much, there are likely half a dozen other reasons. Read a book. Hopefully, you get the concept.

Now let's break the concept down into the possible scenarios and let's do it using Phelps Dodge, a copper producer, and Newmont mining, a gold producer. Addendum: I actually just reread this post after typing it up last night. I realized I never explicitly stated the trade involves going long Newmont and short Phelps Dodge. Most probably knew this but as a point of clarification I am inserting this point. You could just as easily do this with copper and gold futures but there are alot of implications as I mentioned above such as which contracts to use, getting creamed by sophisticated traders, contract expiration, time value losses, etc. As an aside, you'll likely be able to do this with the gold ETF and a copper ETF at some point in the future. Rumors of a copper ETF have been floating for some time.

Today, the spread between copper and gold is historically very extreme. It's the most extreme on this side of the trade as it has been in twenty years. In fact, going from memory, I believe it is nearly as extreme as it was in the early 1970s commodity explosion. I believe copper topped somewhere near $2.50ish and gold was around $400ish in 1973. Today copper is still around $3.70ish and gold around $650ish. So, the spread in 1973 was about 160:1 and today it is about 160:1. This is all from memory as I am too lazy to look it up but exact numbers aren't relevant to the strategy. Now compare this to the opposite extremes when the gold/copper ratio has reached approximately 500:1. ie, Gold at $250 to $350 and copper at $0.50 to $0.70.

What sentiment does this ratio share with you at its extremes? At 160:1 it is telling you the future economic growth is expected to be off the charts positive. At 500:1 it is telling you economic growth is going to be in the gutter. There could also be implications as it pertains to inflation but let's try to keep this as simple as possible. So, what's the economic future? Will the economy boom on or will it crater. Well, who cares? Seriously. Who cares? In spread trading you really don't care. All you care about is making money on the spread and you want the spread to change. Both stocks go up and it is possible to make money. Both go down or one goes down and one goes up, ditto. Remember this point. In a recession, higher order capital goods typically fall faster than money or money equivalents. Copper is an input into the capital goods sector. While gold is neither, it is considered by many a pseudo money equivalent. Especially in times of heightened uncertainty, deflation or inflation. Back to the trade. Two variables=>Four possible outcomes. Remember the time in the cycle when you consider these four scenarios. No statistics, historical returns or probabilities. That would mean I would actually have to do something. If the trade is of interest to you, the data for such analysis is available.

1) Copper rises and gold rises. In this scenario, you make money if gold rises more than copper. Given high commodity prices eventually creates inflation and economic malaise, is copper going to rise faster than gold at this time in the cycle? Especially if the cycle is slowing or worse? Likely not because from these prices if copper made a significant run, we'd be looking at hyper inflationary pressures and gold would do a moon rocket likely beyond that of copper. High odds of a positive return in the trade.

2) Copper rises and gold drops. Again, similar scenario to the first outcome. At an extreme spread where copper as a percent of gold is already at twenty year highs, what is the chance the spread will go against you with gold moving the opposite direction to copper? High odds of a positive return in the trade.

3) Copper falls and Gold rises. If an economic slow down is near, economically sensitive copper has likely peaked and on a relative term will likely fall faster than gold. High odds of a positive return in the trade.

4) Copper falls and Gold falls. Again, a plausible outcome if inflation is tamed or the economy slows. Yet, in periods of higher risk, copper is going to fall faster than gold which is a hedge for both inflationary and deflationary environments.

There you have it. The trade hasn't set up yet as far as what it would take for me to enter but I'm anticipating it will relatively soon. Go make some money. Just don't do it using my post as it isn't advice.
posted by TimingLogic at 10:19 PM