Here in mid-September, all of the stores are gearing up for the Halloween season. Costumes, toys, decorations, and candy by the truckload are now overflowing the shelves. That got me to thinking that it might be a good time to take a look at the spread between the cocoa market and the world sugar market. I know I’m stretching things a bit with this one, but I just can’t help it…I have a sweet tooth!
Looking at the market price data since the early ‘70s, there appears to be some correlation between the price of cocoa and the price of sugar. These two ‘soft’ commodities have often trended together. However, there are plenty of times where the two markets diverged as well. The divergence would sometimes persist for a year or two at a time. Somehow, the cocoa and sugar markets would still manage to get back together.
Correlation without Causation
Although we see several time periods where cocoa and sugar trekked together through bull markets and bear markets, it is hard to pinpoint a strong fundamental relationship between the two markets. Oh, there is the obvious point that sugar is used to sweeten cocoa for consumption. The multi-billion dollar Hershey Company will certainly attest to that. However, sugar is also used heavily in ethanol production and cocoa is not. Furthermore, it is not likely that the crops of cocoa and sugar are affected by the same weather situations. The Ivory Coast is the largest grower of cocoa on the planet and Brazil is the world’s largest sugar producer. They are not exactly close neighbors.
Nonetheless, there is some sort of observable synchronicity between cocoa and sugar. Enough so that a daring (and well-capitalized!) spread trader might want to take a shot at it from time to time. You just need to be selective about when to do it.
Wait For the Extremes
As you probably have guessed by now, our philosophy is to wait for inter-market spreads to reach historic extremes before we start getting excited. This is because we want to find a spread relationship that is mean-reverting and then start looking for a reason to bet on a mean-reversion once it has neared or reached previously unsustainable levels. In other words, we like to bet only when the odds are favorable.
A cocoa futures contract controls 10-tons of cocoa and a sugar #11 futures contract controls 112,000 lbs. of sugar, so comparing the two can seem a bit tricky. To simplify things, we just convert the spread to the difference between the values of one cocoa futures contract and one sugar #11 futures contract.
Four decades of history indicate that a middle-of-the-road area for the cocoa/sugar spread is somewhere between +$2k (premium cocoa) to ‘even money’ (where the two contracts have the same value). The spread is historically ‘expensive’ when it reaches +$10k or higher and it is historically ‘cheap’ when it inverts and sinks to -$5k (premium sugar) or lower.
A week ago, the spread between the nearest-delivery cocoa contract and the nearest-delivery sugar contract surpassed the +$15k mark for the first time since the spring of 2010. There are only five other times in the last four decades when the cocoa/sugar spread has made it to +$15k or higher (basis the nearest-futures monthly chart). Each time the spread eventually reversed and dropped all the way back below the ‘even money’ mark. Perhaps this could happen again?
For the last few years, cocoa has traded at nose-bleed levels not seen since the late 70s/early 80s. Price extremes in a market will certainly cause a related spread to hit extremes as well. Therefore, we filter the spread by looking at the ratio between the two markets.
On Friday the ratio between nearest-delivery cocoa contract and the nearest-delivery sugar contract hit a nearly six-year high of 2.16:1 (a cocoa contract is worth a little more than twice as much as a sugar contract). Historical charts suggest that a normal level is somewhere around 1.2:1. A ratio of 2:1 has only materialized a handful of times in the last few decades. Therefore, the high ratio level confirms our suspicions: the cocoa/sugar spread really is expensive! Spread traders should be looking for a setup to get short.
Trend Change Signals
Unfortunately, a spread between the nearest-delivery cocoa contract and the nearest-delivery sugar contract would mean that we have to use the December cocoa contract and the October sugar contract. This is a problem because the October sugar contract goes off the board in a week and a half. So we are going to focus our efforts on the March cocoa/sugar spread instead.
We see at least three possible technical setups that a trader could use to identify a trend change:
First, consider the basic bullish pattern of higher highs and higher lows. In the last six months, the March cocoa/sugar spread has only traded below a prior month’s low one time. Also, the spread has made higher monthly highs for five consecutive months. Therefore, a break of a prior months low would alter this pattern and indicate a possible trend change.
Second, the March cocoa/sugar spread has closed above the rising 50-day Moving Average every single day for nearly three months straight. A close below the 50-day MA (currently around +$11,602) for the first time since late June would signal a bearish trend change.
Third, consider zooming out to a larger timeframe to filter out some of the noise on the daily timeframe. The March cocoa/sugar spread has closed above the rising weekly 10-day Moving Average every week since the first half of May. A weekly close below the weekly 10-day MA (somewhere around +$12,150 starting on Monday) would signal a bearish trend change.
The weekly 10-bar MA is similar to the daily 50-bar MA since a weekly bar consists of approximately five trading days. However, a filtering effect occurs since a crossover of the weekly moving average requires an end-of-week close below the moving average. It will escape the possible whipsaw of a mid-week close below the 50-day MA that gets reversed a day or two later. This can come at a cost, though. If the 50-day MA is broken at the start of the week and the spread continues to plunge for the rest of the week, the spread can be significantly lower by the time the weekly trade signal kicks in. You have to weigh the pros and cons and decide what best suits your own style.
For tracking purposes, the blog will make a hypothetical trade by selling one 10-ton March cocoa contract and simultaneously buying one 112,000 lb. March sugar #11 contract if the spread closes below the rising weekly 10-day MA or closes below the August low of +$11,216, whichever occurs first. Initially, the spread will be liquidated on a two-consecutive day close $500 above the contract high that precedes the entry.
If the ratio between March cocoa and March sugar surpasses 2:1 before a reversal signal is triggered, we may revise the trade to short one cocoa contract against two sugar contracts. We will post an update is this occurs.