Agricultural Market Curve Shape

Corn, Wheat and Soybeans

Corn curve shape:

2007 contango then lower

2008 contango then lower

2009 contango then higher

2010 backwardation then higher

2011 backwardation then higher

2012 backwardation then ?

 

Corn is evenly split between contango and backwardation with two contangos followed by declines and two backwardations followed by gains. This market, so far, is the only market yet to show roll yield.

 

Wheat curve shape:

2007 backwardation and contango then lower

2008 contango then lower

2009 contango then higher

2010 contango then higher

2011 contango then higher

2012 contango and backwardation then ?

Out of four contango years, three were higher. Out of two partial backwardation years, one is lower and one is unknown. Again, wheat shows no clear pattern except for contango showing higher prices which contradicts the premise of bullish backwardation. 

 

Soybean curve shape:

2007 mixed then lower

2008 contango then higher to unchanged

2009 backwardation then higher

2010 contango then lower

2011 mixed/flat then higher 

Soybeans show some negative roll yield but not much.  Contango markets were split higher and lower as were mixed markets and backwardation was lower.   

Corn and soybeans are the only markets reviewed, so far, to show any indication of roll yield.  For the periods shown, corn appears to have roll yield and soybeans barely shows the same while roll yield does not appear in wheat markets. 

 

Softs Curve Shape: Coffee, Sugar and Cocoa 

The following markets, again, show no clear long-term pattern of lower prices following contango.

 

 

Coffee futures curves:

2007 contango then lower

2008 contango then higher

2009 contango then higher

2010 backwardation then lower

2011 contango then lower

2012 contango then ? 

ICE coffee was contango four out of five years and split evenly lower and higher.  2010 was in backwardation and, contrary to theory, was lower.  Coffee shows no indication of negative  roll yield.

 

Sugar futures curves:

2007 contango then unchanged to higher

2008 contango and backwardation then higher

2009 steeply contango then sharply higher

2010 steeper contango then lower barely lower and higher

2011 backwardated then lower

2012 contango then ?

The interesting sugar curves show markets in stress (2009-2010) yet still no signs of negative roll yield.  Three contango years were followed by mostly higher, higher and sharply higher markets.  The backwardation year was lower and the mixed years were mostly higher.  Again, sugar shows few signs of negative roll yield.

 

 

 Cocoa Futures Curves:

2007 backwardated and contango then higher

2008 contango then sharply higher

2009 contango and backwardated then lower

2010 contango then lower

2011 contango then higher

2012 backwardated and contango then ? 

Cocoa was sharply bullish from 2007 to 2009 and then sharply bearish to 2012.  The three contango years were split higher, higher and lower.  The partly backwardated years were also split higher and lower.  I see no discernible pattern in cocoa. 

The remaining markets, cotton and frozen concentrated orange juice follow. 

 

 

 Cotton curves show the following:

2007 contango then lower

2008 contango then higher

2009 contango then sharply higher

2010 backwardated then lower

2011 backwardated then lower

2012 contango then ? 

Those following cotton may remember the extreme stress the market experienced due to flooding in Pakistan which accounts for the extreme pricing of 2010.  Yet, even in spite of such extreme “positive roll yield”, cotton was lower as of the subsequent year end.  Of the three contango years, the market was split lower, higher and higher.  The remaining backwated year the market was lower.  While on a wild ride, cotton shows no signs of roll yield, negative or otherwise.   

Finally, the last market in the VistaCTA basket, frozen concentrated orange juice is shown next.

 

 

Here we go again:

2007 contango then sharply lower

2008 contango then sharply higher

2009 contango then higher

2010 contango then lower

2011 backwardation then lower

2012 contango then ? 

OJ shows four years of contango with three subsequently higher and one lower. The 2010 backwardation was followed by lower price.  I just don’t see that one data point validates a theory refuted by handfuls of counter datapoints.  Again, there appears little evidence of negative roll yield in oj. 

To sum up: only two markets, corn and soybeans, out of the five ags in the VistaCTA commodity basket show any signs of negative roll yield, and those signs are specious, at best.   

The results of energy, metals and, now, ags, show very little evidence for long-term negative roll yield effects, i.e. bullish backwardation signals or bearish contango signals.  Curve shape, in general, may just be another characteristic of commodity market forward curves that has no predictive value for long-term commodity investors.   

I did not address the seasonal characteristics of many commodity markets.  Such are clearly shown in the forward curves of natural gas, heating oil and gasoline as well, more subtly, in those of some agricultural markets.  In fact, the inflections of a few mixed curves coincide with old crop/new crop designations.  Such designations may occur over shorter periods than those used for this study.  A review of shorter term curve shape effects may modify our results and may be more applicable to shorter-term investors.

Metals Curve Shape

This will be a quick look at metals which are the quintessential “cost of carry” contango markets.  Gold and silver are the definition of non-perishable commodities and nearly always in contango.  Copper futures curves are almost always in contango as the supply chain may be subject to episodes of sudden shortages.

Following is a chart of the year-end forward curves for the Comex Gold futures contract

 

All gold futures curves are contango and each year gold prices rose.  On 12/13/2012, gold was in steep contango and next year’s gold price is unknown. Gold has been one of the most bullish commodity markets of the last ten years and, at the same time is always in contango.  .  Gold is the textbook case that refutes the argument of bearish contango.

Silver follows closely in gold’s footsteps and is sometimes called the “poor man’s gold” since silver’s price level is so much lower. 

Silver, which has some industrial use, is again in contango except for the anomalous far back months of the Dec 2011 and 2012 curves.  And again, contango appears to show no relation to future declines. 

Copper, a truly industrial metal, is subject to occasional unexpected supply shocks, mainly from strikes in copper producing regions.  Under such conditions we would expect to see the futures curve in backwardation.  Supply interruptions can be moderated by large inventories.  The cost of carrying large inventory is relatively low in today’s zero rate world.  Rising rates may bring backwardation back to copper markets.  Looking at the copper curves we note:

2007 contango and backward then much lower price in 2008

2008 contango then much higher in 2009

2009 contango then higher in 2010

2010 backwardation then lower in 2011

2011 contango and backward then higher in 2012

2012 contango and backward then ?

I just don’t see the pattern of contango followed by lower prices or the converse.  Again, generalized curve shape has no predictive value in copper or any other metal futures contract.    

The Rest of Energy Contango

In the prior post, crude oil forward curve shape did not appear to have any ability to predict future return and thereby refuted a popular assertion that “backwardation is bullish” or, conversely, “contango is bearish”.

This post will look at the balance of the energy sector futures: heating oil, natural gas and the reformulated gasoline contracts.  As with crude oil, let’s look at the forward curves on the last day of the prior five years (12/31/2007 to 12/31/2011) and the curve as of our current reference date, 12/13/2012.

Below are the forward curves for the NYMEX No. 2 Heating  Oil futures contract.  Note the undulating effects of seasonality from winter demand.

For example, the leftmost blue curve, showing  the 12/31/07 trade date, has spot at $2.6531 per gallon, Feb08 at $2.6444 and so on out to the Nov10 contract at $2.4436.  In short, the 2007 futures curve is in downward sloping backwardation.  According to conventional wisdom, the backwardated curve is an indicator of higher prices.  A quick look at the year-end 2008 heating curve shows heating oil $1 lower.

 

While 2008 is definitely NOT a typical year, quick looks at the other years shows poor results using curve shape as a predictor, summarized below:

2007 backwardation then sharply lower in 2008

2008 contango then sharply higher in 2009

2009 contango then higher in 2010

2010 contango then higher in 2011

2011 backwardated then higher in 2012

2012 backwardated then ? 

Heating Oil Conclusion: Curve Shape is NOT a predictor of future returns.  

To round out the energy complex, note the following curve charts and corresponding conclusions:

 

 

All Natural Gas futures curves were contango and then all lower!

RBOB 2007 backwardated then sharply lower in 2008

2008 contango then higher 2009

2009 contango then higher 2010

2010 backwardated then higher 2011

2011 backwardated then higher 2012

2012 backwardated then ? 

Conclusion: Throughout the entire energy complex curve shape does not appear to be a predictor of future returns.

Next post will consider the metals and agricultural markets and then look at commodity index rolls, in general.

 

State of Contango

Contango or normal forward curves are alive and well in most commodity markets. As of this writing (the Dec 13 2012 close) the commodities in the VistaCTA commodity basket have the following curve shape characteristics:  

Crude- bump then backwardated

Heating oil- backwardated

Natural gas- sharply backwardated

RBOB gasoline- backwardated

Corn- sharply backwardated

Wheat- normal to flat to slightly backwardated (mostly flat)

Soybeans- backwardated

Gold- contango

Silver- contango

Copper- contango then back down

Coffee- contango

Sugar- contango

Cocoa- mostly contango

Cotton- contango

Frozen Concentrated Orange Juice- contango  

Of the fifteen names in the basket, eight are contango and seven are primarily backwardated.  The question remains, does this mean anything?  Many commodity analysts maintain that contango is a bearish or negative indicator for long investors.  Many ETF and commodity index providers also point to “negative roll yield” (i.e. the presence of contango) to account for the dismal performance of some commodity indexes and ETFs.  Likewise, inverted market curves are claimed to be bullish for long investors.  

Let’s test this assertion.  In the next few posts I’ll review the predictive value of curve shape name by name in the VistaCTA basket.   

WTI Crude Oil Forward Curves

Let’s first consider the NYMEX WTI crude oil futures contract.  There are many ways to do this but a simple way is to look at the futures curves on the last day of the year for the past five years and see if one year’s curve shape can predict the next year’s performance.

The chart below shows the crude oil forward curves on the last day of the year for five years (2007 to 2011) and the forward curve as of the 12/13/2012 close. (I am using “forward curve” and “futures curve” interchangeably here, although, technically, there is a difference.)  The vertical y-axis shows price per barrel and the x-axis shows the contract expiration month.  For instance, the bottom red curve, for the 12/31/2008 trade date, starts at $40.59 per barrel for the March 2009 contract and ends at $74.32 per barrel for the October 2014 contract. 

Aside from the crazy 2008 extreme contango year, 2007 was sharply backwardated; 2010, 2011 and 2012 were generally backwardated, as well.  Did this backwardation predict anything, let alone predict a future bull market?  Looking at the curves note the following:

2007 backwardation then sharply lower in 2008

2008 contango then sharply higher in 2009

2009 contango then higher in 2010

2010 backwardation then higher in 2011

2011 backwardation then lower in 2012

2012 backwardation then ? 

The results show that in the three backwardated years, the markets ended higher, lower and higher.  In the two years of contango, the markets were both higher.  If nothing else, the forward curves indicate that, since 2007, contango is NOT bearish, at least in WTI crude oil.   

Crude Oil Conclusion: Curve Shape is NOT a predictor of future returns.