Friday, March 2, 2012

Do Food Commodity Prices Follow Oil Prices?

People always seem to be interested in the comparison of food commodity prices with oil prices and many have proclaimed that food equals oil.

This next graph going back to the 1950's, taken from this OECD publication, demonstrates that food prices have not followed very closely with oil prices. (Note that commodities are priced in USD's.)


Because the above graph is difficult to read, I've highlighted "food" and "crude oil" in blue and red in the following version:

Concerning this graph, from the report:
This graph shows the evolutions of IMF indices of market prices for primary commodities. The increase in food prices took place in the context of a general rise in commodity prices led by crude oil and metals. However, the 42% rise in food prices and in beverage prices over the period 2006-08 has been modest relative to crude oil prices (51%) but large relative to metal prices (8%).
What are some of the reasons that food prices don't necessarily follow oil prices?

While it is partly true in the industrial agricultural system that "food equals oil," there are many other factors which affect food prices, including the definition of "food" used in making the comparison. Below, I've listed some of them.
  • The dollar's value compared to currencies of other food exporting and importing nations.
  • Supply and demand.
  • Amount of food used for biofuel production.
  • Available infrastructure in transport and storage of food.
  • The price of natural gas.
  • Economic health of each nation.
  • The amount of global meat consumption.
  • Weather.
  • Population growth.
  • The percent of food wasted.
  • Transport prices (not always the same as oil prices, as, for example, currently we have excess bulk shipping capacity which has lowered shipping rates).
  • Government Ag policies and price support programs.
  • Trade agreements.
  • Geopolitics.

The OECD report, which was studying various influences upon price volatility in the agricultural commodities, concluded this:
In the current decade, the agricultural products that are shown to be the most correlated with the crude oil price, based on monthly data, are butter, whole milk powder and soybean oil. In the case of annual data, products with the highest correlations are maize, whole milk powder, wheat and butter. The least correlated are always beef and sugar.
Overall, the study did not find increased volatility in agricultural commodities in recent years as compared to the past fifty years, although the character of the global market has admittedly been changing.

There was one consistent finding most notable, which was not surprising:
The high price events of the past fifty years have typically followed a similar pattern – a price hike in one year followed by a sharp drop in the following year - for most commodities. In addition, past surges in aggregate agricultural prices, as represented by an index of food prices took place in the context of a general rise in almost all commodity prices, crude oil and metals in particular.
Ahh, yes, the laws of supply and demand.


submit to reddit

3 comments:

  1. Nice! Glad to see somebody looking into the assumptions rather than taking them at face value. Also pleased to see biofuels and meat consumption rates on the "likely causes" list. I was at a farming conference this past weekend and they were talking about the crop insurance program (heavily subsidized) and how it's switched to revenue insurance (so we're subsidizing to give people a guaranteed income) and how it's being calculated (high!). I haven't wrapped my head around all the implications yet, but it seems like high guaranteed revenue for commodity crops would induce a land rush in ag areas, and an increase in sensitive lands getting plowed under (this program is exempt from environmental protection clauses)... I feel like it's going to connect to higher food prices too, but like I said before- I'm slogging through a lot of new-to-me material. Any insights?

    ReplyDelete
  2. Molly,
    Excellent remarks and observations here. Thanks.

    Yes, policy is really everything. You might want to follow the writings of the Environmental Working Group, and they have just written on the crop insurance program subject, too.

    We have overproduction which is supported by policy of ethanol to consume corn and the crop insurance program which both, yes, drive up land prices.

    ReplyDelete
  3. One has but to look at energy return on energy invested to gauge to likely correlation between food prices and oil prices. The business of agricultural production requires the input of a considerable amount of energy, and rising oil prices therefore impact farm profitability, which inevitably leads to higher prices for the commodities produced.

    Energy, oil especially, is used in the growing, processing and transportation of food, and therefore food prices will always share some level of positive correlation with energy (oil) prices.

    I consider the present global situation in the is respect to lend itself to a transfer of wealth, from those consuming commodities, to those producing them.

    If we consider that China uses 2 barells of oil per person per year, compared to 17 in South Korea, we can assume that for China to reach the level of industrialisation as South Korea, it would require an extra 44 million barrels of oil per day to sustain (taking into account production decline). Thenm throw in India and other large populations also entering into their most resource intensive phase of indutrialisation, then we see huge demand for oil with finitie resources, which will result in higher prices - but not just of oil, of all commodities where energy is required to produce them.

    In short.... buy an oil field and a farm!

    ReplyDelete