1
Food prices Is there a long term problem in
feeding the world?

• Christopher L. Gilbert
• University of Trento, Italy cgilbert_at_economia.unit
  n.it

IDEC Crisis Workshop, 26-27 June 2009, Trento,
Italy
2
The recent background

• Major grain and oil seed prices rose sharply in
  2007 through to mid-2008.
• Palm oil led, followed by wheat maize (corn)
  lagged. Sugar remained flat.

Prices fell back, although not to their original
levels, in the second half of 2008, rice less so
than other grains. Prices have staged a recovery
in 2009, although not on a dramatic scale.
3
The wider story

• Movements in agricultural prices were less
  dramatic than those in either metals or energy,
  both of which also rose earlier than ags.
• Agricultural raw materials (natural rubber
  excepted) did not participate in the boom.
• These developments have provoked concerns that
  food prices may be higher over the longer term.

IMF commodity price indices
Many poor and middle income countries remain net
importers of grains. High food prices impacts
negatively on many poor households, in particular
in the cities.
4
Food prices and the crisis

• The sharp rise in food prices in 2007-08 came at
  the end of the long boom and as the financial
  crisis had already started.
• High food prices were the principal way in which
  the crisis impinged on the developing world in
  2007-08.
• A number of commentators have suggested that high
  commodity prices resulted from excess speculation
  and were a bubble phenomenon. A June 2009 U.S.
  Senate report has accused speculation of driving
  up wheat prices in 2008. Phillips and Wu (2009)
  have claimed to have identified a sequence of
  bubbles rolling across asset markets, starting
  with NASDAQ (1999-2000), through U.S. real estate
  (subprime bubble) through to oil in 2008. Ags may
  not have been immune.
• Differently from this, other commentators have
  suggested that food prices may now be high over
  the medium to long term.

5
Structure of this talk

1. The evidence from history.
2. Agricultural investment
3. Biofuels demand
4. Futures markets
5. Dollar depreciation
6. What explains the 2006-08 food price spike?
7. Why are prices rising again?
8. Conclusions

6
1. The evidence from history
7
The historical view

• Prebisch and Singer famously argued that the
  terms of trade of primary producers relative to
  manufactures declines over the long term.
• If world food supply is becoming more difficult,
  we should expect a rising food price trend,
• The historical record is more complicated.
• The choice of deflator is important - export
  unit values and developed country CPIs have
  significant services content which rises over
  time.
• I deflate by the US PPI so that the price is then
  measured relative to other wholesale products.

8
Wheat 1842-2008
The recent rice is prices is quite small by
historical standards
The price trend was upward until 1900 but since
then has been downward
US prices deflated by US PPI
9
Maize, 1860-2008
As in wheat, the recent rice is prices is quite
small by historical standards
The price trend was upward until WW2 but since
then has been downward
US prices deflated by US PPI
10
IMF Agricultural Foods Index, 1957-2008
The 2006-08 spike is very modest relative to that
of 1973-74
The price trend was flat or slightly positive
until 1973 but since then has been downward. It
is now flat again
Deflated by US PPI, 2000 100
11
What the record shows

• Food price trends have been variable over the
  past 150 years periods of rising prices (late
  C19), flat prices and falling prices (past three
  decades). Overall, the trend has been down
  relative to producer prices in general.
• Prices have fallen in the context of major
  increases in production and consumption. There is
  no evidence of demand growth leading to higher
  prices over the long term. So either
• agricultural supply is infinitely elastic in the
  long run, or
• productivity growth has offset any tendency to
  declining returns.

12
Will the future be like the past?

• We cannot be sure.
• Are there reasons to believe the contrary?
• Two factors may be relevant here
• Lack of investment in agriculture over recent
  decades.
• High energy prices (likely to continue for many
  years) may result in food crops being diverted
  into use as biofuel feedstocks.

13
2. Agricultural investment
14
Agriculture and development

• Most poor countries remain predominantly
  agricultural in Sub-Saharan Africa (SSA, not
  South Africa) 34 of GDP and 64 of employment is
  agricultural.
• Poverty is concentrated in rural areas - in 2002,
  three out of four poor people in developing
  countries lived in rural areas.
• Substantial reduction of poverty, as foreseen in
  the Millennium Development Goals, requires a
  rural and agricultural focus.
• Agriculture contributed one third of the growth
  in SSA between 1993-2005.
• The World Bank estimates that 81 of the
  worldwide reduction in rural poverty over
  1993-2002 was due to improved conditions in rural
  areas (19 to migration).
• Agriculture can therefore deliver poverty
  reduction and growth.

15
Developing countries have been under-investing in
agriculture
Agricultural investment has been neglected, in
ag-based developing countries. This trend is also
evident in East Asia
Thanks to Luc Christiaensen, UNU-WIDER
16
Yield improvement is possible
Actual and potential (on-farm demonstration)
maize yields in Africa.
Thanks to Luc Christiaensen, UNU-WIDER
17
Developing countries and the recession(Ravallion,
2009)

• Growth in developing countries will slow as the
  result of the recession but the poorest countries
  will generally be least affected.
• Growth over the past 15 years has moved large
  numbers of people put of 2/day poverty but
  often only just out. These people remain
  vulnerable.
• Slower growth will slow the (net) move out of
  poverty. Some households will move back into
  poverty.
• Impacts will be heterogeneous and will depend on
  circumstances. Households dependent on
  remittances will be worse off. Mining (deep and
  artisanal) will generate less. But some crops
  (cocoa, coffee, sugar) will generate greater
  revenues than in recent years.

18
Urban-rural terms of trade

• In developing countries, a rise in food prices
  shifts the terms of trade in favour of rural
  producers and against urban consumers.
• In many countries, this change would correct a
  long-standing urban policy bias and to that
  extent should be welcomed.

• In 2008, higher world prices were not fully
  transmitted to developing countries because of
  high freight rates. Costs rose because of high
  fertilizer prices. Changes in the terms of trade
  were therefore small.

19
Investment Summary

• In developing countries, the necessary investment
  can be forthcoming but it cannot be left entirely
  to the market. Government action will be
  necessary to develop infrastructure and in
  extension.
• Land is available for agriculture in developed
  economies, e.g. by the elimination of set aside.
  Again, governments need to provide the policy
  framework.
• Larger areas of unexploited or low yield land ia
  available in middle income countries such as
  Russia, Ukraine and Kazakhstan.

20
3. Biofuels
21
Demand and supply

• Movements in prices in 2006-08 were common to a
  large number of agricultural commodities.
• Ag economists generally see demand as stable with
  price rises resulting from supply shocks.
  Although there were some supply problems in 2006
  and 2007, harvests were generally good in 2008.
• Stock levels for many major ags had become low,
  but this was not new in 2007 and 2008. Low stocks
  can explain why a shock has a large impact
  (positive or negative) but cannot explain the
  origin of the shock.
• Because standard explanations of price change are
  unavailable, economists have tended to seize on
  biofuel demand to explain the 2006-08 price
  changes.

22
Biofuels (Mitchell, 2008)

• Biofuels demand was responsible for the largest
  part of the rise in food prices but resists the
  temptation to quantify this share. Abbot et al
  (2008) concurred with this view.
• Maize is the main feedstock crop in the US,
  oilseeds hold that position in Europe, Brazil
  uses sugar cane. Thailand uses cassava while palm
  oil has been most important elsewhere in south
  Asia.
• The global use of maize for feed rose by 1.5
  over the four years 2004-07 while its use as a
  biofuel feedstock grew by 65 over the same
  period. 70 of the increase in maize production
  over this period has gone into biofuels.
• The expansion in maize production was largely at
  the expense of soybeans the 23 increase in the
  US area devoted to maize in 2007 was associated
  with a 16 decline in soybean area.
• The eight largest wheat exporting countries
  expanded the area devoted to rapeseed and
  sunflower by 36 over the period 2001-07 while
  wheat area in the same countries fell by 1.
• Mitchell concludes that biofuels demand drove
  food prices over 2006-08

23
Oil and biofuels Schmidhuber (2006)
Schmidhuber (2006) argues that the prices of
crude oil and fertilizers define a break-even
price for maize and palm oil at which of
biodiesel yields zero profit and similarly for
sugar and ethanol.
if the food demand curve for maize intersects
with the supply curve at a higher price than this
threshold p1, it is uneconomic to use maize as a
biofuel feedstock. The maize price is independent
of the oil price, If the intersection is beneath
the threshold, at po, maize is used in biofuels
production. The long run demand curve becomes
horizontal at the threshold pulling the maize
price is pulled up to this level. A higher oil
price raises the maize price threshold and maize
prices should move in step with oil prices. Maize
becomes a petro-commodity.
24
What do we know?

• The World Bank analysis (Mitchell, 2008) has been
  very influential. It has result in enormous
  pressure being imposed on U.S. and European
  governments to cut back on energy-security driven
  mandates on biofuel production. The E.U. has
  reduced the mandate from 10 to 2.
• Mitchell does not produce any direct evidence for
  the claim that biofuel demand drove up food
  prices. Instead, the argument is that there is no
  other explanation (a residual argument). Is it
  the case that no other explanation is available?
• Schmidhubers analysis leads us to expect that
  there should be a close link between oil price
  changes and food prices over recent years. This
  provides an indirect check on Mitchell.
• I will argue that Mitchells case is overstated
  and that there is as yet little evidence that
  should lead us to restrict biofuels production.

25
4. Futures markets
26
Price formation

• There are active futures markets for many of the
  most important agricultural commodities wheat,
  maize, soybeans, where prices rose sharply over
  2006-08, and also cocoa, coffee, cotton and
  sugar, where there was no boom.
• Active trading allows markets to efficiently
  incorporate information about supply and demand
  fundamentals. If non-fundamentally based trading
  takes place, futures markets can act as a
  distorting lens.
• If prices become too high wont Warren Buffett
  sell? If there are too few experts relative to
  the amateurs, and if the experts have short time
  horizons (for example, because of quarterly
  reporting), they will tend to follow the amateurs
  hoping to get out in time (De Long et al, 1991).
• Many commentators (Desai, Masters, Phillips,
  Soros) have suggested that commodity futures
  prices were a type of bubble over 2006-08.

27
Index investment in futures markets

• This is a relatively new phenomenon.
• Investors have identified commodities as an
  asset class . They see portfolio
  diversification advantages in adding a proportion
  of commodity futures to equity and bond
  portfolios Gorton and Rouwenhorst (2006).
• These position differ from traditional
  speculative positions in several respects
• They are almost invariably long.
• They are typically rolled forward and turn over
  slowly.
• They track specific indices (e.g. DJ-GSCI) rather
  than taking positions on specific markets.
• They can be large in relation to the overall
  market.

28
Index Fund Values and Shares U.S. Agricultural Markets Index Fund Values and Shares U.S. Agricultural Markets Index Fund Values and Shares U.S. Agricultural Markets Index Fund Values and Shares U.S. Agricultural Markets Index Fund Values and Shares U.S. Agricultural Markets
31 Dec 2007 31 Dec 2007 30 June 2008 30 June 2008
bn Share bn Share
Corn 7.6 25.8 13.1 27.4
Soybeans 8.7 26.1 10.9 20.8
Soybean oil 2.1 24.8 2.6 21.7
Wheat 9.3 38.2 9.7 41.9
Cocoa 0.4 11.3 0.8 14.1
Coffee 2.2 26.0 3.1 25.6
Cotton 2.6 33.0 2.9 21.5
Sugar 3.2 29.0 4.9 31.1
Feeder cattle 0.4 23.2 0.6 30.7
Live cattle 4.5 48.4 6.5 41.8
Lean hogs 2.1 43.6 3.2 40.6
Total 43.1 26.9 58.3 27.1
Since 2006, the CFTC has published figures on the
offsetting futures positions taken by index
providers. These can account for up to 40 of all
outstanding positions on these markets. Elena
Corazzella and I have constructed an index of
these positions, IF, which we use in the
subsequent analysis.
29
Index positions and ag prices move broadly
together
30
5. Dollar depreciation
31
Exchange rates and prices

• Most ag prices are denominated in terms of the
  U.S. dollar.
• A decline in the value of the U.S. will rise
  these prices.
• A simple argument (Ridler and Yandle, 1972) shows
  that the elasticity must be less than unity if
  not, a uniform depreciation of the dollar against
  all countries would raise prices in all
  currencies, not just the dollar, upsetting market
  balance.
• This result extends to the elasticity for a
  general set of exchange rate changes provided the
  exchange rate index is appropriately weighted.
  The weights reflect production and consumption
  shares and elasticities.
• In the case that elasticities are equal across
  countries, the elasticity is one minus the U.S.
  dollar share or production and/or consumption.
• Exchange rate effects on prices are small but
  consistent over time. Accurate estimation
  requires a long sample.

32
Exchange rates and ag prices move broadly together
This correlation (0.96) is even higher than that
with futures index positions (0.81). The two
explan-ations compete with each other
33
6. What explains recent food prices?
34
What explains food price rises?

• There are three candidates
• Dollar deprecation if the dollar is worth less,
  dollar food prices will be higher. Unrestricted
  estimation gives an exchange rate elasticity in
  excess of unity. I impose an elasticity of 0.53
  estimated over data from 1971-2008.
• The high oil price oil and related products (in
  particular fertilizers) are inputs into food
  production. Biofuels demand will increase
  consumption.
• Futures market activity we have seen bubble-type
  activity in other markets, so ag markets may not
  be immune.

35
Cointegration

• I have 39 monthly observations from 2006/1 to
  2009/3. This forces a parsimonious analysis.
• Is there an equilibrium relationship between
  these three variables and the IMFs agricultural
  food price index in the Engle-Granger sense that
  the food price index reverts back to the level
  implied for equilibrium.
• I investigate this question in two ways
• I use OLS to estimate the long run equation, as
  the initial stage of the Engle-Granger two stage
  procedure.
• I use the Johansen procedure to test for the
  cointegrating rank in a VAR(2) model linking the
  levels of the four variables.
• In both cases, I establish the presence of a
  single cointegrating relationship. However, the
  unrestricted Engle-Granger procedure gives an
  exchange rate elasticity in excess of unity.
  Imposition of a unit elasticity is acceptable.

36
(No Transcript)
37
Analysis of Peak Food Price Change (restricted vectors) Analysis of Peak Food Price Change (restricted vectors) Analysis of Peak Food Price Change (restricted vectors) Analysis of Peak Food Price Change (restricted vectors) Analysis of Peak Food Price Change (restricted vectors) Analysis of Peak Food Price Change (restricted vectors)
30 month change Johansen Engle-Granger Engle-Granger
30 month change Johansen One step Two step
Dollar depreciation lnX 17.5 9.3 9.3 9.3
Oil price lnO 71.1 12.8 11.8 11.3
Futures index positions lnFIP 51.6 13.0 14.0 24.0
Total 35.1 35.0 41.5
Residual - 21.4 21.4 14.9
Agricultural food prices lnIAF 56.4 56.4 56.4 56.4

• The oil price elasticity is almost exactly equal
  to the 17 pass-through estimated by Baffes
  (2007). The estimated impact is less than the
  15-20 estimated by Mitchell (2008). This impact
  can be accounted for in terms of upward shift in
  the supply curve without recourse to a shift in
  demand.
• All three estimates imply some inflation of food
  prices from futures market activity.

38
Getting behind the numbers

• The preceding analysis identified dollar
  depreciation, the rising oil price and futures
  activity as the proximate causes of the 2006-08
  price rises. Where does this leave biofuels
  demand?
• To the extent that food prices were driven by
  exchange rate changes, there is no room for a
  biofuels explanation.
• There are three possible channels
• The higher the oil price, the more attractive
  becomes biofuels production. Anticipating this,
  the market bids up the prices of grains and oil
  seeds in relation to rises in the price of oil.
  The low estimated oil price elasticity makes this
  unlikely.
• Index futures positions may be speculations on
  future biofuels demand. This seems implausible
• Biofuels may be in the residual (unexplained)
  component of the food price change (but how can
  we know?)
• I suspect that biofuels demand is indeed an
  intriguing and perhaps significant story
  relating to particular markets (Cooper and
  Lawrence, 1975) and not a major explanation for
  recent food price changes.

39
7. Why are food prices rising again?
40
2009

• Food prices are rising again in 2009 from
  post-Lehman lows. Why?
• Two factors are important
• Fertilizer prices are only now falling.
• Restocking.
• Rice played an important element in this process.

41
The rice story

• Rice is the principal food of many poor people.
• An increase in the rice price has a negative
  impact on poverty levels worldwide.

Bangkok rice export price. Source World Bank
The cause of the 2007 rise in the rice price was
the Indian governments misguided decision to
prohibit rice exports. (India is a major rice
exporter).
The governments objective was to protect its
population from the effects of the rising wheat
price but at the expense of the poor in
rice-importing countries.
42
Trade or stocks?

• Developing countries, in which food is the major
  item of household expenditure, are necessarily
  concerned with food security.
• Two strategies are available
• Food security stockpiles financially costly
  since resources are tied up in stocks,
  administratively costly and prone to corruption.
• Trade less costly and efficient if shocks are
  uncorrelated across countries.
• Development agencies, particularly the World
  Bank, have pushed for trade-based food security.
  This did not work in 2008, particularly in rice
  (Haiti, Philippines).
• Governments have learnt that they cannot rely on
  trade and hence are rebuilding depleted
  inventories. This is pushing up world food prices.

43
Trade based policy measures adopted(Joe Dewbre,
OECD)
44
Fertilizers
Prices moved up sharply in late 2007, lagging
food prices. Phosphates and urea turned down in
mid-2008, too late for the southern hemisphere
spring.
Most fertilizer prices are now reasonable for
current northern hemisphere crops. Potash remains
suspiciously high.
45
The outlook

• Provided harvests are reasonable, I expect food
  prices to come down over the coming year
  fertilizer prices are falling and restocking
  should be over.
• However, the dollar remains fairly low and may go
  lower, pushing dollar prices up.
• Low oil prices has resulted in biofuels
  production has being very unprofitable in Europe
  and the U.S.A. over the past nine months. In the
  U.S., it has been sustained by the biofuels
  mandate. Prospects look better with oil at 70/bl
  but maize prices remain above the Schmidhuber
  threshold.
• My view is that 70 is near the maximum OPEC can
  hope to sustain without encouraging production of
  non-conventional oil. But if oil does get back
  towards 100, this may pull food prices further
  up.

46
8. Conclusions
47
Conclusions (1)

1. The world can feed itself but only with the
   necessary investment.
2. Developing countries can increase yields
   relatively easily we do not need a second
   green revolution.
3. The role of biofuels demand in driving up food
   prices has been exaggerated. It is premature to
   take action to limit biofuels production.
4. Dollar deprecation and futures index investment
   were more important factors in driving up food
   prices. They are highly correlated and it is
   difficult to disentangle them.

48
Conclusions (2)

• Attention should be given to the operation of
  commodity futures markets and their effects on
  prices. It would be extreme to prohibit certain
  types of actors or investments but greater
  transparency is desirable.
• Agricultural prices are still high because of
  high fertilizer prices and reversion to
  stock-based food security. Governments should
  push for a new WTO protocol limiting the ability
  of countries to impose food export bans or taxes.
• My expectation is that food prices will fall
  over the coming year.

49
Thank you for your attention