1. AAMP Training Materials
Module 3.4: Price Stabilization
Nicholas Minot (IFPRI)
This is the fifth section of the Price Analysis Module which is part of the AAMP Training Materials.
Exercises for this section are found in the excel spreadsheet entitled:
Module 3.4 – Price Stabilization.xls

2. Objectives Understand definition & measures of price instability
Identify sources of price instability
Use Excel to simulate price instability
Examine the relationship between price instability and income instability
Evaluate the effect of trade on food price stability
Discuss the role of buffer stocks in stabilizing food prices

3. What is food price instability?
Food prices in Africa are not constant over time
A great deal of variation is present which can impact incomes
This variation is called food price instability
The figure above shows inflation adjusted wholesale maize prices in Dar es Salaam, Tanzania. Over the sixteen year period displayed, there has been a great deal of variation in maize prices. Dar es Salaam is the largest city in Tanzania, and maize is one of the most commonly consumed staple foods there. Looking at the figure, there appear to have been price changes of over 100% in a single year.
Seeing this figure, one might ask:
How unstable are Tanzania’s maize prices relative to other countries?
What are the effects of maize price instability on incomes?
How can instability be measured?
Can price stability be moderated?

4. Measuring Food price instability
Measuring food price instability using CV
Coefficient of variation (CV) = standard deviation / average
Adjusted coefficient of variation = CV with correction to remove effect of the time trend
Calculating CV in Excel
= stdev(range) / average(range)
Example: = stdev(b3:b40) / average(b3:b40)
Simulating a random variable in Excel
=norminv(rand(), mean, stdev))
Example to generate a random variable with mean = 200 and CV = 20%, standard deviation will be 40 so
= norminv(rand(), 200, 40))
The coefficient of variation (CV) expresses the variability of a sample as a percentage of the mean. It is very useful for comparing between data sets with widely differing means, such as maize price between regions or countries. It gives an intuitive sense of the dispersion of the data.
In Excel, it is easy to produce a column of numbers which take on a specific CV for comparison.

5. Random variable CV examples
The examples above were created using random variables in Excel. At CV = 10%, the data remain close to the mean of As variation increases, the data points stray farther and farther from the mean.

6. Generating random variables - Excercise
Open [Ex1 - CV of Random Variables]
Examine the formula for the column of numbers
What does $C$3 mean?
Press F9 to recalculate several times
Why do the numbers and the graph change?
Change standard deviation to 40, then 60
What happens to the graph?
Change standard deviation to 10, then 5
Look at column G and recalculate several times
Why are “actual” average and std deviations different than column C?
The exercises for this module are in the excel workbook entitled: Module 3.5 – price_stabilization.xls
$C$3 fixes the reference to cell C3, even when the equation is copied to other cells.
The numbers change because F9 recalculates, which generates a new set of random numbers
As standard deviation increases, the points are farther from the mean, implying that prices become more instable
As standard deviation decreases, the points are closer to the mean, implying that prices become more stable
“Actual” numbers differ from column C for the same reason that flipping a coin 10 times won’t always give you 5 heads and 5 tails

7. Magnitude of maize price instability
After working with the CV of random variables, this figure should be very interesting. It compares seven African countries’ coefficients of variation on maize prices. In general, food prices in Africa are very volatile compared to developing countries in Asia.
Thinking back to the CV of random variables exercise, imagine what Ethiopia’s maize prices must look like in a line graph with over 50% CV. In comparison, think back to Tanzania’s data, graphed in slide 5, which has a CV of about 26%. How does such high price variation impact on household incomes among the poor? How does such high variation impact maize producers in these countries?
By comparison, for six Asian countries, the CV for rice prices ranged from 12% in Bangladesh to 25% in the Philippines.

8. Price Stabilization – Model
Examine [Ex2 – Stabilization] of the Excel workbook
This model will be used in several exercises
The yellow boxes contain parameters that can be changed to simulate markets for different commodities and countries
General Assumptions – Sets characteristics of the domestic market
Trade assumptions – Sets characteristics of international trade and policy assumptions
Buffer stock assumptions – Sets buffer stock policy and cost assumptions
More…
Now that you have examined what price instability is and how it can be measured, look at the model of price stabilization in Excel. The model allows you to compare different market situations and compare the outcomes.

9. Price Stabilization – Model
The boxes highlighted in light green contain the outcome of the assumptions made in the yellow boxes
Average and CV of several variables of interest
Graph 1 shows prices with and without international trade
Graph 2 shows prices with and without a buffer stock
Examine [Data – Stabilization]
Shows how the outputs are calculated based on the inputs
Note: You can not change values in the green boxes. All of the cells in the worksheet are protected except those highlighted in yellow.

10. Causes of food price instability
Variation in domestic commodity supply
Particularly commodities that are not widely traded e.g. bananas & root crops
Seasonality in prices
Differences in harvest size
Variation in world price of commodity
Particularly commodities imported from world markets (e.g. rice & wheat) or exported to world markets (e.g. coffee)
Maize is more tradable than cassava but less than rice & wheat
Large effect in but generally little effect
Only 13 of 62 food prices in Africa showed significant link to world prices
Tradability refers to the degree to which the commodity is traded on international markets, which determined whether local prices are influenced by international prices. If a significant portion of local consumption (say 20%) is imported every year, the commodity is generally considered tradable. Likewise, if a significant share of production is exported, it is considered tradable. If only a tiny portion is imported or exported or if trade is only sporadic, then the commodity is considered non-tradable. For many African countries, maize falls in a grey area between tradability and non-tradability, since there is cross-border trade and occasional imports from South Africa, but the volumes are small and generally do not occur every year.

11. Causes of food price instability
Changes in size of harvest
Changes in size of harvest
Changes in price
Changes in price
Quantity
Quantity
Elastic supply & demand
Inelastic supply & demand
Many staple foods have inelastic supply & demand (meaning they are not responsive to changes in price)
In this case, small changes in supply (size of harvest) can have large impact on prices

12. Causes of food price instability
Food policy
Trade policy
Buffer stocks
Variation in demand (e.g. holidays)
Changes in closely related markets
Markets in neighboring countries
Markets for substitute commodities (maize & cassava)
Speculative bubbles

13. Supply and price instability – Exercises
Open [Ex2 – Stabilization]
Examine variation in domestic supply of a commodity
Increase CV of production
Decrease CV of production
How does it affect CV of prices?
Price elasticity of demand
Set price elasticity of demand at -0.3, -0.5, and -1.0
How does elasticity affect food price instability?
Why does inelastic demand make food prices more volatile?
What factors determine whether demand is elastic or inelastic?
High variation in production  High price instability.
This is the norm in most African countries that rely on rain-fed agriculture.
Price elasticity of demand = the percentage change in demand given a 1% increase in price. Price elasticity of demand is always negative. For example: If elasticity is -2.0, a 1% increase in price causes a 2% decline in demand.
If the price elasticity of demand is highly inelastic, meaning that demand does not diminish much with increases in price, then food prices can be very volatile. In countries in which large proportions of the population favor one staple food, price elasticity of demand for that food is highly inelastic. In places in which the population has many substitutes for the favored staple food, then price elasticity of demand tends to be more elastic.
Message: A diversified food basket can reduce food price instability.

14. Trade and price instability
Trade sets a natural “price band” within which market prices must stay
But band may be very wide if distance and transport costs are high
Taxes on imports and exports make the price band wider
Import parity price sets the upper limit of price band
Definition: Cost of imported product including transport and taxes
Export parity price sets the lower limit of price band
Definition: World price of an exported good minus cost of taxes and transportation from a certain location to world markets
The import parity price is higher than the world price of a good because it includes import tariffs, unloading, cost of transportation, profit, & risk premium. Export parity is less than world price because you need to subtract export taxes, the cost of transportation to the port, loading, and so on.

15. Trade and price instability – Exercise
Open [Ex 2 – Stabilization]
There are 2 ways to compare with and without trade
Compare the first and second section of the green-shaded table
Compare the red (no trade) and green line (with trade) in figure 1
Change transfer cost to/from world market from $150 to $175
What are the maximum/minimum prices with and without trade?
What is the CV of price with and without trade?
Add 30% import tax and 30% export tax
What are the maximum/minimum price with/without trade?
What is the CV of price with/without trade?
Without trade, prices reach higher peaks and lower troughs than when borders remain open  Higher CV on price.
Message: Closing borders to international trade eliminates the price-moderating effect of the IPP and EPP price band.
Adding import taxes raises the IPP and adding export taxes lowers the EPP  Higher CV on price.
Message: Taxes on trade widen the natural “price band” that international trade provides.

16. Food price stabilization in theory
Idea of a buffer stock
Buy when price is low (bumper harvest)
Sell when price is high (drought year)
Effect is to raise the price when low, lower price when high
Price band policy
Set ceiling price and floor price
Buffer stock willing and able to sell “unlimited” quantities at ceiling price
Buffer stock willing and able to buy unlimited quantities at floor price
Effect is to keep price between ceiling and floor price

17. Food price stabilization in practice
Public food reserves are typically managed by state-owned enterprise
Reserves in main staple cereal and 1 – 2 other grains
Root crops and cooking bananas generally not included because of perishability
Food reserves use different types of interventions
Not just buying and selling, but import and export policy, government imports and exports, regulations of grain marketing etc.
Food reserves do not use consistent buy/sell rules
Intervention depends on budget resources, politics, etc.
Food reserves in developing countries have multiple objectives: Price stabilization, preparation for emergencies, support farm price, keep down consumer prices etc

18. Buffer stock and price instability – Model
The model calculates
Quantity that has to be bought and sold each year to keep price inside the price band
Quantity in storage (initial stock – all sales + all purchases)
Trading costs = cost of buying or selling stock (revenue is negative)
Storage cost = quantity in storage x cost per ton (initially $50/ton)
Transport cost = cost of moving commodity to/from warehouses
Interest cost = opportunity cost of money used to buy stock
Total annual cost
More…
The buffer stock and price instability model is also found in [Ex2 – Stabilization]

19. Buffer stock and price instability – Model
Cumulative cost over the simulation (negative = revenue)
Balance left over = Original budget minus accumulated net costs
Probability of exhausting funds
Probability of exhausting stock
Probability of exceeding storage capacity
Numbers will vary for each recalculation

20. Price instability and income instability
Only affects households if it causes variation in income and consumption
Food price instability
Real income (purchasing power)
Consumption
REDUCED WELFARE
Importance of commodity in consumption or as source of income, level of diversification
Ability to smooth consumption with savings, credit, sale of assets, etc.
Level of income, degree of risk aversion
Food price instability affects real income (purchasing power), but size of impact varies with degree of dependence on food sales or food purchases.
Instability in real income causes instability in consumption, but size of impact can be modest if household can smooth consumption with savings, sale of assets, etc.
Variation in consumption reduces household welfare, but size of impact depends on level of income and degree of risk aversion.

21. Price stability and income stability – Exercise
Set price elasticity of demand to -1.0
What is the CV of gross farm revenue without price stabilization?
Why is it so low?
Set buffer stock price band at 350 and 300
What is the CV of gross farm revenue with price stabilization?
Why is gross farm revenue more unstable with price stabilization?
Set price elasticity of demand to -0.5
Compare CV of gross farm revenue with and without stabilization
Why are the results different?

22. Price stability and income stability – Explanation
For farmers, price stabilization may actually destabilize income
Without Price Stabilization
With Price Stabilization p
In a bad year, high price offsets low output.
In a good year, there is a low price, but high output p
Fig. 1
Fig. 2
Variation in output not offset by changes in price = more income instability
In the figures above, income is represented by the area of the yellow and blue rectangles (price x quantity). When the output is high, income is the yellow rectangle, and when output is low, income is the blue rectangle.
Figure 1 represents the situation without price stabilization. Changes in price and quantity offset each other, so little change in income (yellow and blue rectangles have similar size).
Figure 2 represents the situation with price stabilization. Changes in output are not offset by opposite changes in price, so there is a large difference in income between high-output year (yellow rectangle) and low-output year (blue rectangle). q q

23. Effect of price band width – Exercise
Open [Ex 2 – Stabilization]
Change from wide price band to narrow price band
Lower limit of band is buying price, upper limit is selling price
Compare price band from 275 – 500 to 300 – 400
What is the CV of price before and after?
What is the frequency of purchase and sale before and after?
How does average annual cost change?
What is the probability of running out of funds over 10 years?
What is the probability of running out of stocks over 10 years?
What is the probability of exceeding storage capacity over 10 years?

24. Effect of price band width – Explanation
Wide band implies
Less price stabilization
Less frequent intervention
Lower cost
Narrow band implies
More price stabilization
More frequent intervention
Higher cost
Remember: Because Excel is creating these figures using the random number generator, your figures won’t match exactly.

25. Effect of level of the price band – Exercise
Open [Ex 2 – Stabilization]
Change the price band to 350 – 500
What is the CV of price?
What is the frequency of purchase and sale before and after?
What is the probability of running out of funds over 10 years?
What is the probability of running out of stocks over 10 years?
What is the probability of exceeding storage capacity over 10 years
Change the price band to 250 – 400
What is the probability of exceeding storage capacity over 10 years?

26. Buffer stock If mid-point is too high: If mid-point is too low:
Buying more than selling
Accumulation of stocks
Eventually exhaust funding or storage capacity
If mid-point is too low:
Selling more than buying
Depletion of stocks
Eventually exhaust stocks
One option: Set mid-point at average of last 3 years
These graphs correspond to price bands in exercises on previous page. The top graph shows the result of using a price band of , as in the first exercise. The bottom graph shows the result of using a price band of , as in the second exercise.
The blue arrows indicate the mid-point of the price band, 425 in the top graph and 325 in the bottom graph.
Note that the exercises will not yield graphs exactly like this because random numbers differ each time they are selected, but the general pattern and the results will be approximately the same: exhaustion of funding or storage capacity if the price band is too high and eventual depletion of stocks if the price band is too low.

27. Conclusions Price stabilization is expensive
Large procurement costs (US$ 80M in Kenya, 2006)
Storage, handling and overhead
State enterprises can’t cover costs with stabilization efforts
Aggregate benefits are small
Most estimates 0 – 4% of farm income
Benefits of price stabilization not pro-poor
Most benefits go to larger commercial farmers and urban poor
Food price stabilization prone to “rent-seeking”
Open borders provide no-cost “price band”
Impeding imports has exacerbated price spike in several cases

28. Conclusions Improve consistency and predictability in govt. actions
Promote private grain storage and imports
Credit, non-intervention and storage rental
Promote consumption of secondary staple crops
Cassava can act as a shock absorber for grain markets
Rationale for 3 month grain reserve
To cover period until commercial imports can be arranged
If price stabilization politically necessary
Adopt a rule-based price band
Adopt wide and market-based price band

29. References
Jayne, T.S., Ballard Zulu, and J.J. Nijhoff Stabilizing food markets in eastern and southern Africa. Food Policy 31, no. 4: 13. Minot, N Food price stabilization: Lessons for eastern and southern Africa.” Prepared for the policy seminar on “Agricultural Risks Management in Africa: Taking Stock of What Has and Hasn’t Worked“, September 6-10, 2010, Lilongwe, Malawi. Rashid, S., R. Cummings, and A. Gulati “Grain marketing parastatals in Asia: Why do they have to change now?” Discussion Paper No. 80. Markets, Trade, and Institutions Division. International Food Policy Research Institute. Washington, DC. World Bank Managing Food Price Risks and Instability in an Environment of Market Liberalization. Agriculture and Rural Development Department. Washington, DC.