Projecting egg prices
A report detailing possible sources of infulence for egg pricing is examined.
The recently released comprehensive report "UC Poultry Memo 81" prepared by Don Bell, poultry extension specialist (emeritus), epitomizes the quotation, "Forecasting is very imprecise especially when it involves the future."
The report provides historical data relating to flock size, egg production and UB prices from 1985 to the present. What is evident is that the forecasting models based on pullet chick placements which generated good correlations with egg prices from 1985 though 1994 are unable to provide reliable predictions from 1995 onwards. Applying regression analyses, Don Bell showed a -0.61 correlation coefficient between hen numbers and price and a -0.64 correlation between egg numbers and price for the period 1985 to 1994. This means that 61% and 64% of the variation in price could be attributed to the determinant factors (hens and egg numbers) respectively. Using the 1985 to 1994 hen population values, it was possible to predict that each additional 1 million layers depressed price by approximately 1 cent/dozen.
For the subsequent period extending from 1999 through 2008, a different relationship between supply (hen numbers or egg production) was determined. Regressing hen numbers against price yielded a weak correlation coefficient of +0.12. In the case of egg production the correlation increased numerically to +0.28. This denotes that increasing the number of hens or the total eggs marketed resulted paradoxically in higher UB prices. This is clearly counter intuitive and suggests that factors other than simple production influenced price, especially in the unprecedented "boom years" of 2007 and 2008 with average UB prices of $1.17 and $1.31/ dozen respectively.
For an appraisal of the various factors contributing to the apparent anomaly, Bell cites the following possibilities which influence the demand side of the pricing equation:
- The disparity between consuming population and number of layers - For 1985 the ratio of hens to population was 1.02. In 2008, the proportion was 0.94 based on 287 million hens and 305 million U.S. consumers. The actual situation may be influenced by the large number of undocumented aliens who are not counted in the population but are known to consume eggs at a disproportionate rate to the census-based "official' population.
- Per capita consumption - This has increased from 242.6 eggs over the 1985-1994 period compared to 253.6 eggs from 1999-2008.
- Breaking - The obvious increase in egg-products from 20% of production in the "early period" to 30% during the past nine years.
- Exports - Shipments of both shell eggs and products have not materially influenced domestic supply to demand relationships, having maintained a relatively constant proportion of 2% to 3% of total production over multi-year periods. It is evident, however, that in the short-term, large export consignments may have had a marked effect on price. This is attributed less to direct consumer response than to the reaction of chain-store buyers to perceived shortages attributed to exports or disease. Buyers have become a progressively larger influence on price and this factor may not be fully appreciated in the context of recent trends.
- Seasonal price fluctuation - Annual cycles are inherent to the industry and usually result in a 20% variation above and below average annual egg price. The timing and amplitude of these seasonal oscillations may have altered the relationship between supply and demand.
- Consolidation - Acquisitions have resulted in fewer industry decision makers who plan flock placements, molting and depletion to balance supply and demand. Smaller-scale producers supplying local markets are obliged to follow prevailing prices since transport of eggs and the emergence of national chains with large distribution centers have altered the traditional relationship between supplier and customer.
- Possible defects in the discovery system - Aberrations in price may be induced or amplified by deficiencies in the discovery system. The marked escalation and fluctuation in price during 2007 most certainly challenged a relatively simple system based on interviews with suppliers and buyers in a rapidly changing price environment.
- Cost of production - During the past three years the industry has entered uncharted territory with regard to ingredient costs. Escalations of large magnitude obviously affect the production side of the equation. Prolonged periods of negative returns influence flock placement and depletion programs resulting in fewer producing hens. The actual cost of production however has no direct effect on price in the short term since there is insufficient leverage to pass on cost in a market which is mildly oversupplied or at equilibrium. In reality, our eggs are "elastic" with respect to price, with small increases in supply reflected in disproportionate depression in realization.
Additional factors not considered by Don Bell in his review of forecasting of egg prices might include the following:
- The effect of an increasing volume of national and house brands.
- Emergence of cage-free, organic and specialty eggs which are selectively purchased by specific segments of the U.S. consuming population.
- The effect of large national chains including members of the FMI and NCCR.
- The state of the economy which has resulted in a greater appreciation of price value relationships for protein, especially among lower and middle income consumers.
- The influence of reduced stocking density associated with the introduction of the UEP Welfare Program.
The obvious discrepancies between forecasts of price and actual realization during the past five years have resulted in the refinement of predictive models. Chick placements alone are an inadequate basis for predicting price. Don Bell introduced short term adjustments based on interviews in addition to 24-month hatch values in order to improve the precision of the University of California prediction model.
Despite his efforts, the latest report contains the caveat: "Solid data for projections are limited to hatch data which can be projected ahead by 5 months. Everything else is subject to the application of historical patterns or guesswork."