Superior price forecasting in volatile commodity markets - Holistic views of short-term market perceptions dramatically improve reliability of petrochemical price forecasting

Superior price forecasting in volatile commodity markets Holistic views of short-term market perceptions dramatically improve reliability of petrochemical price forecasting For petrochemical firms, purchasing and trading commodities are critical business functions. Price forecasting provides essential support for day-to-day business decisions, especially as unexpected price movements can badly damage a firm's competitive position and bottom-line. “ PA has exceeded our expectations Working with commodity experts from a leading petrochemicals firm, regarding forecast PA Consulting Group (PA) applied its dynamic simulation methodology to accuracy after we develop a robust commodity Price Forecaster. When applied to the global had unsuccessful market for a particularly volatile commodity – where 40% of all month-to-month projects with seven contract price changes exceed 7% in magnitude – PA substantially increased other consultants. the reliability of the firm’s monthly price forecasting. The approach of combining Increasing forecasting accuracy Dynamic As one of the world’s largest buyers, PA's client has been considered the global Simulation with our price forecasting benchmark for the commodity in question. Yet, seven previous experts’ knowledge attempts to computerise their price forecasting failed. When PA was asked to has proved to be tackle the problem, the team first gathered the views of client personnel on the the best approach workings of this particular commodity market. Subjecting those opinions to for short-term extensive dynamic testing filtered out much “knowledge” that proved commodity price (counterintuitively) to be unrelated to short-term price movements. forecasting.” PA's resulting dynamic Price Forecaster focuses on the shifting perceptions of Project Leader the big players in this commodity market, rather than on the fundamentals of the Client market itself. 1 Via a user-friendly interface, the Price Forecaster quantitatively expresses the strength and timing of the causal links between the factors that shape players' shifting short-term perceptions and resulting commodity price movements. Integrating market views and information across organisational boundaries The client firm had long employed an informal committee-based approach to price forecasting. But like many large companies, the organisation was highly compartmentalised with little sharing of market data and expertise across related functions and business units. This resulted in fragmentary, incomplete knowledge and sometimes in conflicting trading and procurement actions. In addition to sharply increasing the reliability of the client's committee-based approach to forecasting, PA's Price Forecaster encourages and reinforces collaboration within the groups that possess relevant commodity market knowledge. For the first time, the client has a rigorous quantitative platform that is applicable across various markets, segments and departments for making, testing, discussing and communicating market views and commodity price forecasts. It also enables commodity experts to run reproducible scenario analyses that ensure learning across organisational boundaries. The Price Forecaster embodies the organisation's best knowledge, making it accessible and rigorously useful under constantly shifting market conditions. Delivering successful forecasts to improve the bottom-line PA's Price Forecaster improved monthly price forecasting reliability well above client expectations: Corporate headquarters • forecasts are within the 5% error rate target in 85% of months and average 123 Buckingham Palace Road less than 2.5% error in those months London SW1W 9SR United Kingdom Tel: +44 20 7730 9000 • error rates average 7.7% in the remaining 15% of months (usually due to Fax: +44 20 7333 5050 stochastic events such as unanticipated plant outages) E-mail: info@paconsulting.com www.paconsulting.com • the Price Forecaster is 88% reliable at predicting large directional changes (of more than 2% in magnitude) in monthly contract prices. PA Consulting Group is a leading management, systems and technology consulting firm, Even on the narrowest of measures, the client's return on investment for the operating worldwide in more Price Forecaster is spectacularly high. Disregarding increased profits from than 35 countries. trading and downstream products, if superior and reliable forecasting enables Principal national offices in Argentina, China, Denmark, the commodity to be purchased in a single regional market for an average of Germany, India, Ireland, 1% less, the client earns back their investment in the Price Forecaster in less Malaysia, The Netherlands, New Zealand, Norway, than two weeks – with an annual return of more than 25:1. Sweden, United Arab Emirates, United Kingdom, United States © PA Knowledge Limited 2008. All rights reserved. 00416-18