Profitability Forecasting Methodology

Introduction

In 2000, the ChemSystems team (then under IBM ownership) began a project to build a new simulation model of the global petroleum and petrochemical industries. The state-of-the-art simulator was designed to be a fully integrated model of the global business dynamics (global material flows and cash flows) using the latest simulation software. It took the 30+ years of Chem Systems knowledge and experience of the global industries and produced algorithms to simulate business dynamics in petrochemicals, polymers, intermediates etc. The resulting simulator has been a major advance in supply/demand and profitability forecasting technology.

Major development of the simulation software was completed in early 2002 and the new "ChemSystems Online® Simulator" is now used to generate all PPE reports and the output to a subscription website. Since early 2003 it has also been available to clients to run private simulations using their own oil prices, consumption drivers, capacity listings etc. The private simulations are held in total confidentiality on Nexant's servers and accessible via the internet or can be run on the client's own intranet or network.

The simulator relates market demand drivers to petrochemical consumption. From a database of petrochemical processes and plant capacity the regional consumption is then compared to the ability to produce. Global trade algorithms, driven by a comprehensive logistics model, add to the simulation of trade and build to a full supply, demand and trade model of the industry. Basic commodity theory dictates that market tightness, measured by average operating rates, is the primary driver of profitability with inter-regional competition and inter-material competition adding to the complexity of price and cost drivers.

Price Influences

The primary drivers of price for most products are a combination of the cost of raw materials and the supply/demand balance of the market. These two combine to drive cost plus margin to derive price.

The variable cost of production is determined from raw material costs and the cost of utilities less credits for co-products. To this are added the fixed costs associated with running the plant consisting of operating labour, maintenance, general plant and works overheads and tax and insurance to give the cash cost of production. Cash costs for the forecast period are projected based upon raw material costs (usually influenced by crude oil) and the other associated costs of production, making assumptions about the reduction of costs over time due to experience curve effects. The margin is determined from the return on investment (ROI) forecast which, in turn, is derived from an analysis of the historical relationship of margin with average industry operating rate. The combination, cost plus margin, making allowance where appropriate for freight and packaging, produces the price.

Secondary influences on the price forecasts include:

• Forecast prices in other regions (USGC and South-East Asia)
  
• Relationship to related products (e.g. inter-polymer relationships)
  
• Profitability of upstream and downstream processes.
  

Analysis of the range of costs in a market uses the Leader/Laggard methodology to model the higher and lower cost producers and to forecast the limits on upper and lower margins.

Standard Leader and Laggard cost models are defined in the Quarterly Business Analysis Supplement and available online in the techno-economics section of our website. The operating rate projections are developed in the market dynamics analysis and are discussed in the Petrochemical Market Dynamics reports and also on ChemSystems Online®.

Experience

The development of cost models for the Leader and Laggard includes an element of cost reduction through operating experience. Historically there has been a progressive reduction in the cash cost of production for both Leader and Laggard models based on technological developments (such as plant capacities and specific consumption factors) and business operational improvements (such as fixed cost reductions). These so called "experience curve effects" are forecast to continue.

As an example, polypropylene cash costs of production (excluding propylene) have declined by an average of over four percent per year in real terms since 1983 and cost reductions are expected to continue through the forecast period but at a lower rate, averaging around one percent per year. This reflects the considerable developments in polypropylene manufacture since 1983.

Driving Forces for Profitability

Extensive analysis has established that prices of petrochemicals and polymers can fall to the cash cost, and sometimes even the variable cost, of Laggard units at the bottom of a cycle. During a cyclical upturn prices will rise to much higher levels such that Laggard, or high cost, producers cover cash costs and generate a positive margin. The extent to which margins rise is dependent upon the tightness of the market, which may be measured by average industry operating rates.

In general, the margin (expressed as percent return on new capital for Leader units) for the petrochemical industry as a whole shows a good correlation with the aggregate operating rate for the petrochemical industry.

Prior to 1980, each region's petrochemical industry could be characterised as being isolated from the rest of the world. Whilst operating rates and margins were still related, a higher return was achievable at a lower operating rate than in subsequent years as the industries became increasingly influenced by external trade volumes and pricing.

Cyclicality of the Petrochemical Industry

Petrochemical industry profitability has been considerably volatile (or cyclical) and this is expected to continue due the combined effects of demand cycling (due to macro economic performance) and of supply cycling (due to the degree of over or under investment in the industry).

The ChemSystems Online® Simulator produces a price and profitability forecast that is inherently linked to the market dynamics scenario projected. Thus projected cycles in operating rates (due to variations in the balance between supply and demand) result in corresponding cycles in profitability. Capacity expansions in the next two to three years are known and factored in to the projections. Beyond the short term the outcome of the cycle is speculative since it depends on unforeseen events as well as economic performance and the degree of over or under investment in the industry.

Trend Prices

In previous years forecast trend prices have been displayed in addition to the cyclical prices. This was valid when the underlying crude oil price was presented as a long term trend. However, the underlying crude oil prices are no longer truly trend and so the validity of presenting petrochemical prices as trends has been challenged.

The desire to use trend forecast pricing is largely driven by doubt over the timing of the cycles as there is clearly no historical evidence that price have ever shown any trend. The use of trend pricing in project evaluation or business planning presents other challenges, particularly the failure to incorporate the implications on cash flow, debt coverage and project return at the bottom of a cycle. The recommended approach to alleviate the doubts over cycle timing is to run sensitivities that vary the timing of the cycle rather than eliminate cycles by the use of trend prices.