A central aspect of the dynamic problem facing the firm in an evolving industry is the decision about additions to productive capacity. Particularly in capital intensive industries, capacity decisions have long lead times and involve commitments of resources which may be large in relation to firms' total capitalization. If the firm fails to add capacity at the appropriate time, it not only loses immediate sales and market shares but also may diminish its long-run competitive position—if the firm adds too much capacity, it can be burdened with unmet fixed charges for long periods of time. From a competitive standpoint, additions to capacity can pose major problems since the matching of capacity to demand is often a major determinant of industry rivalry and profits. The problem is most acute in industries producing standardized products, where product differentiation does not protect firms against mistaken capacity decisions of others.
HFCS demand also depended to some extent upon the HFCS capacity that was actually built. Large users of sugar were unlikely to convert to HFCS unless enough capacity was on-stream that they could be assured of gaining adequate supplies and would not be at the mercy of only one or trants two HFCS suppliers. Thus, demand and supply were interdependent; this guaranteed some temporary excess capacity if high demand for HFCS high was to occur.
The profitability of HFCS would depend on the spread between HFCS was costs and prices. Prices would be a function of the capacity utilization in intries the industry and of the price of sugar. If the demand for HFCS were close narket to or exceeded the HFCS capacity put on-stream, then HFCS would be end in • hough expected to be priced relative to sugar (at approximately a 15 percent discount). However, if there were substantial excess HFCS capacity, then HFCS would likely be priced relative to cost. A significant part of HFCS capacity costs were variable, the major element being the cost of corn. However, vert to ssured there were fixed operating overhead and significant capital costs. Judging from experience in the starch and corn syrup markets, in periods of substantial overcapacity the price of HFCS was likely to be near variable costs.
There is of course no logical reason why firms could not have expectations that were inconsistent with the likely behavior of rivals. But it does seem reasonable to assume that as part of the process of generating and refining their expectations, firms will check their projections for aggregate industry capacity against their analysis of competitors’ decisions in light of industry capacity. Anticipating that other firms are engaged in the same process, a single firm should expect industry capacity expansion to be the result of predicted choices about rivals behavior, and it should expect rivals to come to the same or similar conclusions. Consistency then simply summarizes the idea that intelligent rivals will converge in their expectations about each others’ behavior.
The combining of demands, investment pacity decisions by individual firms, and industry capacity expansion into cash wh flows over time involves an economic model of the industry. That model for corn wet milling is laid out in sections 8.3 to 8.8. The model takes capacity, demand, and sugar prices and generates prices, profit margins, ;e that and capacity utilization rates.. The latter are then used to project cash havior flows for various possible investment decisions. The point is that there is a static industry model built in here.
As Professor Winter points out in his insightful remarks in the comment to this paper, our hypothesis is that the evolving oligopoly, deprived of the full complement of markets required to guide investment decisions, calculates the equilibrium in the market and then makes the associated investment decisions. This calculated equilibrium is not the same one that would result with the full set of futures markets, because of the increased uncertainty that is created by the absence of those markets.
One additional aspect of HFCS demand is built into the model, reflecting buyer behavior for this important input to their products. As described aboved, major buyers are unlikely to change over to HFCS unless sufficient capacity in on-stream to serve their needs without making them vulnerable to interruptions and bargaining power by a few suppliers. As a result, it is assumed that the high demand scenario cannot occur unless substantial capacity is added. This interdependence of supply and demand will be further reflected in the capacity/conversion scenarios described below.
Thus, the pattern of capacity expansion from 1976 onward depends upon the demand-growth scenario that actually occurred. Further, as argued earlier, high demand growth was very unlikely when capacity was low because the major buyers of HFCS were reluctant to switch over from sugar until there was sufficient capacity in the HFCS market to meet their needs. Later, in assigning probabilities to demand scenarios, we make the probability of high demand zero when capacity expansion is low.
In formal terms, the reduction of uncertainty transforms the game from one with a single equilibrium to one with many equilibria. Without uncertainty, the optimal strategy for each firm is to expand capacity so as to supply whatever demand is not ocvered the rivals’ known capacity. Thus, there are multiple equilibria depending on whether one firm gets a jump and supplies the whole market or various combinations of firms do. The problem for the individual firm is to ensure that the equilibrium that is achieved is as favorable to it as possible. That entails expanding as fast as possible itself, but ensuring that others do not. This requires trying to preempt by means of public announcements of major capacity and contracting very early for the construction of facilities. If preemption by many firms and fails, then massive overcapacity will result. Thus, with no uncertainty about demand, a new form of risk may emerge to replace. With significant uncertainty about demand, this form of risk is not important because preemption is not a rational strategy under extreme risk-taking behavior. From inspection of figures 8.2, 8.3, and 8.4, it is clear that as uncertainty about demand is reduced, a firm’s risk-return frontiers or options become flatter and flatter, and strategies of preemption (and therefore multiple equilibria) increasingly likely.
In most markets, including the corn wet milling industry, it makes little sense for the firm to create mistakenly high expectations about demand in its rivals. This is because doing so will create excess capacity, hurting the firm as well as its rivals, though its rivals will be hurt more if they build mistakenly. Such an incentive not to overinflate rivals' expectations goes directly against the pressure to paint an optimistic picture of the future to the capital markets.
The second reason for employing scenarios relates to the sequential character of capacity decisions. Capacity decisions are not made on a once-for-all basis at a single point of time by every competitor. There is in fact a sequence of decisions through time, with information from the market about demand, prices, and competitors' behavior pouring in. As a practical matter, this must also be simplified. The approach taken here is to characterize the industry's evolution in terms of scenarios for purposes of predicting the "first round" decisions of competitors. The alternative is to try to deal with the full sequence of decisions, a procedure that also quickly becomes impractically complicated.