Nucor Case Summary
. The Steel Industry
Similar to our discussion of e-commerce markets - albeit in a very different industry - we begin with a discussion of the challenges to capturing value in the steel industry. The fundamental difficulty here is intense rivalry. Above all, this arises from the fact that steel is basically a commodity, with very limited opportunities for product differentiation. While there are different segments of the steel market, within each segment products are nearly identical. This is especially true within the low-end, primarily construction market that Nucor serves. In such a commodity market, the only variable that firms have to attract customers is price, generating the lose-lose price competition which is the defining characteristic of intense rivalry.
This price competition is intensified by several other features of the steel industry. First, demand is flat, or even shrinking, implying that firms must fight with each other to capture market share. Second, the import share has grown to over 20%. And taking advantage of lower costs through cheaper labor, lighter regulations, and government subsidies, these imports have very low prices. This effectively eliminates the possibility that the large U.S firms will be able to collude to keep prices up; such collusion would simply yield more share to the imports. Finally, the large integrated firms have substantial excess capacity - largely because they built plants when demand was greater and imports were a small threat. The investment costs from these large mills are sunk, so these firms are willing to push prices all the way down to variable costs in order to generate positive cash flow.
Added to this internal rivalry is the threat posed by substitute industries - aluminum, plastics, and advanced composites. The growth of these substitutes contributes to the flat demand for steel, and further restricts the ability of steel firms to maintain positive price - cost margins.
Together, these factors imply that steel is a fully competitive industry. In such industries, the only way to generate positive profits is through Ricardian rents. That is, firms with lower costs can steal share with small price cuts - small cuts are generally enough due to the extreme price sensitivity of steel buyers - and thus increase quantity while maintaining positive price - cost margins.
Note that through the 1950s, the steel industry dealt with this rivalry threat by maintaining a collusive oligopoly, with a stable price level set by U.S. Steel. As such, the steel industry basically functioned as a monopoly under the control of U.S steel, thus preventing destructive price competition.
However, the long-term trouble with this cartel was that it limited incentives to innovate. In a competitive market, firms are quick to adopt cost-reducing innovations in order to steal share, as discussed above As we discussed in class, this can be seen as a prisoner's dilemma. Consider two competitive firms who begin with equivalent technology and are splitting the market and making $2 million each in profits. A new technology is introduced which offers a slight decrease in costs, enough to steal the entire market if one firm implements it alone. Implementing the technology costs $1 million. The game grid for this choice can be written as:
Don't Invest
Invest
Don't Invest
(2, 2)
(0, 3)
Invest
(3, 0)
(1, 1)
This is a classic prisoner's dilemma - no matter what the other firm does, I should invest (either to steal share or protect my part), so we end up both investing, splitting the steel market (recall demand is flat, so there's no additional demand after the investment), and making less as a result of the investment cost.
However, under a cartel, the only reason to invest is the slight increase in margins that may be possible with lower costs, which generally won't cover the investment cost. So, part of the collusion can be seen as an arrangement to cooperate and not invest, thus maintaining higher profits for the full industry.
However, the result of this was a lack of modernization by the integrated firms. By the 1970's they were stuck with 40 million tons of open-hearth furnace capacity, when this technology had been surpassed by oxygen furnaces. Hence, the integrated firms had very high costs, creating an opportunity to enter with low costs and thus steal share. But the huge fixed costs required to build a traditional steel-mill made it hard to take advantage of this opportunity. Recall that the integrated mills had already sunk these investment costs. So, the relevant cost for them in current decisions - that is the avoidable cost - was just the production costs. However, a firm considering entering had not yet sunk the investment cost, so this was a relevant, avoidable cost in the entry decision. Hence, if an entrant hoped to steal share with lower costs and thus prices, it needed to have its investment cost per unit + production costs per unit < the incumbent firm's production costs per unit alone, a challenging proposition in an industry with large investment costs. Put differently, to preserve their share, the integrated firms were willing to push prices all the way down to their production costs. Since entrants needed to get under this price to steal share, they needed to have the total fixed and variable costs from the project under this price in order to have positive margins and profits.
Who was able to deal with this? Foreign firms, who had already built their mills and thus sunk their investment costs, and thus could compete on equal footing with the integrated U.S. firms, stealing share by taking advantage of their better technology and lower production costs. So, as discussed above, imports rapidly entered the market in the 1960's, breaking the cartel. After the cartel was broken, the integrated firms entered the investment race, building new oxygen furnaces. However, they were largely playing catch-up and the imports could only supply about 20% of demand, so an opportunity remained for U.S. entrants, if they could find a way to deal with the entry barriers.
2. Minimills
A side effect of the shift to oxygen furnaces and general decline in production among the integrated firms was a decrease in their demand for scrap steel. This increased availability and lower price of scrap steel created the opportunity to form minimills, smaller operations which produce steel, primarily for the low-end of the steel market by melting scrap. So, through the 1960's and 1970's, minimills took advantage of the entry opportunity created by the high production costs of integrated firms.
What are the advantages of minimills? Recall that the key challenge to entering is ...
This is a preview of the whole essay
2. Minimills
A side effect of the shift to oxygen furnaces and general decline in production among the integrated firms was a decrease in their demand for scrap steel. This increased availability and lower price of scrap steel created the opportunity to form minimills, smaller operations which produce steel, primarily for the low-end of the steel market by melting scrap. So, through the 1960's and 1970's, minimills took advantage of the entry opportunity created by the high production costs of integrated firms.
What are the advantages of minimills? Recall that the key challenge to entering is to get investment + production costs below the integrated firms production costs. Minimills effectively reduced both the investment and production costs of making steel. Investment costs were reduced by lowering the minimum efficient scale and thus lowering capital costs per ton from roughly $1000 to roughly $100. These investment costs were reduced both because minimills can efficiently produce at smaller scales - by avoiding the full process of melting iron ore to make steel - and because minimills are designed to last for only 10 years or so. So in both per-year and total terms, minimills reduce economies of scale and thus make entry a lower-cost proposition.
The real key to minimills is that they allow both lower investment costs and lower production costs, rather than requiring a tradeoff between the two. Lower production costs come partially from the use of modern furnaces and casting technology. In addition, minimills tend to focus on narrow product lines, eliminating the cost of re-configuring the production line for different runs. Finally, their small size allows location in areas with cheaper, generally non-union labor, and lower electricity prices. This partially helps to offset the major cost risk that minimills face - higher energy prices or higher scrap prices disproportionately increase their prices and may eliminate their production cost advantage. However, they remain sensitive to high scrap prices, and if those prices spike up too much in certain periods they may be forced to exercise the option of shutting down until prices fall.
3. Nucor's Practices
Through the 1960's and 1970's, many firms took advantage of this technology and entered the steel industry using minimills. Leading the way was Nucor. Between 1972 and 1986 Nucor expanded from 6 to 16 plants, yielding a capacity increase from 0.2 million tons to 2.1 million tons and a production increase from 0.14 million tons to 1.7 million tons. With low costs they were able to maintain positive margins on this quantity, and net earnings grew from $4.7 million to $46.4 million.
However, the story of this success can't just be minimills. Many firms used this technology, and they all served the low-end, commodity portion of the steel industry. So, to grow in this sector while generating profits requires Ricardian rents - lower costs than the other minimills, not just the integrated firms. And Nucor was able to do this by doing the two things they do well, "build steel plants economically and operate them efficiently." These skills allowed Nucor to be the 2nd most productive steel firm in the world, and the most productive in the U.S., producing 981 tons of steel per year per employee.
How did they do it? Through a combination of practices which generated both lower production costs and lower investment costs, the two key variables throughout this discussion. Start with the lower production costs. Perhaps the key to this is Nucor's incentive system. Worker compensation starts well below industry norms, but grows with the output of the worker's production group. These production bonuses amount to 80 - 150% of base wages, so that Nucor's average wages are actually higher than the industry average. So, quite simply, workers are paid more if they produce more, generating more productive workers and thus lower production costs.
Note a couple key features of this system. First, it is based on group performance. Since steel is produced by groups of employees, this is important to generate teamwork and internal monitoring by the group. Note that it might not work as well in industries which depend on the ideas or innovations of individuals. Second, it is tied directly to quantity, the performance measure that these groups can control. Bonuses are paid weekly, with ongoing postings of progress toward weekly goals on the front of each plant. So each individual in the group can see that one bad day will have a direct impact on their pay, and the members of the group have an incentive to monitor one another. This provides an excellent way to deal with the agency costs faced by large organizations. Firms need a way to align the incentives of their employees with the incentive to make profits - here the contribution of each employee is simply to make more steel and that's how they're paid. Contrast this with the standard "incentive" programs like stock options or profit sharing. While those pay people based on profits, individuals have effectively no control over these measures, so such programs have little impact on the performance of any individual.
Note that this fits with the decentralized organization at Nucor. Groups need to be small enough that individuals see their direct contribution to output. So, Nucor uses production groups of 20-40 individuals. This is consistent with the use of minimills, where production lines can be controlled by much smaller groups of employees. But Nucor pushes it further, with a very flat organizational structure consisting of only 5 layers of management, with most decisions made at the plant level. This limited bureaucracy also serves to directly reduce Nucor's overhead costs, further bolstering their low-cost position.
Note that this incentive program will not work in all situations, potentially limiting its scalability, but also helping to protect it from imitation. First, it works well with an easy to measure performance standard like quantity. And this is the right performance measure in Nucor's low-end segment of the steel market, in which steel is a commodity, cost is everything, and the way to lower labor costs is to get more production per worker. However, anywhere where quality matters and firms seek B advantages, including the high-end of the steel market, this will be harder to implement. While it may be possible to measure the quality of the steel produced or the level of customer service provided, this would be much more costly than counting tons limiting the competitive advantage that can be gained by implementing such a program. And as performance measures become more complex, the link to individual effort becomes more tenuous, reducing the incentive effects of the program.
Further, this program is clearly only possible with non-union employees. Unions generally place strict restrictions on the form of compensation, and this type of low-base wage/performance-based structure is something they consistently fight. Finally, such a goal-oriented system is not for everyone. Nucor overcomes this largely by screening employees based on psychological tests, rather than steel-making experience. This only works in conjunction with Nucor's excellent training within the production groups, and might be impossible in more complex, larger, integrated mills.
In addition to these incentive programs, Nucor has a set of practices which lead to greater efficiency and thus lower production costs. First, within the groups, employees are continuously trained by more senior employees, in the process learning multiple functions in the firm. This leads to more productive employees and allows full capacity utilization at all times, as workers can shift to the part of the process requiring the most attention. Note that this fits with the incentive program, since bonuses are paid on group performance, members of the group have an incentive to train one another. Second, Nucor ultimately hires the people who build its plants to work in them. This may create a certain "pride in ownership," and certainly increases workers' knowledge of the plant's function. Again, this probably works best for small minimills. Finally, Nucor is able to reduce employee turnover - both by keeping workers happy so they don't quit, and by resisting layoffs. Newer workers are less productive and require high training costs, so this further reduces production costs. And this is where the more indirect incentive programs come in - profit sharing, stock-options, college allowance for employees' children, etc., probably do not directly lead to more production, but they make employees happy and thus reduce turnover. And Nucor's flat salary structure, and "share the pain" programs during downturns in the steel market certainly help as well.
The second key component of Nucor's strategy is lower investment/construction costs. Nucor has extensive experience building and remodeling minimills, and thus organizes its own construction in optimal ways that other firms or external contractors would have a hard time mimicking. A key component of this is that Nucor designs and builds plants concurrently, reducing engineering costs to 2% of total costs, from the industry standard of 10%. This relies on their experience, as they are able to make adjustments "on the fly," and the fact that they are building smaller, simpler minimills. This method also reduces the time to project completion, which allows lower payments to all external workers and contractors and allows positive cash-flows more quickly.
How did Nucor come up with this set of practices? Put simply, experience. Nucor has extensive experience building and operating steel plants, so they have moved down a learning curve. And this learning curve, combined with the strategic fit of all these elements, prevents other firms from copying these strategies. Absent the learning and fit, these practices would simply represent operational effectiveness. But the strategic fit requires a group of decisions and tradeoffs, so that Nucor is really operating in a unique position in the market. And the learning curve indicates that, rather than simply moving out to the productivity frontier as firms do with operational effectiveness, Nucor is constantly shifting the frontier out in a way that other firms can't keep up with.
As discussed in class, we can depict the overall learning process as follows (make sure you understand all the connections...)
Lots of Investment and Growth
Experience Operating Plants
Employee Incentives Investment Practices
And Practices.
Decreased Production Costs Decreased Investment Costs
Increased NPV
of Investments
More Investment
and Growth
So, this is a classic learning curve advantage. More experience generates lower costs, lower costs allows more investment and growth, that generates more experience, etc. The advantage builds on itself, and thus is very difficult to copy.
The final question that needs to be asked about a learning curve advantage is how the firm keeps the learning internal and spreads it to all parts of the firm. That is, while individuals can simply remember their experience as they learn, firms need to find ways to institutionalize this learning process. Nucor does this in several ways. First, they have formal channels - monthly operating reports from each plant that are shared with all plants, and formal corporate meetings with plant managers and functional groups. Second, many of the practices discussed above also foster learning. Most importantly, lower turnover keeps the information inside the organization and allows time for senior employees to train more junior employees in the learned practices. And by running its own construction projects, Nucor keeps the learning about building projects in house, particularly by hiring the builders to be permanent employees. Finally, the incentive structure contributes to learning. Each production group has an incentive to seek out information about how to operate productively. And Nucor posts lists of the most productive groups and allows employees to visit the plants where those groups work. Because production standards for bonuses are paid relative to market average production figures, groups do not compete with one another and are therefore willing to share information and learning.
4. The Investment Decision
In 1986, Nucor finds growth opportunities in its traditional segment of the steel market declining. Note that this is more of a problem for Nucor than many firms, as Nucor's strategic advantage comes from continuing to learn through growth and investment. As these growth opportunities slow, so does Nucor's traditional productivity growth. The number of employee hours per ton of steel fell from 3.9 to 2.9 from 1978 - 1982, but only from 2.9 to 2.4 in the next four years, from 1982 - 1986.
Hence, Nucor is considering a move into the flat-sheet segment, which accounts for more than half of overall steel demand. Recent technological developments have made this possible for minimills. Nucor is focusing on the possibility of implementing the compact strip production (CSP) process developed by SMS, an unproven but potentially efficient way of producing 2" thin-slabs of steel which can then be fairly easily pressed into flat sheets. The possibility of making this investment should be viewed as an investment opportunity for Nucor. Where does the opportunity come from? From its operating and building practices, combined with its ability to learn from new investments, both of which suggest that this investment is likely to be worth more (have higher NPV) for Nucor than other firms. So, there are ex-ante limits to competition, Nucor's competitive advantage in essence gives it a first-shot at this technology, and the fundamental decision is whether to exercise the option or wait for more information.
The spreadsheet used in your homework suggests that, under reasonable assumptions, the return from this investment is probably marginally higher than the opportunity cost. Note that while 7.06% may sound low, this is the after-tax, and after-interest return, so the opportunity cost needs to be computed in these terms. A 12% gross return is reduced to 5% after deducting these interest and tax percentages, so the 7.06% return sounds reasonably good on pure NPV terms.
However, experimenting with the assumptions used to arrive at this IRR suggests that the return is subject to a great deal of risk. And we know that, given this risk, the NPV of the project should be compared to the option value of waiting, which is increased by the amount of risk which will be resolved by waiting, and decreased by any costs to waiting, such as lost first-mover advantages or lost market opportunities. So, we now turn to the risks and costs to waiting.
The key risks to this investment project can be broken into three basic categories. First, it is not clear how well CSP will actually work, nor how high the costs will be, for a full-scale implementation. And changes in investment or production costs could greatly impact the actual return from the project, easily pushing it negative. So, there may be value to waiting to gain more information about the actual efficiency and costs of the method. However, a couple issues weaken this argument. First, SMS seems to be done with its internal testing of the technology and is now ready to commercialize. So, if Nucor wants more information, it will probably have to allow another firm to enter first. This could potentially be a major cost to waiting, as discussed more fully below. Second, the discussion of Nucor's practices indicates that it is distinct from other firms in many ways, generating higher NPV's from most projects than its competitors. So, it is not clear how much Nucor will learn from the experience of other firms.
Second, there is the risk that the technology will be leapfrogged by something better. This is probably the most important risk and the strongest argument to wait. Recall that sales of steel are all about low costs and prices. So, if another lower cost technology is developed soon after this investment, Nucor will be forced to drop prices to compete, and the NPV of the project could easily turn negative. However, most of the potential technologies are focused at direct casting of steel into flat-sheets, and these are expected to be 10-15 years off. So, they are really after the life of the CSP project and thus not impacted by this decision. The one exception seems to be the technology developed by Mannesman-Demag, which would produce 1" slabs, rather than CSP's 2" slabs. The best reaction to this may be to compare the best estimates of the NPV's of these two methods today, and then go ahead with the better one. I believe Nucor has done that and selected CSP. But if more information is likely soon, this provides the best argument for high option value to waiting in order to improve this choice.
Finally, most of the other risks revolve around fluctuations in the demand for steel, steel prices, scrap prices, energy prices, etc. While these are all real risks, they do not really provide an option value to waiting. If Nucor waits one year, they will face a new starting level for these variables, but the variables will still fluctuate from that point forward. So, this uncertainty is not resolved by waiting and thus does not provide a reason to wait.
The option value to waiting - primarily from learning more about alternate technologies - needs to be weighed against the price of the option, that is the cost to waiting. The traditional first-mover advantages here are somewhat limited and short in duration. However, there are two other arguments that moving today may be very important. The first is the risk of entry. To see this, begin by noting that demand for flat-sheet steel is stagnant (or should I say flat?), so that new entrants are likely to have to steal share. And it seems most likely that minimills will primarily serve the low-end of this market, the construction sector. The total demand in this sector is only 2.5 million tons. The first entrant will face challenge enough to steal 1 million tons from the integrated firms who will aggressively defend their share by pushing prices down to their production costs, as discussed above. So there is a real chance that there's not room for a second firm. Thus, a classic economy of scale argument suggests that, since there may be room for only 1 firm, the market may well go to the first firm. In that case, the cost to waiting would be losing any chance to enter, while the benefit to moving today would be deterring future entry.
The second argument is similar in spirit. Exhibit 12 B of the case indicates that the production cost advantage of CSP is much larger against un-modernized, integrated mills than against modernized mills. Recall that a new entrant needs to have its investment costs + its production costs < the integrated firms' production costs to make profits. The small production cost difference vs. modernized mills suggests that this equation will not be satisfied against those firms. And U.S. steel, Bethlehem steel, and others are beginning modernization programs. So, if Nucor hopes to steal share, the time is now! Further, if Nucor can get in first, then when the incumbent firms think about investing to modernize, they will have to compare production + investment costs to Nucor's production costs, and Nucor will win this comparison. So, Nucor may actually be able to deter the modernization, if they get in today.
Still, there remains the risk that Nucor will jump in first to deter entry and modernization, but then end up losing money if the CSP technology is leapfrogged. So, there remains an argument to wait. However, one more factor needs to be considered. Recall that the NPV of the project should be increased by the value of any options created by the project. And recall that Nucor's core resource is really their ability to learn from investments and internalize that learning to improve future performance. But they have little experience in building plants this large, competing in flat-slab, using this type of technology, etc. So, this operation would contribute extensively to exactly the sort of learning that Nucor is so adept at utilizing. This would create the option of bringing more CSP plants online at lower costs and higher returns, if this opportunity presents itself. This may further reduce entry risks: if the thin-slab market grows enough for multiple plants, Nucor's experience may enable them to get there faster and cheaper than potential entrants. And if new technologies appear, as expected, around the turn of the century, Nucor may be in a much better position to evaluate and implement those technologies due to the learning from this project. And even if this project has to end early, due to technological leapfrogging sooner than expected, Nucor's ability to apply their learning and experience to a new investment should be seen as the terminal value of this project, which would be missing if Nucor chose to simply wait for new technology to arrive.