If you want to start to read the beginning of the series, click here for part one.
2.0 The Tail Wags The Dog – How Tactical Decisions Run Company Strategy
2.1 Competitiveness over a lifecycle
An industry always improves, in the sense that it is always better at what it is doing today than it was 10 years ago, because of technology development (the industry doesn’t, however, make more money due to the real price decline). We illustrate this point with the red line in Figure 2-1.
At some point, a company will build a new mill, either a Brownfield or a Greenfield. Or it will add a new production line in an existing mill. We illustrate that with the blue line starting at point A, where the mill is new and state of the art. It is clearly above the red line. Over time, the mill will be improved (the blue line goes up) but it can never follow industry average. At some point, it will cut across the red line and become less competitive than the industry average. Decades later, the company will come to the conclusion that it is better to shut down the mill because the gap between it and the industry average has become too large. This is a simple illustration of a life cycle of a mill.
A mill will slowly travel along the blue line to the right. Most mill investments will, at best, slow down the loss in relative competitiveness, not improve it. Only a few investments will take it to the left (improve mill competitiveness), and then in almost all cases it will only be a small shift. For example, a company makes a really large replacement capex in a mill. It replaces 10% of the mill with state of the art technology (like a so-called recovery boiler in a pulp mill). The time from the decision to invest in the mill to turn key (completion of investment) is often 1-3 years. Replacing 10% of the asset base shifts the mill to the left, but 90% of the mill ages another 1-3 years in the process. The net effect for the mill from the point of decision to turn key is not very large.
We will use Figure 2-2 (p.17) for our discussion about allocation of capital. Let’s say we run a business with 10 mills. They are all making more or less the same type of product aiming at a certain market. We assume that our port-folio of 10 mills, on average, is exactly on the industry average (i.e. we have half of our capacity in mills that perform better than the industry average, and half of our capacity in mills that perform below industry average).
Therefore, it is not unlikely that we would have three mills (making up 50% of our capacity) that are the more competitive mills, and seven that are less competitive. Let’s categorize the 10 mills. We have our Category C mill, the mill we understand we will close in 1-3 years since it isn’t performing well at all. We have our three close to state of the art Category A mills. Finally, we have our six Category B mills, good providers of our product mix, etc., but to different degrees less competitive.
Question #1: “In a normal year, how would your company invest in those 10 assets?” Most companies do not have relevant information to make the categorization, but let’s give this a try. We exclude possible investments in Brownfields/Greenfields, we just look at those 10 assets.
If we begin with Category C, it contains the mill that we know we will not keep for very long. We would prefer to spend 0% of capexes on this mill, but we need to spend some capital in that mill for it to stay open another quarter, another year. Most companies spend 3-5% of their capital on this category of mills.
How about Category A?
That category is relatively close to state of the art technology. They “need” less capital. Additionally, the opportunities for this category are perceived to be not that many (since they already are relatively good). They have less environmental issues, fewer safety issues, etc. On the other hand, they represent 50% of the capacity – clearly the need to invest here exists. Most companies would invest 20%, sometimes up to 25% in this category.
Finally, we have Category B.
Six mills (twice as many as in Category A) with an increasing technology gap (aging assets). Often these are relatively complex mills – two or three of everything instead of one – and not always a “linear” layout. There are usually environmental and safety issues. Category B gets 75% of the capital.
When considering the answer to Question 3, a company must include all the discussions/analyses/etc. on, “Should we close or keep, should we make another investment?” In considering this question, include the time spent trying to fix the mill, evaluating closure costs, and eventually when the mill closes. We had a client once who hesitated over a Category C mill for a long time. Once they even had one of their senior VPs on the corporate jet going to the mill to communicate its closure. When he landed, he got the message, “We got another large order, come back home.” This can go on for years, which is why we see such a high percentage of resources allocated to Category C mills.
Category A doesn’t need resources.
Those mills are working just fine. In a project, we had a number of years ago, the project manager from the client’s side had been a mill manager for their best Category A mill for five years. He told us that the head office hadn’t contacted him even once during those years.
Category B gets 65% of the resources.
When we read the pink pages we often read about CEOs from the capital-intensive industry explaining the quarter’s poor performance with reasons such as prices not developing the way they expected, increased costs, and shaky demand. Certainly, one can explain a quarter or two with those factors, but the capital-intensive industry often fails in delivering shareholder value year in, year out. To blame the above factors for more than a couple of quarters is like saying that one lost a sailing race due to a head wind.
Prices not developing as expected, costs being up and demand being shaky is nature. It is a fact for all players in a capital-intensive industry. Technology development, even if slow, brings down prices compared to costs for existing assets – this is a fact. Companies invest in better technology in an attempt to avoid that.
Now, we cannot say that this table, for a certain company, in a certain year is “incorrect.” All we can say is that if one spends resources like this over time; one will weaken and eventually ruin the company.
We would say that all companies and all CEOs we have worked with for the last 23 years have had the ambition to do well. They want to do the right thing, to have the company make more money, but too many organizations fail.
There are several reasons for why many companies don’t deliver expected returns or fail. Why over time they don’t follow the stock market’s index, as an example. Here we show you all you need to drain a company of cash flow, and even sometimes, as shocking as it sounds, run them into the ground: All that is needed is to have the company’s future somewhere (green), but spend its resources elsewhere (red). Done year in and year out, decade after decade, results in a poor performing company. Or a company in Chapter 11.
We were engaged in answering this question in the late 90s. We came up with a long list of reasons but pared the list down to three basic reasons why companies fail.
The first two reasons are present in all companies, and they have to be there. They cause an extremely costly problem, but they are necessary.
The third reason is anything from somewhat present to very present in almost all companies. Although it doesn’t have to be present at all, companies can throw it out today – there is no point whatsoever in having it involved in the capital allocation process.
Reason #1: The capex process is a bottom-up process. It has to be. The mill knows the issues and how to fix them (supported by central technology/manufacturing resources). So, the mill manager will come to the head office and say “I have an issue, and now I know how to solve it.” This issue could be quality, cost, safety, environmental, etc. The head office will say, “We know. We have been discussing this now for two years.” On paper, the capex necessary to fix the issue makes sense.
Reason #2: Delta calculations are used to evaluate the benefit of a capex project, and the projects are evaluated one by one, in isolation. Going back to what we discussed in part one, basically all companies do what the textbook says, what we have been taught to do: they calculate the payback/NPV/IRR for each capex project individually, in “isolation” (we assume all data is accurate and correct).
So, a mill manager comes to the head office saying, “I have an issue, and now I know how to solve it. The payback is two years.” The head office is likely to reply, “We know. We have been discussing this for three years. The payback makes sense; it even looks a bit conservative.” They respond this way because an individual analysis of the issue tells the company the manager’s solution is an appropriate and reasonable way to proceed.
Reason #3: ROCE (or ROOC, Re, EVA, or any other P&L and balance sheet based measure) is used to prove the performance of the mill. It is bad enough that companies use these measures at the group level. Some claim that the errors in these measures are canceled out when applied on the company/group level. This is incorrect, and to use any of these measures on a mill level gives totally corrupt information (it doesn’t matter how many adjustments you make in, for instance, EVA). Let us give you a relevant example:
Take the blue line from Figure 2-5. It represents the life cycle of a mill, let’s say 60 years. We lay it out as the X-axis in a graph, with the categorizations A, B and C, Figure 2-6.
It would be very volatile, but for the purpose of our example, we can “normalize” it. Look at Figure 2-7 whose EBITDA margin development represents a mill from its Greenfield state to closure 60 years later.
In reality, the EBITDA margin will jump up and down. One can work in a mill for 20 years without noticing the EBITDA margin trending downwards.
The mill carries a lot of initial capital in the first 15 years (accounting rules), but cash flow from year five is superior to any other year after that. The EBITDA margin is at its peak. Approximately at the point when it goes from being competitive to less competitive than the industry average, the mill’s accounting based performance measures will skyrocket.
**If you are ever to take the seat as a mill manager, make sure you do this a year before the initial capex has been written off in the books. Then make sure your bonus is set on ROCE or EVA; you’ll be rich. Cash flow and EBITDA margin will be down, but you’ll be just fine.
So, with “Reason #3” a mill manager comes to the head office saying: “I have an issue, and now we know how to solve it. The payback is two years. And, we still contribute well to the company’s ROCE”. Why would anyone ever say no to a capex with those attributes? No red flags anywhere in sight.
ROCE does not, in any way, indicate competitiveness. Hence it should never be involved in any capital allocation discussions.
Any business wants to think that it sets a system-wide strategy for its assets and that its capexes follow that strategy, right? Like this:
We claim that the capex process is so strong that it actually sets the strategy for the assets, in our example, the mills. The result of the selected capexes for a mill year in and year out determines the strategy of that mill and as a result, the whole company, (however, it does not determine the fate of a mill. Even stronger external forces determine that).
So, it is not that an asset strategy determines which capexes are chosen for a mill. Companies don’t even have well-established, uniform and thorough processes for setting the asset strategy for their system of mills, (the reader may think “but we do in our company”, especially if said reader is responsible for that process – we’ll challenge that statement any day).
So, the picture actually looks like this:
All companies that invest capital in fixed assets (even if they are not necessarily capital intensive) need to fix this issue. All companies have to have an Asset Strategy that governs the capex allocation.
But the way to achieve it is not at all the way one might think it should be done.