Most external dependencies are self-evident to senior management in any capably run business. Amazingly, very few of these dependencies have been formally evaluated as to their relative to revenues, costs and earnings. Our first step in the analytical process is to identify and rank the external dependencies. We’ll have the opportunity to change the prioritization of these dependencies later (and over time), as long as we make certain we’ve at least captured the primary dependencies here.
While literally endless layers of influencing dependencies could be identified, we especially want to determine a top ten to fifteen. The principles of Pareto analysis tell us that once these influencers are ranked properly, the relative impact of each factor is substantially diminished the further we go down the list. As we’ll discuss later, another part of our practical focus on strategy is to realize that while there may indeed be a 121st most important external dependency and a 168th most important external dependency as well, we need to focus in order to get meaningful results. In this case, that means simply narrowing down to a top 10-15 factors. Later in the process, we will concentrate corrective (meaning, adaptive) effort on the topmost of those 10-15, get them [properly addressed, and then move on down the list.
Modeling primary financial impacts
Our next step is to create models of each of the primary factors. While several firms, including ours, have proprietary means of assisting clients in developing these financial models, the essential element is to be able to estimate cause and effect. We need to assemble an understanding of the business, based on the pre-existing financial models currently in use, and past history along with forecasting input from key managers of each area in question.
In what direction, and to what extent, do we expect that a 10% increase in consumer inflation will affect revenues? To what extent would we expect profit margins to be impacted due to a 23% increase over time in transportation costs?
Notice that we are no longer asking functional managers about the likelihood of a given external condition. What we are quantifying in this step is the relationship between the external variable and the internal results. Already in this step, we’ve taken the process beyond the context of asking individual managers to forecast the future, which was never ultimately the process we were after in the first place. Instead, we’re asking that the manager understand and be able to quantify the relationship between things affecting input costs and the ultimate financial results from their area in the context of overall corporate performance. In our experience, this is a legitimate point of knowledge for most managers that we should expect them to be able to address – if not, we might ask if they are really in command of the basics necessary to manage their area of responsibility.
While it should be acknowledged that there are several very sophisticated organizations that have done a substantial portion of this type of analysis for themselves (public utilities, insurance companies, among others), the reality is most business enterprises, even those with very large external dependencies, have not. Not too many years ago I had been meeting to review an acquisition with two Senior VPs of a top-3 US airline. After wrapping up and on the way to lunch, I asked them about their fuel hedging program. They responded that they really did not engage in fuel hedging (they still don’t to this day), and that “fuel prices are nearly unpredictable”.
As we know, one key competitive advantage on the part of Southwest in particular has been their fuel hedging operation, particularly as volatility in global oil markets spiked in early 2008. With billions at stake across the airline industry in annual fuel costs, we can start to see that many industries where it might have been assumed external dependencies had been well identified and management processes put in place years ago to address them, the reality is that, just as in this example, there are gaps everywhere. Setting aside the fact that we may or may not have the sophistication to put some of the “risk-mitigation” strategies in place, the practical reality is that there are many options for addressing the situation, once we have first identified its impact on the business.
While some of this behavior may fall into the category of “corporate denial”, it has some similarity to the individual experiencing pain who wants not to visit the doctor and have tests done for fear of what they may learn. We need to honor shareholders and other stakeholders who depend on the business for results. We’ve all heard the old adage that “failing to plan is planning to fail”. Yet in the final analysis, if we have not identified the relationships between at minimum the top dozen or so external input factors and our business results at some quantifiable level, then it’s effectively as if we’ve said that their impact is zero.
This is true due to the absence of actionable information in the absence of the analysis. Therefore, unless we know that zero is the actual impact resulting from the relationship (meaning, no relationship), then we know we are operating on a false premise. And if we are to be honest with ourselves, even the assertion that the accurate relationship is zero, would itself have to be based on analysis in order to support that conclusion.
Up until the moment we are in possession of a credible analysis of the dependencies, we are implying that there is no relationship between the top external factors and our ability to produce predictable business results across the enterprise. This highlights the urgency of completing this seemingly academic exercise, and at the same time serves to explain why and perhaps how so many companies in so many industries have been brought up short and suddenly found themselves in literally unrecoverable trouble.
Assess probabilities
The next step is to assess the relative probabilities of individual variables reaching forecasted levels. Of course, we believe the forecast represents the most likely scenario or it would not be the forecast. Yet, as a practical matter we need to attach a probability since some forecasts are simply “stronger” than others. For instance, we have a greater probability in the fed funds rate, and therefore the cost of capital, being at a level consistent with what treasury futures would predict six months out, as opposed to the probability that fuel prices would be at nearly any level that we might predict a year from now. This is also why we look to establish ranges as described earlier.
As with other data, we need to test our estimated probabilities against external, independent sources. Thankfully, we have not only very competent and objective analysts available in most of these variables, but in many of them we also have markets. The predictive capability of markets has been demonstrated time and again to be highly accurate, although certainly not infallible. But market indicators, in league with analyst-developed probabilities, can develop a much clearer picture, one which gives us at least our starting place.
Along with other parts of the process, keep in mind that the mechanisms used to establish probabilities can themselves be adjusted over time. As we see unfavorable and perhaps repeated surprise factors develop around certain probability forecasts, we can re-evaluate our sources and the evaluation methods we use internally to develop our probabilities. Where did the surprise originate, and how is our data gathered in such a way as to not capture the thing that created the surprise? The ability to track our projected probabilities over time against the actual outcomes, will give us an ever greater ability to refine our methods and gain greater accuracy over time.
Rank each exposure: Probabilities/Impact = Exposure
Our next step is to perform a simple two-dimensional ranking of exposure. For our purposes her, we are defining exposure as:
P/I = E
Where
P = Probability of a forecast condition, stated as a percentage
I = Impact in earnings terms as variance from current earnings (EBITDA) if the condition materializes
And
E = Exposure, ranging from 1 to 100
From this calculation we prepare a graphic analysis, plotting each of the conditions in relative terms. This allows us a much greater understanding of both the conceptual exposure to external dependencies, as well as a truly quantitative picture. Much like the pilot of a large commercial aircraft, we now have instruments that can measure the effects of the external conditions on our business that correlate with wind direction, barometric pressure, temperature, humidity, and allow us to start to understand how they affect the results we can produce.
The resulting chart will resemble this format:
% | * *
Probability |* * *
| * *
|_ *___*______________
$ Impact
At this point there may be certain revelations in terms of how we look at the business. There also may be a tendency to call the results into question. While our process to arrive at the analysis can usually, and should, stand some refinement over time as pointed out earlier, it is important here to follow the process to completion, particularly in the first pass, then go back to make further refinements. Recall that we’re out to formulate a strategy that we can own, and a practical strategy is of greatest value when we do not “let the perfect be the enemy of the good”.
The next step is the identification of options and selection of specific strategies. Now that we have identified the relative effects on the business, we need to determine the optional strategies available to us, their cost and time required to implement each potential strategy, and select from among them the steps most practically suited to the business and its capital constraints.
