Total shareholder return (TSR) is a beguiling metric. Return on invested capital (ROIC), return on equity (ROE), growth rates, risk adjustments, and so forth rise and fall and interact in the various ways that make financial analysis intriguing, but at the end of the day, the number that tells how well a company has done for its shareholders is TSR. Because TSR represents the ultimate bottom line of financial performance, it can’t help but command the attention of board and senior management.
On the other hand, TSR frustrates management by seeming almost random in its operation. Management may believe it is doing everything right, setting ambitious plans, executing those plans and meeting its forecasts, yet find that its TSR is no better than average. Meanwhile, the highest TSR in its peer group is likely to be posted by a company that was in trouble no more than a year ago. It all seems unfair.
Management can be forgiven for concluding that TSR is unpredictable, owing more to the latest investment fad than to any rational assessment of company performance.
TSR is a paradox among financial metrics—dominant in assessments of past performance yet peripheral in plans for future performance. This paradox can be resolved. Just as TSR serves as the ultimate arbiter of past performance, it can be harnessed as the prime touchstone of planning for future performance.
Let’s start with what is obviously good about TSR. First, as metrics go it’s admirably comprehensive. It tells the investor how much his or her investment value has grown. Value growth, of course, is not the only strategic aspiration utilities pursue. Reliability, reputation, strong customer service, community citizenship, and safety are the foundational elements of any utility’s mission. But to the extent that a utility also wishes to create shareholder value, TSR cuts to the chase.
Second, it’s simple. Find out how much the price of shares has risen during a given period, add the dividends received during that time, divide by the original investment, and you have the total return. If no dividends have been received, TSR reflects simply the change in share price. If no change in share price has occurred, TSR reflects simply the payout of dividends. Because of its simplicity, it can be researched and compared to any set of peers through a brief online visit to virtually any financial reporting service.
However, TSR also has its limits, which are sometimes not fully appreciated. It tells us how well an investment has performed over a given period— not necessarily the same thing as telling us how well management has performed. TSR is the metric that never stops asking, “What have you done for me lately?”
Posting a high TSR is like getting the academic award for most improved. A student who goes from D to B beats the student who goes from B+ to A-, or the student who consistently gets A’s. When a board gives management a target of being a first-quartile TSR performer each year, it is asking for something that is by definition extremely difficult, if not impossible, to deliver. It’s asking management year after year to rank among the most improved. As Table 1 indicates, no utility over the past five years has consistently accomplished that feat.
The reason is apparent. When a company’s good fortune and superior management has endowed it with strong growth prospects, that expectation of strong performance is reflected in its share price. To elevate that share price further, management must offer the market new expectations of even better performance. In time, any management runs out of opportunity or ideas for that next increment of improvement. Its rate of improvement subsides to the mean. It still may be a very well run company, rendering strong and reliable earnings, but its reputation for strong performance is now baked into its share price.
There’s a related problem with TSR as a planning tool. It reflects market dynamics, and management is perennially suspicious of market dynamics. Investor preferences seem to change with the seasons. Today’s favored strategy is tomorrow’s pariah. As far as TSR is concerned, it seems all that sensible executives can do is manage the fundamentals and then cross their fingers.
A good planning metric provides specific guidance in two ways: in making choices for future action and in evaluating current management performance. In light of the drawbacks just noted, how can TSR serve either to guide choices or to evaluate performance?
To use TSR as a guide to management choice, we need to recognize, or at least pretend to recognize, that the market’s seemingly fickle behavior is not irrational. To be sure, from day to day the stock market reacts to a varied assortment of events, rumors, enthusiasms, barely perceived trends, and the like. But the market is trying to sort out three fundamental assessments:
(1) How much is this company earning?
(2) What is the expected growth rate of those earnings? and
(3) What is the risk associated with that expectation?
If an investor accurately can appraise those three elements, she will know what a stock is worth. Current earnings are (usually) easy to determine. Growth and risk are tougher—indeed, they’re inherently speculative—and it’s primarily the elusiveness of these two elements that leads the market to bounce around in response to various signals as it tries to sort out the telling indicators from the static.
So let’s stipulate for now that the market is rational but inquisitive for new information, that over any short-term period it may under-react or over-react in unpredictable ways to new information, but that over the long run it will register a reasonably accurate appraisal of earnings, growth, and risk (see “Doing the Stock Market Math,” P. 47).
On this assumption, which, true or not, is axiomatic to any valuation discussion, the market will price stock at the present value of the company’s future earnings stream, as best it can figure that out.
The next step is to translate the generic elements the market uses—earnings, growth, risk—into more specific tools that management can employ in improving that present value. Given any particular earnings stream, the generic aim is to increase expected growth or to reduce risk or to do both, but that generic aim needs to be decomposed to become actionable.
For this purpose, the concept of economic value is useful. Economic value, of course, is simply the value created by an activity over and above the capital cost employed by that activity. This is the metric that is invoked in the application of economic value added (EVA) techniques and the like.1 The analytic steps are straightforward.
• Decompose the overall enterprise into natural P&L portfolio elements. These are likely to be differentiated on the basis of core business activity, risk profile (and therefore cost of capital), and perhaps financial structure. Treat corporate or shared overhead functions as a separate unallocated cost center so as to strip out any distortions that might arise from allocation.
• On the basis of current plans, project cash flow out for five years for each entity.
• Determine the weighted average cost of capital (WACC) appropriate to that particular differentiated cash flow. Most corporate finance or planning departments have their favored methods for determining this number, most such methods are defensible, and your company’s financial assumptions ought to be internally consistent, so just do what the Finance Department says on this one.
• Discount future cash flows at the relevant WACC to determine present value, using some reasonable method (back to the Finance department) for calculating a terminal value at the end of the 5-year forecast period. This is the value of the portfolio element. If capital assets are associated with that element, you can subtract the book equity value of those assets to determine how much economic value is being created (or wasted). Add the individual values to determine enterprise value.
• Divide enterprise value by the number of outstanding shares of common stock.
These steps provide a simple but illuminating profile of the company’s value position, the kind of profile every board of directors and every management team should have at its fingertips. The illumination comes from distinguishing the capital cost and growth plans of constituent element, understanding where value is being created and lost (in most companies value is being lost somewhere), and how the pieces fit together to create overall enterprise value.
Bear in mind that this display offers a picture of the value of the enterprise today given the plan you have developed for future performance. Some of these plans will doubtless unfold in future years, and today the market may know nothing of them, or may have only a general directional impression of them.
One of the interesting inquiries spawned by this kind of picture is to try reconciling the value implied in your plan to the value currently registered by the market. If they’re close, then your plans and the market’s perceptions likely are in alignment. If the market value is significantly lower, it may be that the market doesn’t know of your plans, or doesn’t believe your projections, or possibly is assigning a higher risk to your activities than you assign. If the market value is significantly higher, then the share price may be due for a downward correction unless your plans can be improved to fill the gap.
For future years in your planning horizon, increase the calculated share value by your enterprise cost of capital, compounding annually. The price paid today embeds an expectation that the return on investment will equal the cost of equity capital.2 This is an important assumption in application of TSR as a management tool, because it means today’s calculated value will increase in step with the equity cost of capital. Over the relevant planning period, projected growth in calculated share value (assuming for the moment no dilution through dividend payouts) will be equal to the cost of capital.
Compare the calculated share price at the end of the planning period to the current actual share price (i.e., what people are paying for your stock today without perhaps fully appreciating your plans). The difference is the growth in TSR that your current plan provides. Note that the putative share value you have calculated at the end of the period is not necessarily a prediction of share price, since some of that value will likely have been paid out in dividends. Regardless of how much or how little value growth is captured in dividends, however, the TSR calculation remains the same.
By putting together some fairly basic calculations in this way management can see that a particular level of TSR is implied in its plans. That level can be raised or lowered by changing those plans. Accordingly, TSR can serve as a discretionary goal that frames corporate planning.
Since industry mean TSR over time will approximate the industry cost-of-equity capital, a company that wants to be among the industry leaders in value growth probably will wish to set a TSR goal in excess of the cost of capital. If its value profile does not provide that kind of growth, the company should consider adjusting its portfolio or burrowing into the constituent entities of that portfolio to see how their individual economic contributions can be improved.
So let’s put in perspective what we have. It would be foolish to contend that the result of these calculations is a firm prediction of share price years from now. Basic value drivers change in ways that management cannot control and probably cannot forecast. Interest rates and inflation rise or fall, the general economy prospers or languishes, the market goes through bear and bull phases, and so on.
These systemic phenomena doubtless will have an effect on the dollar value of a share of stock five years from now that cannot be calculated today. As with any planning tool, what the calculation described here provides in your projection today, based on the best information available today. Part of the planning process, of course, is to keep that projection updated with some regularity and to make whatever corresponding adjustments are called for by your TSR aspirations.3
Another point is that one cannot predict exactly when the market is going to swing into alignment with the calculated value perspective. Some of the plans that underpin that valuation may take time to roll out. The market might not be convinced that a plan will work until it starts to see results.
And, of course, some plans simply may not work. What the series of calculations described above tells us is the TSR that is embedded in the plan. Whether that TSR matches the TSR that ultimately occurs is just as dependent as any other forecast on the quality of the plan and its execution.
The question then becomes, how do we manage to this long-term goal? How can we translate TSR into an actionable basis for performance management?
The answer is, indirectly. TSR should not be a direct management measure. For reasons discussed earlier, TSR in the short run is highly misleading. Even the best-run companies will place in the first quartile among peers in some years and second or even third in other years. (And some of the worst-run companies may find themselves in the top quartile in a given year owning to a rebound from some slip-up.)
Over time, superior companies doubtless will be among the top TSR performers, but from year to year their quartile rankings will, indeed must, shift. Instead of a yearly TSR goal then, we suggest a 5-year TSR goal, based on achieving a long-term targeted TSR level. That level may be set with the intention of placing the company over that extended period in the top quartile of shareholder return, but management should not be evaluated on whether the target is reached during any shorter segment of that period.
A long-term TSR target is meant to encourage management to think in the long term, and to focus on the issues it is able actually to manage rather than casting a nervous eye continually over its shoulder at how the market is reacting on any given day, quarter, or even year.
So TSR should be a long-run goal. The metrics of performance management, on the other hand, must operate in the short run. A company can’t wait five years to take a reading on how well management is performing. In this respect, TSR works powerfully, but indirectly.
Recall from the TSR projection illustrated above that in the sequence of calculations that determines long-term TSR the only real variable under management control is the plan— shown in the chart as “value of plan above market.” The short-term evidence of how well management is performing toward meeting its long-term goal is whether it’s making plans that will put it on the right TSR trajectory, and whether it’s accomplishing the substance of those plans.
Management therefore should be asked two basic questions:
Each of these questions, of course, leads to a host of subsidiary questions that lie beyond the scope of this article. But notice that even these very general questions are quite different from those questions that commonly frame management performance issues. Questions like, How much are you going to reduce costs next year over this year? What is your projected revenue growth? Are you managing within your capital budget? These latter questions are important, but their relevance and their relative priority need to be framed. We suggest that the framing is best achieved by anchoring all economic performance measures in the specifics of a long-term TSR plan.
TSR should be an important goal and measure of management. Properly employed, it directly links management priorities to the interest of shareholders, and it provides a long-term perspective on the plans, investments, and strategies needed to build enterprise value. Improperly employed, however, it creates expectations for management performance that are inherently unattainable.
In reaction, management reaches for other standards of performance, and a metric of vital import is thereby sidelined in favor of more complicated, sometimes conflicting, potentially sub-optimizing metrics. TSR should play a more prominent role in management choices. Techniques are readily at hand for providing it the central billing it deserves.
1. EVA never has found particular favor among utilities, largely because it is typically treated as an all-or-nothing analytic approach that must be administered comprehensively and pervasively if it is to be used at all. When applied in this way, it can occupy more time in analysis than seems justified by the gain in management insight. But EVA does not need to be applied to the nth detail. It is illuminating at the corporate level in understanding the economic contributions of different elements of the corporate portfolio. It then can be applied in cascading levels of detail to whatever depth management finds useful, and no further. Even at a high level of generality it provides important insight into an enterprise’s sources of value.
2. It’s important not to confuse return on equity or return on assets with return on share price. Whereas return on equity can be higher than the cost of equity capital (thereby creating value) or lower (thereby wasting value), the expected return on share price always will tend to equal the cost of equity capital. If the expected payout in dividends and price appreciation drops, the price will drop; if the expected payout rises, the price will rise. In either event, the percentage return gravitates toward a consistent cost-of-equity capital. The wide swings in share price observable over the past few years may seem to belie this contention, until one reflects that in the wake of Enron, California, etc., investors’ “traditional evaluation” of utility risk became unhinged. That disorientation (combined with substantial re-evaluation of growth expectation for many companies) drove down share price steeply. Once calm returned with the industry’s strategic retrenchment and with pervasive balance-sheet reconstruction, the sense of untamed risk subsided and share prices rebounded. Needless to say, this was an unusual period, but not inconsistent with the basic risk-growth price relationship discussed herein.
3. This is an article, not a textbook, and neither space nor the reader’s patience permits discussion of all the exquisite complications that can be layered on this arithmetic process. For instance, what if the company issues new shares and new equity? What if the company re-engineers its finances to modify its WACC? How does a contemplated merger affect these calculations? And so forth. The basic answer is, this approach still works, you just have to make the peripheral adjustments necessary to take account of each of these complications.