How EVA works against GATX.

Glasser contends that EVA penalizes companies - like his - whose assets yield long-term returns.

Economic Value Added has been touted in boardrooms for the last few years as the best available method to judge the performance of companies and their managers. While I agree we should seek new ways to measure company performance, I am convinced that the EVA approach can penalize companies that invest in assets with long-term returns. For that reason, I think EVA should not be the only method used to evaluate a company's performance.

This is true for companies such as GATX. We lease $4 billion in transportation equipment such as railcars, pipelines, and commercial aircraft. Investments in these assets, as you might expect, are for the long term. Our assets generate a regular revenue stream while we hold them, and a lump sum (which we call disposition gains), when we sell them. Our goal is to invest wisely for the best long-term returns. The irony is, the more money GATX invests, the lower its EVA. When we restrict our investments, our EVA improves.

The EVA principle is a sound one: To get managers and employees to think and act like investors. For example, an EVA-oriented investor would compare the return from a stake in the company to any other valuable investment by holding management accountable for effective use of the corporation's cash and assets. Managers, therefore, are expected to earn a rate of return on capital higher than the project's cost of capital, thus increasing value - and creating wealth - for investors.

FIGURE 1

GATX CORPORATION
1193 EVA Calculation ($ in millions)

1992 Capital                   3,767
Assumed WACC                    9.3%
Cost of Capital                  350

Less:

1993 NOPAT                       174
EVA                            (176)

As part of the EVA calculation, the GATX reporting basis required
these adjustments: Debt: includes off-balance sheet items.
Allowance for Losses: Noncash, "precautionary." Goodwill:
Cumulative impact of noncash accounting requirement. Construction
in Progress: Not yet operating. Cumulative Unusual Losses: Cost of
unsuccessful efforts.

Analysts and investors have used many methods to analyze companies' cash flows and the sources and uses of their cash to generate future earnings. EVA deviates from more traditional approaches by assigning an opportunity cost to cash flow. Opportunity cost reflects the value of capital in terms of the loss of the opportunity to use it elsewhere. Generally, this cost is measured by what the capital would earn in the safest investment around: the yield available on U.S. Treasury bills. By assigning an opportunity cost to cash flow, EVA reveals the actual return managers are getting on the investors' dollars.

It is the rate of return on capital (after-tax operating profit) divided into shareholders' equity (total equity capital), minus the opportunity cost of that capital. Under the EVA model, the opportunity cost is the return investors expect on the company's mix of outstanding equity (stocks) and borrowings. The bigger the difference, the higher the EVA. In fact, the EVA calculation is used to determine the worth of new transaction and business opportunities.

FIGURE 2

GATX CORPORATION
1993 Net Income and NOPAT ($ in millions)

                           Net Income     Adjustments     NOPAT

Gross Income               1,087          (17)            1,070
Operating Expenses         528            (30)            498
Interest                   152            (152)           0
Goodwill Amortization      4              (4)             0
Loss Provision             30             15              45
All Other                  269            0               269

Pre-Tax Profit             104            154             258
Income Tax                 (51)           53              (104)
Equity Earnings            20             0               20

Net Income/NOPAT           73             101             174

EVA calculations require a restatement of Reported Net Income to
calculated Net Operating Income After Taxes (NOPAT). Figure 2
reflects the following adjustments: Gross Income: an insurance
settlement is deleted from Gross Income. Interest: Unleveraged.
Goodwill: Noncash Accounting Requirement. Loss Provision: +/-
Write-offs. Taxes: Cash Taxes After Adjustments. Depreciation:
"Economic" Asset Usage.
FIGURE 3

GATX CORPORATION
1993 Capital-Financing Approach ($ in millions)

                        Book Basis     Adjustments     EVA Basis

Debt                    1,940          326             2,266
Deferred Liabilities    619            0               619

Equity Equivalents:
Allowance for
Losses                  0              96              96

Goodwill
Amortization            0              15              15

Construction in
Progress                0              (41)            (41)

Cumulative
Unusual Losses          0              181             181

Shareholder Equity      590            0               590

Unidentified            0              (13)            (13)

Total Capital           3,149          564             3,713
FIGURE 4

GATX CORPORATION
1993 Capital-Operating Approach ($ in millions)

                          Book Basis     Adjustments     EVA Basis

Cash & A/R                877            0               877

Loss Allowance            (96)           96              0

PP&E-Gross                3,305          (41)            3,264

Allowance for
Depreciation              (1,343)        0               (1,343)

PV of Operating Leases    N/A            326             326

Goodwill                  137            0               137

Cumulative Goodwill
Amortization              N/A            15              15

All Other                 269            0               269

Cumulative Unusual
Losses-AT                 N/A            181             181

Unidentified              0              (13)            (13)

Total Capital             3,149          564             3,713

Of course, adding value is better than losing it. But components of EVA penalize older, capital-intensive companies, for which the payout is long-term (30 years or more, in GATX's case). The cash-flow generating capabilities of GATX are far different from those of a consumer-products company whose assets begin generating high revenues and returns almost immediately, but are often short-term in nature. As a result, businesses such as GATX are nearly always on the bottom of Stern Stewart's Performance 1000 list (the most recent of which was released in December 1995). For GATX, it is more important how we add value over time than to improve our absolute position on the list. Although in the short term, GATX may receive lower annual returns on its capital investment, over a 30-year period, our returns are extremely competitive with those higher on the Performance 1000 list.

The EVA formula works against companies that have cyclical periods of reinvestment. Over the last five years, GATX has reinvested more than $3.5 billion in new assets. In the previous five years, less than half that amount was invested. Recently, as reinvestment for new long-term opportunity was high, our EVA was lower. When our reinvestment was lower, our EVA was higher. Since GATX's assets are typically long-term in nature, they provide returns and economic value over the course of several years.

GATX UNDER EVA

The EVA calculation is as follows (the data used here are from 1993, the most recent information available at this writing):

EVA = Net Operating Profit After Taxes (Total Beginning Capital X Weighted Average Cost of Capital). Figure 1 shows GATX's EVA calculation, which is based on the calculations made in Figures 2, 3, and 4.

UNATTRACTIVE OPTIONS

As a result of these calculations, GATX reviewed a variety of scenarios in which its EVA could be increased by using the Stern Stewart approach. One way is to increase the company's NOPAT on existing capital by investing less capital. GATX could invest in higher-return projects, but that would mean changing the inherent risk profile of the company, which is unattractive to company management and likely would be undesirable to its investors.

We have concluded that the EVA framework is not substantially different from the Return on Total Capital Employed methodology, which we currently use to measure our return on capital investments.

Most important, we believe that EVA is an excellent model to determine whether a company's management is adding or losing value. But EVA discounts long-term value inherent in a capital structure such as GATX's, where constant reinvestment replenishes assets that are sold at the end of a lease term or are retired from service.

We will continue to watch EVA because it is an increasingly common measure in the financial community. But we believe the quality of our cash flow, supported by long-lived assets on long-term leases to our customers, provides our shareholders with consistently acceptable returns.

We think our market leadership, commitment to meeting customer and shareholder needs, the quality of our cash flow, and our overall risk profile give us substantial opportunity for long-term growth.

James J. Glasser is chairman and CEO of $1.5 billion GATX, based in Chicago. The company leases transportation and distribution equipment such as railcars, pipelines, ships, warehouses, and commercial aircraft. The company also operates a financial-services unit that focuses on owning or managing transportation and distribution assets.

Ralph L. O'Hara, controller, and George S. Lowman, director of communications and investor relations, contributed to this article.