Comments on the Cost of Capital Methodology Used by the Canadian Transportation Agency Manitoba is pleased to submit comments in response to the Consultation Document issued by the Canadian Transportation Agency ( CTA or The Agency ) on November 4, 2010 concerning the Agency s cost of capital methodology review. These comments also refer to the Review of Regulatory Cost of Capital Methodologies by the Brattle Group that was attached to the Consultation Document ( the Brattle Report ). I. Background This inquiry is being conducted in respect of two railways, the Canadian National ( CN ) and the Canadian Pacific ( CP ), both of which are highly profitable, as demonstrated in Table I: Year Table I Canadian National Book Value Earnings Return on Canadian Pacific Book Value Earnings Return on Book Equity per Share per Share Book Equity per Share per Share 2005 14.65 2.35 16.0% 23.56 2.81 11.9% 2006 16.49 2.92 17.7% 26.86 3.40 12.7% 2007 20.97 3.40 16.2% 35.61 4.32 12.1% 2008 18.49 3.04 16.4% 31.95 3.33 10.4% 2009 22.66 3.08 13.6% 37.82 2.62 6.9% 2010(1) 25.31 4.52 17.9% 30.01 3.67 12.2% 2011(2) 29.35 4.65 15.8% 34.15 4.50 13.2% (1) Book value Sept 30, 2010; Earnings 9 mos. Annualized (2) Value Line Forecast Source: Value Line and 3Q 2010 Financial Statements With the exception of a single year by one railway (CP in 2009) both railways have consistently earned returns in the double digits every year since 2005, and they are 1
Jan-00 Jun-00 Nov-00 Apr-01 Sep-01 Feb-02 Jul-02 Dec-02 May-03 Oct-03 Mar-04 Aug-04 Jan-05 Jun-05 Nov-05 Apr-06 Sep-06 Feb-07 Jul-07 Dec-07 May-08 Oct-08 Mar-09 Aug-09 Jan-10 Jun-10 Nov-10 1= Jan. 2010 expected to continue to do so in 2010 and 2011. Indeed, the earnings per share for both railways are forecast to reach all-time highs in 2011. These high profits are reflected in the share prices of the two railways. Chart I shows the extent to which the stock performance of both railways has far exceeded that of the New York and Toronto exchanges. Chart I 12 Comparsion of Stock Prices from 2000-2010 10 8 6 4 CNI CP S&P 500 TSX 2 0 Source: Yahoo.com For both railways, the current price of common shares (US$67.50 for the CN; US$65.50 for the CP) is close to or at their all-time highs. The Agency s finding as to railways cost of capital affects their allowed revenue from the carriage of grain, principally from the Prairie Provinces. In Decision No. 512- R-2010, the Agency found that the CN s grain revenue was approximately $463.9 millions and the CP s grain revenue was $464.0 millions. In 2009, CN s total revenue was $7,367 millions, and CP s revenue was $4.303.2 millions, so that statutory grain 2
revenue accounted for a relatively small portion of the total revenue of each railway, 6.4 percent for the CN; 10.6 percent for the CP. The revenues earned by the railways in the carriage of grain represent a direct deduction from the revenues received by grain producers. That is because the prices of grain, unlike the rates charged by the railways, are not determined by or even related to the costs of producing grain. Rather, they are established by the interplay of demand and supply for the respective crops worldwide. The railways enjoy cost-based rates and revenue; grain producers do not. While the Agency has a statutory obligation to provide an adequate rate of return to the railways, it also has a public interest obligation to ensure that the returns allowed are the very minimum needed to meet the railways capital return requirements. For this reason, any ambiguity or uncertainty in cost of capital determination should be decided in favour of a lower return. Manitoba also stresses the importance of the fact that the Agency s overall costing model consistently overstates rail costs because it is not adjusted for rail productivity increases. In this regard, it contrasts with the rail costing metric used in the U.S. by the Surface Transportation Board ( STB ) that is adjusted quarterly for rail productivity increases. Rail productivity in the United States is measured by the Rail Cost Adjustment Factor (RCAF). RCAF is based on data produced by the railroads which is reviewed by the STB each quarter. As shown in Chart II, RCAF when adjusted for productivity, shows a consistent record of increasing rail productivity and declining rail costs over a 21 year time span (1989-2010) This highlights a major issue that is not thus far been addressed by the Agency in the cost of capital review, but increasing rail productivity is a significant contributor to the railways profitability from grain transportation and should be recognized as such. 3
Chart II SK Snavely King The Rail Cost Adjustment Factor (RCAF) 1.4 1.2 1 Rail Cost Adjustment Factor Rail Cost with Productivity Has Decreased since 1989 RCAF-U is Rail Cost Unadjusted for Productivity RCAF-A is Rail Cost Adjusted for Productivity Costs and Productivity both as Measured by Rail Industry 4Q 2007 = 1.0 0.8 0.6 0.4 0.2 RCAF-U RCAF-A 0 1989 1992 1995 1998 2001 2004 2007 Data Source: AAR, ICC, and STB The RCAF is a logical candidate for use in updating rail costs to the current quarter. The Rail Cost Adjustment Factor is frequently used in negotiations and other rate related matters. The RCAF is based on data assembled by the AAR and largely collected from the railroads. The RCAF is reviewed and adjusted as appropriate by the STB on a quarterly basis. The RCAF-A 20 year declining cost pattern, with costs adjusted for productivity, suggests the potential for significant revisions to variability estimates. 1996-2010 Snavely King Majoros & O Connor, Inc. 15 The grain transportation at issue in this proceeding is generally efficient and benefits from productivity gains similar to those shown in Chart II. For example, over 80 percent of grain rail shipments are now loaded by high throughput grain elevators in train blocks of 50 cars or more a major efficiency and cost-reducing operational strategy for railways. As well, since 1999, rail line abandonment across Western Canada has shed many of the railways high-cost branch lines, further reducing costs for the railways. Ultimately, these costs are transferred to: Provincial governments (who maintain the road network and support shortline railways). Farmers (who must pay for the increased trucking) and Grain companies (who are forced to construct new high throughput elevator facilities capable of loading multiple car blocks). We urge the Agency to bear these issues in mind and reflect these productivity gains and capital enhancements in making its capital cost assessments. 4
II. Capital Structure The Consultation Document inquires whether the book value or the market value of capital should be used in calculating the overall return to capital. The Agency should use the book value of the components of capital structure for three reasons. First, the alternative of using market values makes the regulatory process not reasonable because it is circular. The market value of common equity is driven by investors perceptions of the prospective profitability of the subject company. The Agency s findings as to the cost of capital affect that profitability. A generous cost of capital results in increased profitability, which in turn increases the market value of equity, which in turn increases the Agency s estimate of the cost of capital. Book values are not subject to the effects of the Agency s findings. In contrast to market values, book values are based on the historical record of investment by bond and stock holders, which is unaffected by the Agency s decisions. transparent. Book values are both reliable and pragmatic because they are stable and Market values, on the other hand, vary daily -- indeed hourly. Notwithstanding the Agency s treatment of the railways capital structures as confidential, they are readily discernable from the annual stockholders reports, as shown in Table II: Table II Capital Structure As of September 30, 2010 (CDN$ in millions) Canadian National Canadian Pacific Long-term Debt $ 6,461 28.3% $ 4,744 42.3% Deferred Taxes 5,119 22.4% 1,819 16.2% Common Equity 11,233 49.2% 4,658. 41.5% Total Capital $ 22,813 100.0% $ 11,220 100.0% Finally, it should be noted that the investments to which the Agency s cost of capital findings are applied are all expressed in book values. It is both inconsistent and incorrect to estimate the cost of capital based on market value and apply that estimate to a book value rate base. 5
III. Treatment of Deferred Taxes Deferred taxes are an allocated amount during the period to cover tax liabilities that have not been paid. They are a non-cash expense that provides a source of free cash flow. This cash is available to the railways at no cost. Most of this tax liability is due to the fact that rail companies, such as CN and CP, often depreciate fixed assets for tax purposes at a higher rate than for accounting purposes. In actual practice, the deferred tax liability is usually postponed indefinitely as financially prudent companies continue to invest and upgrade their capital infrastructure. The Consultation Document inquires whether the Agency should continue to give weight to deferred taxes at zero cost in the capital structure. Manitoba submits that it should continue this treatment. The alternative, which is to subtract deferred taxes from the rate base (investment) to which the cost of capital is applied, fails the test of pragmatism. This treatment is feasible only when the rate base corresponds to the entirety of the company s capital. That is not the case in any of the applications of cost of capital by the Agency. In assessing the railways grain transportation costs, for example, the Agency applies a cost of capital factor to the investment used to transport grain, which is only a portion of the total investment of the major Canadian railways. IV. Cost of Long-term Debt The Agency should use the embedded cost of debt as of the end of the most recent quarterly reporting period. The suggestion that forecasts of interest rates should be used fails the test of reasonableness because it would apply different interest costs to the railways debt than the railways actually pay. This procedure also fails the test of transparency because the methodology by which analysts forecast interest rates is proprietary and probably largely judgemental. It fails the test of pragmatism because forecasts will vary both among analysts and over time. The Agency should use the yield-to-maturity method of estimating the cost of debt. This procedure factors in the costs of debt issuance and any discount or premium 6
that the railway may have received when it initially sold the bond issues. It does not, however, reflect current discounts or premiums on the outstanding debt. V. Cost of Equity The Brattle Report discusses four methodologies for estimating the cost of equity, the Capital Asset Pricing Model, Discounted Cash Flow, Risk Premium Approaches, and Comparable Earnings. A. Capital Asset Pricing Model (CAPM) The Brattle Report discusses the issues associated with implementing the CAPM in considerable depth, and these comments will not repeat or elaborate on those issues. We recommend that the Agency continue to use the CAPM as its basis for finding the cost rate of equity. The CAPM is reasonable because it is consistent with the objective of providing federally regulated railway companies with a fair and reasonable return. Unlike the other methodologies reviewed in the Brattle Report, the CAPM relies on a measure, the beta, which is discrete and specific to each railway company. The other methodologies rely on measures that are based largely on groups of proxy companies. The CAPM is reliable and pragmatic because it relies on externally generated factors that are not subject to interpretation and judgment. In this regard, Manitoba strongly objects to the Agency s practice of holding its calculations of the CAPM confidential. The CAPM does not use competitively sensitive information. Rather, it relies on inputs that are drawn from public information on markets and interest rates. This review would be much more informative if the review included an evaluation of the Agency s actual calculations and data assumptions. In addition to the conventional application of that methodology, the Agency should consider the results of the Fama-French Three Factor Model, described in 7
Appendix B (page 117) of the Brattle Group report. This model is based on the findings of E. Fama and K.R. French that company size and the market-to-book relationship, in addition to the beta, strongly influence equity returns. This model enhances the CAPM by avoiding the implicit assumption that the beta is the sole measure of a company s risk relative to the market. An example of this variant of the CAPM is presented in Attachment A hereto. This implementation of the Fama-French approach to calculating the cost of equity for U.S. railways has been presented by the U.S. Surface Transportation Board ( STB ) in rail rate reasonableness proceedings. In these proceedings the STB was examining whether prior cost of capital decisions needed to be restated due to implantation of the CAPM and 3 Stage DCF approaches (and the discontinued use of the single stage DCF model). The Fama French approach was used as one of the measures to show that prior cost of capital determinations were within reason. The STB calculated the cost of equity using the Fama-French approach using Morningstar s cost-of-equity estimates approach. The Consultation Document poses a number of specific issues with respect to the implementation of the CAPM. 1. Should the Agency use U.S. data? The Agency should continue to use Canadian data because the investments in grain transportation to which the cost of capital is applied are entirely within Canada. 2. Should the Agency use short-run or long-run inputs to the CAPM model? The Agency should use yields on government bonds in the one to three year horizon as the best indication of the risk-free rate. Yields on shorter term debt, e.g. six months, reflect fluctuations in liquidity and are likely to be unstable. Yields on longer term debt reflect the risk that future inflation may erode the value of the underlying bonds. They are therefore not fully risk-free. 8
Betas are conventionally computed over five years, and this appears to be the appropriate time period to reflect current conditions but still mask variations resulting from short-term ephemeral influences. B. Discounted Cash Flow ( DCF ) The theory underlying the DCF approach is reasonable and consistent with valid economic principles. The procedure founders, however, on issues of implementation. The most common implementation uses long-term earnings growth rates that are forecast by investment analysts. The Agency correctly observes that these forecasts are unstable, that is, they vary over time. They also vary dramatically among analysts. Thomson Financial surveys investment analysts for their forecasts of earnings and dividends. For each railway, Thomson found four analysts willing to make forecasts of long-term earnings growth. For the CN, the highest estimate was 11.0 percent, the lowest was 7.1 percent. For the CP, the spread was even greater, the highest was 12.0 percent; the lowest was 4.0 percent. The only way to smooth these variations is to survey a large number of analysts and to use a large proxy group of financially sound, dividend paying companies in the same industry. Here again, the DCF founders on the issue of implementation. There are only two large Canadian railways. The sample can be expanded by including U.S. Class I railroads, but there are only three that fit the requirements of the DCF approach. The Burlington Northern Santa Fe (BNSF) has been acquired by Berkshire Hathaway and is no longer traded publicly. Berkshire Hathaway now reports BNSF data with its other utilities holdings. The Kansas City Southern Railroad does not issue dividends. That leaves only the Norfolk Southern, the CSX and the Union Pacific to add to the sample of railways. Five companies is a very small proxy group for DCF estimation. The foregoing involves problems of reliability and pragmatism. As applied to the railway industry, there is a further problem of reasonableness, which is that the median earnings growth rates currently forecast for the two major Canadian railways (9.5% for 9
the CN and 10.2% for the CP) are unsustainable in the long run. The Office of the Parliamentary Budget Officer currently forecasts the Canadian economy to experience a growth rate of real Gross Domestic Product ( GDP ) of 1.7 percent through 2030. 1 It predicts a long-term inflation rate of 2.0 percent. 2 The sum of these forecasts is a longterm growth in nominal GDP of 3.7 percent. The predicted earnings growth rates for the railways are multiples of this forecast national GDP growth. The projection of these railway growth rates into the indefinite future would imply that the railways consume more and more of the national output over time a totally unreasonable assumption. Ultimately, the sustainable rate of the railways earnings growth is constrained by the growth of the economy in which they operate. Confronted with this problem, the U.S. Surface Transportation Board has employed a three-stage DCF model 3. This model uses three growth rates, (1) growth in years 1-5 is assumed to be the median value of annual earnings growth rate in the 3 to 5 time horizon as forecast by railroad industry analysts. (2) growth in 6-10 years is the average (mean) all analysts growth rates from the first stage. (3) growth in 11 years and onwards is long run nominal growth rate for the U.S economy. If the Agency uses the DCF model at all it should use this or a similar variant of the DCF application C. Risk Premium Approaches The Brattle Report discusses two risk premium estimation techniques, historical and prospective. The historical approach identifies the differential between experienced returns on equity and bond yields over an extended period of time. The prospective 1 Office of the Parliamentary Budget Officer, Fiscal Sustainability Report, February 18, 2010, page 11. 2 Id. Page 12. 3 U.S. Surface Transportation Board Docket Ex Parte 664 (Sub-No.1), Use of a Multi-Stage Discounted Cash Flow Model in Determining the Railroad Industry s Cost of Capital, Decision served on January 28, 2009 10
approach calculates a DCF return during a number of recent periods and derives the risk premium by comparing those results with bond yields. The historical approach mirrors the procedure used to determine the market risk premium in the CAPM, only without any company-specific inputs. In this regard, it fails the reasonableness test because it does not ensure that the railways will receive a fair and reasonable return. It is clearly inferior to the CAPM. As described in the Brattle Report, the prospective approach to the risk premium model involves finding the rate of return on equity in order to find the rate of return on equity. This circular process begins with DCF studies of the equity return for either the target company or a group of comparable companies during several historical accounting periods. This procedure encounters all of the problems of reasonableness, reliability and pragmatism with the DCF that are outlined above. Assuming, however, the validity of the DCF results, the equity returns are then compared with bond yields during the same periods. The differences are then plotted to yield a statistical relationship between bond yields and equity return requirements. The obvious flaw in this methodology is that it starts with the solution, i.e. the DCF rate of return. Moreover, it only loosely relates the results to the target railways. These are clear violations of the reasonableness test. For the foregoing reasons, we recommend that the risk premium procedures not be employed in finding the railways rate of return. D. Comparable Earnings As described in the Brattle Report, the comparable earnings approach involves the selection of a number of unregulated companies with risk metrics similar to the target company. The average earned return on book equity during recent reporting periods is then assumed to be the return the company would have received if it were investing the 11
same amount in other securities possessing an attractiveness, stability and certainty equal to that of the company s enterprise. 4 Aside from the weaknesses pointed out in the Brattle Report (p. 59), this approach suffers from the fundamental flaw that accounting returns do not describe comparable earnings of unregulated companies because they are not available to investors. The return on book equity will match the return available to equity investors only in the rare instances when the market value of the stock matches its book value. Otherwise, the return available to the investor is the company s earnings relative to the market value of the stock, not its book value. For most successful unregulated companies, that market value is greater than the book value, so that the accounting return on book equity is an overstatement of the return the company would have received if it were investing the same amount in other securities possessing an attractiveness, stability and certainty equal to that of the company s enterprise. Since this approach clearly fails the test of reasonableness, we recommend that the comparable earnings approach not be employed in finding the railways rate of return. VI. Assessment of a Grain Risk Adjustment The carriage of grain is a risk-reducing factor for the railways for two reasons. First, the railways have a virtual monopoly over the long-distance carriage of Canadian grain, which reduces the risk of transporting this commodity relative to most of the rest of the railways traffic. A demonstration of this effect was recently performed by William E. Avera in a rate case involving the Baltimore Gas & Electric Company in Maryland. Dr. Avera selected two lists of companies, one a group of 13 regulated gas and electric combination utilities with S&P bond ratings of A-, BBB+, or BBB, and with Value Line safety ratings of 1 or 1 and financial strength rates of B++ or better. The other group was 69 unregulated companies with comparable safety and financial strength ratings. He 4 Northwestern Utilities Limited v. City of Edmonton (1929), Supreme Court of Canada. 12
conducted DCF studies using six different approaches to developing the g or growth factor. The results, as shown in Table III were as follows: 5 Table III Cost of Equity Estimates V Line IBES First Call Zacks br+sv Price Average 69 Unregulated Companies 11.9% 12.6% 12.8% 12.7% 12.2% 13.7% 12.7% 13 Regulated Utilities 9.9% 10.7% 10.4% 10.5% 10.9% 12.5% 10.8% Difference, regulated v. Unregulated -2.1% -1.9% -2.4% -2.3% -1.2% -1.2% -1.8% Note: Differences may not subtract exactly due to rounding. In every DCF application, the indicated return for the regulated monopoly companies was at least 1.2 percent lower than the corresponding return for the comparable unregulated companies. The average differential was 1.8 percent. These results demonstrate that quite regardless of comparable safety and financial strength ratings, investors regard monopoly companies as significantly less risky than competitive companies. The same risk differential applies to monopoly vs. competitive segments of the railways transportation markets. The second reason that grain transportation is less risky is that the volume of grain is not tied to the general economy, as with the other commodities and merchandise that are transported by rail. This fact was supported in 2009 when most rail traffic experienced significant declines in volume due to the economic recession. During this period grain volume actually increased (due to good crop growing conditions and a healthy beginning inventory of grain stocks) and worked to lessen the overall drop in freight revenues for major Class 1 railways in Canada. The Agency in 1985 ruled that grain traffic actually reduced the overall risk for railways because grain revenues were not correlated with business cycles in the economy as were revenues for other non-grain traffic. 5 Maryland P.S.C. Case No. 9230, Testimony of William E. Avera, Exhibits WEA-2 and WEA-4. 13
These lower risk characteristics should be reflected in the cost of capital applied to the investment associated with grain transportation. In the past, the Agency has applied a reduction of 100 basis points to the grain-related rate of return on equity. The foregoing evidence suggests that the adjustment should be as much as 180 basis points. Manitoba recommends a downward adjustment in the rate of return on equity of investment associated with grain transportation of between 100 and 180 basis points (1.0% to 1.8%) to recognize the risk-reducing characteristics of this traffic. VII. Conclusion Manitoba supports the methodology that the agency has employed for finding the cost of capital for grain transportation. Changes in the cost of capital methodology should only be considered by the Agency upon a more comprehensive costing review that would implement productivity adjustments in estimating the grain revenue Cap. In addition, the Agency should consider providing a 100 to 180 basis point reduction in the grain-related cost of equity to reflect the much lower risk associated with the regulated monopoly characteristics and the counter-cyclical nature of grain transportation. 14