EFTA01415820.pdf

31 October 2017 Railroads Canadian Rails Deutsche Bank Markets Research North America Canada Industrials Railroads Industry Canadian Rails Date 31 October 2017 Initiation of Coverage Flipping the Script - Buy CP, Sell CNI Seldon Clarke, CFA Associate Analyst Amit Mehrotra Research Analyst Kenya Watson Research Associate Chris Snyder, CFA to Initiating Coverage of Canadian Rails We are expanding our coverage of Transportation companies with initiation of coverage of the Canadian railroad industry. We are positive on Canadian Pacific (CP, Buy) and cautious on Canadian National (CNI, Sell), as we see overall market share and earnings trajectory driving a reversal in recent relative value trends. All told we forecast CP to grow EPS at double the rate of CNI (+30% for CP vs. +15% for CNI cumulatively through 2019), which together with capex and free cash trends should drive re-rating at CP and de-rating for CNI. In this report we present a primer on the Canadian rail industry, with deep dives on Canadian Pacific and Canadian National. Deutsche Bank Securities Inc. Distributed on: 31/10/2017 20:03:27 GMT Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. DISCLOSURES AND ANALYST CERTIFICATIONS EFTA01415820 ARE LOCATED IN APPENDIX 1. MCI (P) 083/04/2017. Obed7b6cfllc EFTA01415821 EFTA01415822 Railroads Canadian Rails Deutsche Bank Markets Research North America Canada Industrials Railroads Industry Canadian Rails Date 31 October 2017 Initiation of Coverage Flipping the Script - Buy CP, Sell CNI Initiating Coverage of Canadian Rails We are expanding our coverage of Transportation companies with initiation of coverage of the Canadian railroad industry. We are positive on Canadian Pacific (CP, Buy) and cautious on Canadian National (CNI, Sell), as we see overall market share and earnings trajectory driving a reversal in recent relative value trends. All told we forecast CP to grow EPS at double the rate of CNI (+30% for CP vs. +15% for CNI cumulatively through 2019), which together with capex and free cash trends should drive re-rating at CP and de-rating for CNI. In this report we present a primer on the Canadian rail industry, with deep dives on Canadian Pacific and Canadian National. CANADIAN NATIONAL (CNI): A Victim of Its Own Success; Initiate Sell/$73 PT We see 1096+ downside in CNI shares as the company's strong outperformance starts to slow- driven by both the law of large numbers (the company already achieves a mid 40's operating margin, up from the high 30's five years ago) as well as catch-up performance from CP. For example, we note that from 2012-2016 CNI increased volumes at more than double the rate of CP, reflecting mix as well as market share gains during CP's implementation of Precision Railroading. The combination of CNI's slower prospective earnings growth and high capex (20% of sales) implies 15.5x P/E under our DCF-derived methodology, implying potential for 3.5 turns (20%) valuation de-rating vs. current trading levels. Initiate Sell. CANADIAN PACIFIC (CP): Shifting Gears; Initiate Buy/$209 PT Following its multi-year implementation of Precision Railroading, CP is shifting gears from cost take-out to top-line growth. As such, we see at least 15% EFTA01415823 upside in shares as CP leverages its reduced cost base, improved service levels, and recent capacity investments to retake market share. We expect this to translate to 30% cumulative EPS growth over the next two years, reflecting midsingle digit revenue growth, significant operating leverage, and accelerated share repurchase. Against this backdrop we see CP's relative valuation discount as unsustainable, which underpins our positive stance to shares. Initiate Buy. Valuation and Risks: We utilize P/E multiples to value rail stocks, with our target multiple assumptions heavily supported by our discounted cash flow models. Risks for the group include recession, industrial production, pricing, and mgmt. execution. For more details please company-specific CP and CNI sections within this note. Seldon Clarke, CFA Associate Analyst Amit Mehrotra Research Analyst Kenya Watson Research Associate Chris Snyder, CFA Research Associate Companies featured Canadian Pacific (CP.N),USD174.74 Canadian National (CNI.N),USD81.21 Source: Deutsche Bank Buy Sell Deutsche Bank Securities Inc. Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MCI (P) 083/04/2017. EFTA01415824 31 October 2017 Railroads Canadian Rails Table Of Contents Executive Summary 3 North American Railroads 10 Rail End Markets Overview and Outlook 23 Railroad Valuation 34 CP Company Overview 40 Management Overview 48 Valuation 49 Financial Statements 51 CNI Company Overview 56 Management Overview 65 Valuation 65 Financial Statements 67 Page 2 Deutsche Bank Securities Inc. EFTA01415825 31 October 2017 Railroads Canadian Rails Executive Summary We are expanding our coverage of Transportation companies with initiation of coverage of the Canadian railroad industry. We are positive on Canadian Pacific (CP, Buy) and cautious on Canadian National (CNI, Sell), as we see overall market share and earnings trajectory driving a reversal in recent relative value trends. All told we forecast CP to grow EPS at double the rate of CNI (+30% for CP vs. +15% for CNI cumulatively through 2019), which together with capex and free cash trends should drive re-rating at CP and de-rating for CNI. In this report we present a primer on the Canadian rail industry, with deep dives on Canadian Pacific and Canadian National. We do not believe current valuation appropriately reflects the relative earnings trajectory and shifting market dynamics within the Canadian rail industry. We believe CP is well positioned to regain market share from CNI as it leverages its lowered cost base, improved service levels, and recent capacity investments to retake market share. In our view, these efforts will help CP achieve industryleading volume and earnings growth over the next several years (ex-CSX). A lower cost base and better service levels should help CP regain share From 2012-2016, CNI increased volumes (revenue ton-miles [RTM's]) at more than double the rate of CP (+3.0% CAGR vs. +1.2% CAGR for CP). While mix likely played a factor, we believe the key driver behind this was CNI's ability to win market share as the lower-cost carrier with superior service. Further, this came at a transitional time for CP as the focus was largely on the implementation of Precision Railroading which likely pushed freight onto other transportation modes as well. As you can see below, CNI had nearly 18% points of cost advantage over CP (as measured by operating ratio) before CP began implementing Precision Railroading in 2012. This advantage has largely been erased, and we expect just a 200bps difference in operating ratios in 2017. Figure 1: Canadian RTM's 2004-2017E (2004=100) 100 110 120 130 EFTA01415826 140 80 90 Source: Deutsche Bank, Company filings 137.5 CP RTM's CNI RTM's 115.3 Figure 2: CNI's cost advantage has largely been erased 10% 12% 14% 16% 18% 20% (2%) 0% 2% 4% 6% 8% Source: Deutsche Bank, Company filings 17.8% 14.2% 2.0% 0.8% (0.0%) Moreover, CP has seen a significant improvement in service levels while CNI's service metrics are essentially inline with 2012 levels. Behind this improvement was CP's multi-year track upgrade program which was completed in 2015. While a more efficient railroad, better service, and lower costs typically all go Deutsche Bank Securities Inc. Page 3 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017E CNI cost advantage vs. CP 2005 EFTA01415827 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017E 2018E 2019E EFTA01415828 31 October 2017 Railroads Canadian Rails hand-in-hand, we believe it is important to note that almost all of CP's service improvements came after 2014, which is when the aforementioned volume trends started to diverge. Figure 3: CP's avg. train speed is up 28% since 2012 vs. down 5% for CNI 15% 25% 35% (15%) (5%) 5% 2013 2014 CP Source: Deutsche Bank, Company filings 2015 2016 CNI 2017 Figure 4: CP's dwell has improved by 13% vs. just 1% for CNI 10% 15% 20% (15%) (10%) (5%) 0% 5% 2013 2014 2015 2016 2017 CP CNI Source: Deutsche Bank, Company filings We believe CP's improvements are beginning to manifest in volume trends. As you can see below, volumes at CP have inflected positively relative to CNI in recent months and appear to be gaining momentum. CP is still playing catchup in Intermodal, though we note the majority of CNI's intermodal growth is coming from International, which we believe carries a lower margin than domestic intermodal, which represents a bigger piece of the pie for CP. Looking out over the next couple of years, we expect this momentum to continue for CP and forecast cumulative RTM growth of 7% by 2019 (vs. 2017) compared to 5% for CNI. Figure 5: CP's volumes began outpacing CNI in recent months 100 110 120 EFTA01415829 130 90 CP RTM's (ex-intermodal) CNI RTM's (ex-intermodal) Figure 6: and Intermodal volumes aren't far behind 100 110 120 130 140 90 CP Intermodal RTM's CNI Intermodal RTM's Source: Deutsche Bank, Company filings Source: Deutsche Bank, Company filings The picture becomes even clearer when we look at the number of carload classifications growing relative to those declining on a yoy basis. As you can see below, CP's carload strength appears to be broad -based and gaining momentum. Page 4 Deutsche Bank Securities Inc. 4-wk avg.( 100=Week 1 '17) % chg. From 2012 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 Oct-17 28% (5%) 4-wk avg.( 100=Week 1 '17) Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 Oct-17 % chg. From 2012 (13%) (1%) EFTA01415830 31 October 2017 Railroads Canadian Rails On the other hand, CNI's carload trends appear to be losing steam with Intermodal and "Other" being the only carload groupings up yoy over the past four weeks. Figure 7: CP carload growth appears to be gaining momentum... Carloads growing YoY Carloads declining YoY 2 7 4 5 2 2 2 2 7 7 7 7 4 5 2 7 4 3433 5 6 566 5 6 4 3 8 76 1 2 3 4 5 6 3 8 1 6 2 3 7 Figure 8: as CNI's carload trajectory appears to be losing steam Carloads growing YoY Carloads declining YoY 7 8 7 7 7 78 4 5 2 1 7 8 99 2 334 7 665 2 1 2 2 2 2 1 66 33 EFTA01415831 44 5 5 2 7 Source: Deutsche Bank, AAR Source: Deutsche Bank, AAR CP is putting the pieces together to leverage a better network After building a better foundation, i.e. lower-cost base and better service, CP has begun putting the pieces together to profitably grow its revenue base. A key piece, in our view, was the appointment of John Brooks as CM0 in February, 2017. In the 2.5 years prior to John's appointment as CMO, Keith Creel was largely responsible for overseeing sales and marketing at CP. At the same time, however, Creel was acting COO and being groomed to be Hunter Harrison's replacement. Combined with the major task of turning around North America's least profitable railroad, its easy to see how marketing/sales efforts largely went by the wayside. In addition to John's appointment as CMO, these efforts have been reinvigorated through a number of initiatives: ■ Trip Plan - Introduced in late 2016, Trip Plan allows customers to track shipments on a car by car basis thereby eliminating inefficiencies and increasing delivery accuracy. This service helped drive on-time performance above 90% in 2016 (well above industry avg). ■ Enhanced intermodal offering - CP recently opened its transload facility in Vancouver which helps steamship lines more efficiently manage their containers. Transload facilities enable the interchange of commodities to different rail cars / containers. This allows CP to utilize a steamship's containers for inland originating traffic destined for ports, thereby reducing costs to its customer by eliminating transportation costs on empty containers. To this point, empty export containers out of Vancouver are up over 40% ytd through September. ■ Expanding sales/marketing footprint - CP announced just last month that it is expanding its sales and marketing presence in Asia. Positions are being added in China and Singapore to expand business with current customers and attract new ones. This follows a number of other announcements over the past several months aimed at improving CP's product/market reach through increased communication with its customers. ■ Dedicated Train Program - Introduced during the 2014/15 crop year, CP's Dedicated Train Program (DTP) allows its customers to control Deutsche Bank Securities Inc. Page 5 EFTA01415832 Jan-16 Mar -16 May -16 Jul-16 Sep -16 Nov -16 Jan-17 Mar -17 May-17 Jul-17 Sep-17 Jan-16 Mar-16 May -16 Jul-16 Sep -16 Nov -16 Jan-17 Mar -17 May -17 Jul-17 Sep -17 EFTA01415833 31 October 2017 Railroads Canadian Rails their own rail assets for a full crop year. This program helps customers more efficiently manage their supply chains and demand for the service continues to grow (up 15% yoy in 3Q). While CNI has long been ahead of the curve in many of these efforts, we believe what will take place over the next several years is more of a natural shift in market share back to CP. Aside from service and cost, the major difference between any major Class I railroad is its geographical footprint. We believe much of the market share which CP lost to CNI over the past several years was simply due to the fact that CNI had a better overall service offering which in many cases outweighed any advantage CP may have had geographically. Now that CP has eliminated any service or cost advantage that CNI was able to take advantage of in the past, we expect to see a natural rebalancing of market share in CP's favor. Relative valuation premium/discount expected to flip in coming years With CNI still trading at a 7.5% premium to CP on a fwd. P/E basis, we believe current multiples imply that recent market share losses at CP are structural and will likely continue long-term. However, we believe CP is turning the corner and expect a natural rebalancing of rail business to support best-in-class volume growth over the next several years. With much of this rebalancing coming at the expense of CNI, we expect the 10% premium that CNI has been trading at to reverse over the next 12-14 months. Figure 9: CP NTM P/E relative to CNI NTM P/E since 2012 1.4x 1.2x 1.0x 0.8x Jun-12 Jun-13 Jun-14 Jun-15 Jun-16 Jun-17 CP/CNI Relative P/E Mean Source: Deutsche Bank, Company Reports +/- 1. std. dev Figure 10: CP's relative premium (discount) to CNI on a NTM P/E basis 10% 20% 30% (20%) (10%) EFTA01415834 0% '12 '13 Source: Deutsche Bank, FactSet '14 '15 '16 CP NTM P/E Premium (Discount) to CNI '17 DCF Framework supports our view on valuation All told we expect rail volume growth to track fairly closely with long-term industrial production and the overall economy. As such, we believe Canadian railroads should generate cumulative long-term top-line growth in the range of 3.5-4% (-2% volume and pricing). With CP in a better position to win market share over the next several years, in our view, we assume top-line growth at the top end of that range (4%) with CNI trending at the lower-end (3.5%) before returning to our long-term assumption of 3.75% growth in 2025+. As shown on the following page, our DCF framework for CP translates to an implied P/E multiple of roughly 17.5x our 2019 EPS estimate. Page 6 Deutsche Bank Securities Inc. EFTA01415835 31 October 2017 Railroads Canadian Rails Figure 11: Our DCF framework for CP translates to an implied P/E multiple of 17.5x CP - DCF Framework Gross Revenue YoY Change (%) EBIT % margin Implied Off BS Interest EBIT + Interest adjustment % margin Cash tax rate (%) Unlevered Free Cash Flow NOPAT % of revenue D&A % of revenue Net Capex % of revenue Working capital Unlevered FCF Time factor Discount factor PV of cash flows Price Target Derivation NPV Net Debt (3Q '17) PV of Operating Leases Implied Equity Value 4Q '18e Sharecount USD/CAD Implied Share Price WACC calculation Historical RFR Equity Risk Premium (ERP) Beta Cost of equity WACD Tax rate Cost of debt % equity % debt WACC Source: Deutsche Bank, Company filings, FactSet 44,356 7,991 288 36,077 138 EFTA01415836 $1.25 2.35% 7.00% 1.01 9.4% 5.5% 26.5% 4.0% 72% 28% 7.9% 2016 6,232 (7.2%) 2,578 2,600 2017E 6,543 2018E 6,831 2,960 24 2019E 7,265 3,226 25 2020E 7,556 3,400 2021E 7,858 3,536 27 2022E 8,172 3,678 29 3,706 2023E 8,499 3,825 30 2024E 8,839 3,978 31 2025E 9,193 4,137 32 Terminal EFTA01415837 Value 9,537 5.0% 4.4% 6.4% 4.0% 4.0% 4.0% 4.0% 4.0% 4.0% 3.75% 2,747 23 26 2,770 2,984 3,251 3,426 3,564 3,854 4,008 4,169 4,292 41.4% 42.0% 43.3% 44.4% 45.0% 45.0% 45.0% 45.0% 45.0% 45.0% 45.0% 22 33 4,325 41.7% 42.3% 43.7% 44.8% 45.3% 45.3% 45.3% 45.3% 45.3% 45.3% 45.3% 23.3% 15.0% 15.0% 15.0% 1,995 2,355 660 2,536 683 2,764 719 2,856 758 2,912 789 2,967 822 3,023 856 3,078 891 3,133 928 3,179 32.0% 36.0% 37.1% 38.0% 37.8% 37.1% 36.3% 35.6% 34.8% 34.1% 33.3% 640 963 10.3% 10.1% 10.0% 10.0% 10.0% 10.0% 10.1% 10.1% 10.1% 10.1% 10.1% (1,066) (17.1%) (55) (1,235) (18.9%) 11 EFTA01415838 0.0 100.0% (1,275) (18.7%) (10) 0.0 (1,325) (18.7%) 7 1.0 (1,417) (18.8%) 0 1.0 (1,478) (18.8%) 0 1.0 (1,541) (18.9%) 0 1.0 (1,607) (18.9%) 0 1.0 (1,675) (19.0%) 0 1.0 (1,747) (19.0%) 0 1,514.0 1,790.1 1,934.1 2,165.2 2,196.3 2,223.0 2,248.3 2,272.1 2,294.2 2,314.4 0.0 1.0 Other Assumptions Revenue growth 2020-2025 Terminal revenue growth Terminal EBIT Margin Terminal Tax Rate Terminal D&A (% of rev.) Terminal Capex (% of rev.) $209 <-- Translates to —17.5x our 2019 EPS estimate 4.0% 3.75% 45.0% <--Reached by 2020 26.5% 10.1% (19.0%) EFTA01415839 0 2,330.1 1.0 100.0% 92.7% 85.8% 79.5% 73.7% 68.3% 63.3% 58.6% 58.6% 2,006.1 1,885.5 1,768.2 1,656.9 1,551.4 1,451.4 1,356.6 32,679.6 (1,812) (19.0%) 16.6% 18.3% 19.9% 21.6% 23.2% 24.9% 26.5% While our DCF framework for CNI translates to an implied P/E multiple of 15.5x our 2019 EPS estimate. Deutsche Bank Securities Inc. Page 7 EFTA01415840 31 October 2017 Railroads Canadian Rails Figure 12: Our DCF framework for CNI translates to an implied P/E multiple of 15.5x CNI - DCF Framework 2016 Gross Revenue YoY Change (%) EBIT % margin Implied Off BS Interest EBIT + Interest adjustment % margin Cash tax rate (%) Unlevered Free Cash Flow NOPAT % of revenue D&A % of revenue Net Capex % of revenue Working capital Unlevered FCF Time factor Discount factor PV of cash flows Price Target Derivation NPV Net Debt (3Q '17) PV of Operating Leases Implied Equity Value 4Q '18e Sharecount USD/CAD Implied Share Price WACC calculation Historical RFR Equity Risk Premium (ERP) Beta Cost of equity WACD Tax rate Cost of debt % equity % debt WACC Source: Deutsche Bank, Company filings, FactSet 77,786 10,305 516 66,965 EFTA01415841 730 $1.25 2.35% 7.00% 0.92 8.8% 5.4% 26.5% 4.0% 83% 17% 8.0% 5,312 5,350 2017E 2018E 5,863 2019E 12,037 13,067 13,418 14,171 (4.6%) 41 42 6,274 2020E 6,498 2021E 6,727 6,775 2022E 6,968 2023E 7,177 2024E 7,393 2025E 14,596 15,034 15,485 15,950 16,428 16,921 7,614 45 46 47 53 Terminal Value 17,555 8.6% 2.7% 5.6% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.75% 5,682 49 50 52 5,723 5,905 EFTA01415842 6,319 6,544 7,017 7,228 7,445 7,668 7,900 44.1% 43.5% 43.7% 44.3% 45.0% 45.0% 45.0% 45.0% 45.0% 45.0% 45.0% 38 44.4% 43.8% 44.0% 44.6% 44.8% 45.1% 45.3% 45.3% 45.3% 45.3% 45.3% 13.4% 15.0% 15.0% 15.0% 4,633 4,865 1,293 5,020 1,355 5,371 1,431 5,455 1,478 5,536 1,525 5,619 1,573 5,669 1,622 10.2% 9 9% 10.1% 10.1% 10.1% 10.1% 10.2% 10.2% (2,695) (22.4%) (35) (2,700) (20.7%) (24) 0.0 100.0% (2,684) (20.0%) (8) 0.0 (2,834) (20.0%) (18) 1.0 (2,919) (20.0%) 0 1.0 (3,007) (20.0%) 0 1.0 EFTA01415843 (3,097) (20.0%) 0 1.0 5,716 1,673 5,762 1,726 55 7,955 16.6% 18.3% 19.9% 21.6% 23.2% 24.9% 26.5% 5,847 38.5% 37.2% 37.4% 37.9% 37.4% 36.8% 36.3% 35.5% 34.8% 34.1% 33.3% 1,225 (3,190) (20.0%) 0 1.0 10.2% 10.2% 10.2% (3,286) (20.0%) 0 1.0 (3,384) (20.0%) 0 1.0 1,791 (3,511) (20.0%) 0 3,127.9 3,433.6 3,683.1 3,950.7 4,013.7 4,054.1 4,094.6 4,101.0 4,104.0 4,103.6 4,126.8 0.0 1.0 100.0% 92.6% 85.7% 79.4% 73.5% 68.1% 63.0% 58.4% 58.4% 3,658.1 3,441.2 3,218.5 3,009.8 2,791.3 2,586.5 2,394.7 56,686.1 Other Assumptions Revenue growth 2020-2025 Terminal growth rate Terminal D&A (% of rev.) Terminal Capex (% of rev.) $73 <-- Translates to 15.5x our 2019 EPS estimate 3.00% 3.75% Terminal EBIT Margin 45.0% <--Reached by 2022 Terminal Tax Rate 26.5% 10.2% (20.0%) Page 8 EFTA01415844 Deutsche Bank Securities Inc. EFTA01415845 31 October 2017 Railroads Canadian Rails Deutsche Bank Securities Inc. Page 9 Figure 13: DB North American Class I Rail Comparative Valuation Table Current Market Price Price Company (Ticker) CSX Corp. (CSX) Norfolk Southern (NSC) Union Pacific (UNP) Canadian National (CNI) Canadian Pacific (CP) Class I Average Buy 51.60 60 Hold 132.80 Hold 116.37 Sell 81.21 174.74 Buy 135 129 73 209 Valuation Rating 30-Oct Target Methodology Blended Avg. 17.5x 2018 EPS 17.0x 2018 EPS 15.5x 2018 EPS 46,596 NTM Cap Dividend ($mm) Yield 1.8% EPS 2016 2017E 2018E NTM EPS Growth 2016 2017E 2018E YoY YoY YoY Price/Earnings 5-Yr. 2016 2017E 2018E NTM Avg. EV/EBITDA 5-Yr. EFTA01415846 2017E 2018E NTM Avg. 2.21 2.80 2.93 2.91 22.5% 26.6% 4.5% 23.3x 18.4x 17.6x 17.7x 14.3x 9.9x 9.4x 10.1x 8.1x 17.5x 2019 EPS 24,362 Note: we are translating Canadian rail numbers to USD at an exchange rate of USD/CAD of 1.25 Source: Deutsche Bank, Company Filings, FactSet 37,669 2.0% 6.45 6.90 7.75 7.61 14.8% 7.0% 12.2% 20.6x 19.2x 17.1x 17.5x 14.1x 10.0x 9.4x 10.1x 8.3x 88,818 2.5% 5.81 6.69 7.60 7.45 14.6% 15.2% 13.5% 20.0x 17.4x 15.3x 15.6x 15.6x 9.6x 9.0x 9.8x 8.8x 59,924 1.9% 3.92 4.26 4.70 4.63 13.1% 8.9% 10.3% 20.7x 19.0x 17.3x 17.5x 15.6x 12.5x 11.7x 11.8x 10.9x 1.2% 1.9% 8.84 10.48 11.93 11.69 14.0% 18 5% 13.8% 19.8x 16.7x 14.6x 15.0x 16.5x 11.4x 10.5x 10.6x 11.1x 15.8% 15.3% 10.9% 20.9x 18.1x 16.4x 16.7x 15.2x 10.7x 10.0x 10.5x 9.5x EFTA01415847 31 October 2017 Railroads Canadian Rails North American Railroads The North American rail industry is made up of seven Class I rail operators, which are defined as having annual revenues of at least $250M or more in 1991 dollars, or about $475M in current terms, who operate roughly 140,000 of track miles across the continent. In addition, there are over 20 regional railroads and 500 local railroads. Total rail volumes in North America last year totaled roughly 35 million carloads (including intermodal) with 26.1M originating in the U.S. (76%), 6.8M in Canada (20%), and 1.3M in Mexico (4%). The largest contributor to North American rail volumes comes from Intermodal (50%), followed by coal (13%) and agricultural products (9%). Figure 14: North American rail traffic originations by country 76% 4% 20% Source: Deutsche Bank, Association of American Railroads (AAR) U.S. Carloads Canada Carloads Mexico Carloads Figure 15: North American rail traffic by type (exintermodal) Other 4% Nonmetallic minerals 13% Autos & parts 7% Metallic ores & metals 8% Forest products 4% Source: Deutsche Bank, AAR Coal 31% Ag & Food Products 16% Chems & Petroleum 17% The carload breakdown, however, doesn't tell us the full story. While Intermodal EFTA01415848 volumes make up roughly half of North American rail carloads, it makes up just one fifth of revenues for the rails. Figure 16: North American revenue contribution by type Autos 9% Industrial Products 19% Intermodal 20% Coal 11% Other 4% Ag Products 18% Chemicals 19% Source: Company reports (includes CP, CNI, CSX, KSU, NSC, UNP) Page 10 Deutsche Bank Securities Inc. EFTA01415849 31 October 2017 Railroads Canadian Rails Before discussing the revenue model and cost structure dynamics, as well as each commodity group in more detail, we provide a description for each commodity and offer a brief overview of the industry structure: ■ Coal (31% of carloads) — Shipments of coal to power plants and industrial customers. Includes anthracite, bituminous, and lignite coal; ■ Chemicals (17% of carloads) — All chemicals and related products such as sulfuric acid, ethanol and fertilizer. Also includes petroleum and refined oil products, such as crude, gasoline, diesel, asphalt, LPG etc. ■ Agricultural products (16% of carloads) - All food and feed-related products: Grain (wheat, corn, oats, etc.), flour, rice soybean meal, etc. ■ Nonmetallic minerals (13% of carloads) - Crushed stone, sand, gravel, clay and glass products used in construction; Salt used for ice control; ■ Metallic ores and metals (8% of carloads) - Manganese ore, iron ore, copper ore, alumina and nickel ore; Finished steel products and scrap; ■ ■ Autos/Parts (7% of carloads) - Finished automobiles and auto parts; Forest products (4% of carloads) - All lumber and wood products (except furniture), pulp mills, paper, paperboard, containers or boxes, etc. ■ Other (4% of carloads) — all waste and everything not included above Railroad Industry Structure The North American rail industry is split into three classes as defined by the Surface Transportation Board (STB): ■ Class I: Defined as having annual operating revenues of $250M or more in 1991 dollars, which adjusted for inflation, is around $475M today. Class I rails account for just over two-thirds of North American track mileage and consists of seven companies. 1. Burlington Northern Sante Fe (BNSF): Based in Fort Worth, TX, has a network of 32,500 route miles in 28 states and three Canadian provinces. Acquired by Warren Buffet's Berkshire Hathaway in 2010; 2. Union Pacific (UNP): Based in Omaha, NE, UNP links 23 states in the western two-thirds of the U.S. (west of Chicago and New Orleans); 3. CSX Corp (CSX): Based in Jacksonville, FL, about 21,000 route miles serving most of the East Coast including D.C. as well as the Canadian provinces of Ontario and Quebec; 4. Norfolk Southern (NSC): Based in Norfolk, Virginia, operates 20,000 route miles in 22 eastern states and the District of Columbia; 5. Kansas City Southern (KSU): Based in Kansas City, Missouri, EFTA01415850 -6,200 route miles operating in ten central U.S. states with the shortest north/south rail route between Kansas City and ports along the Gulf. Also operates in central and northeastern Mexico. 6. Canadian National (CNI): Based in Montreal, Canada, operates over 20,000 route miles connecting all four major ports in Canada (Prince Rupert, Vancouver, Halifax, & Montreal) to the Gulf of Mexico. It is the largest railway in Canada. Deutsche Bank Securities Inc. Page 11 EFTA01415851 31 October 2017 Railroads Canadian Rails 7. Canadian Pacific (CP): Based in Calgary, Canada, operates approximately 12,500 of route miles spanning six Provinces in Canada and 13 states in the U.S. (serving cities like Minneapolis, Milwaukee, Detroit, Chicago, and New York). ■ Class II & Class III: Defined as railroads with revenues above $40M (Class II) or under $40M (Class III). Together they account for just under onethird of North American track mileage, made up of over 20 regional railroads (Class II) and over 500 local/short line railroads (Class III). Class II railroads connect with Class I carriers for long -haul shipping, and Class III provide rural communities with links to larger networks. Because of their small size short line rails are typically consolidated. Figure 17: Regional exposure by railroad Geographical Footprint Region Northwest U.S. Southwest U.S. Midwest U.S. Southeast U.S. Northeast U.S. Canada Mexico UNP CSX NSC BNSF KSU CP CNI wr wjC 4 iiirc 4 4 4 4 4 wr g g wr g g g g g g g Source: Deutsche Bank, Company filings North American carload and intermodal trends Total NA carloads have declined at a 1.6% CAGR since 2010 reflecting a 7.6% decline in Coal volumes offset by strong growth in sectors like autos/auto parts (+5.7%) and chemicals (+2.9%). Total traffic, which includes carloads and intermodal volumes, has increased at a 0.6% CAGR since 2010 — reflecting a 3 3% growth in intermodal traffic. From a regional standpoint, Mexico has seen the strongest growth with carloads increasing at a 4.4% CAGR amidst strength in autos (+7.2%) and intermodal (+8.3%) while the U.S. has been the laggard. Figure 18: Longer-term movements in NA rail traffic by type 2010-2016 (rebalanced to 100) 100% EFTA01415852 110% 120% 130% 140% 150% 60% 70% 80% 90% Source: Deutsche Bank, AAR Autos/Parts, 139% Intermodal Chems Nonmetallic Forest Ag. Metallic Figure 19: Longer-term movements in NA rail traffic by country 2010-2016 (rebalanced to 100) 100% 105% 110% 115% 120% 125% 130% 135% Coal, 62% 2010 2011 2012 2013 2014 2015 2016 Canada Mexico 129% 101% 110% 104% 2010 2011 2012 2013 2014 2015 2016 U.S. Source: Deutsche Bank, AAR Total NA Page 12 Deutsche Bank Securities Inc. EFTA01415853 31 October 2017 Railroads Canadian Rails More recently, we have seen significant strength in Canada due to strong Intermodal (up 11% YTD), Chemicals (up 9% YTD), and Metals traffic (up 28% YTD). To that point, Canadian carloads are up 9% yoy through 3Q 2017— which compares to up 3% for the U.S. railroads. Figure 20: YoY changes in U.S. vs. Canadian rail traffic Q1 2015 - Q4 2017e 10% 15% 7% (15%) (10%) (5%) 0% 5% 3% 4% 12% 9% 6% 1% 0% U.S. Rails Note: U.S. rails include CSX, NSC, and UNP. Canadian rails include CP and CNI Source: Deutsche Bank, Company filings Canadian Rails Revenue discussion The revenue model for rail companies is fairly straight forward at a high level - volume (typically measured in carloads) x price (measured in average revenue per carload). Total company revenue per carload can depend on various factors, such as mix of volume (different commodities have different price points), length of haul, movements in core/underlying price and fuel surcharges. In 2016, the seven Class I railroads generated roughly $78 billion in revenue This marked an 8% decline from 2015 amidst a 4.2% decline in rail carloads and 4.0% decline revenue per carload (yield). The decrease in rail traffic was largely the result of weaker industrial production activity across North America due to the collapse in commodity prices, looser truckload capacity (hurts domestic intermodal volumes), and the continued shift away from coal dependency (coal carloads down 20% in 2016). The decline in yield, which has a few more moving parts than volumes, was largely the result of lower fuel surcharge revenue ($2.5 billion cumulatively, or $60/carload) and mix headwinds as core-pricing (essentially same-store pricing) was up in the low-single digits across the industry. EFTA01415854 We note 2016 was largely a continuation of downward trends which began in early 2015 and things appeared to bottom in late 2016. To that point, we expect revenue for the industry to be up roughly 6% in 2017 reflecting easy comps, increased industrial production activity, a rebound in coal volumes (largely due to a weaker USD), and increased fuel surcharge revenue. Below we depict Class I revenue trends from 2010-2018E and a market share breakdown by rail. Deutsche Bank Securities Inc. Page 13 YoY Chg. In Carloads Q1 2015 Q2 2015 Q3 2015 Q4 2015 Q1 2016 Q2 2016 Q3 2016 Q4 2016 Q1 2017 Q2 2017 Q3 2017 Q4 2017E EFTA01415855 31 October 2017 Railroads Canadian Rails Figure 21: Class I rail revenue trends 2010-2018E 100,000 40,000 50,000 60,000 70,000 80,000 90,000 13.8% 5.0% 4.6% 7.3% 6.0% 5.2% -5.8% -8.2% (10%) (5%) 0% 5% 10% 15% Total Revenue YoY Chg. Source: Deutsche Bank, Company Reports (we are assuming BNSF & KSU grow in - line w industry) Source: Deutsche Bank, Company reports Figure 22: Class I rail market share by company & region in 2016 NSC 13% CSX 15% CNI 12% CP 6% UNP 26% BNSF 25% KSU 3% U.S./Mexico: 82% Canada: 18% When we look at the rails individually, carload mix plays an important factor visa-vis overall volume growth. For example we note CNI's total traffic increased at a 1.7% CAGR 2010-2016, vs. +1.2% for NSC, +0.2% for CSX, -0.7% for UNP, and -0.9% for CP. We believe the variance largely reflects mix with CNI EFTA01415856 being the least exposed rail to Coal (-7.6% CAGR) while nearly 60% of NSC's 2016 traffic was represented by historically faster growing carload types such as autos/- parts (5.7% CAGR) and intermodal (3.3% CAGR). Additionally, we note that within mix there is a market share component for rails that operate in similar locations which needs to be considered as well. Given the impact mix, fuel and underlying pricing can have on revenue, we focus on underlying traffic growth to gauge organic prospects, and provide comparative exposure based on end market growth trends. We provide more detail on CP and CNI in the company specific sections. Figure 23: Exposure by key end markets for NA railroads under coverage 2016 Exposure Autos/Parts Intermodal Chemicals Ag Products Coal Various Other CAGR (2010-2016) CSX NSC UNP CP 5.7% 3.3% 2.9% (0.1%) (7.6%) 0.1% Source: Deutsche Bank and company filings CNI 7.5% 6.1% 10.2% 4.9% 5.0% CSX 43.6% 53.3% 38.6% 38.7% 41.6% NSC 10.9% 6.6% 12.7% 14.5% 11.5% UNP 7.4% 8.3% 11.6% 19.5% 11.6% CP 13.0% 12.4% 13.8% 12.1% 6.4% CNI 18% 13% 13% 10% 24% CAGR (2010-2016) 0.2% 1.2% (0.7%) (0.9%) 1.7% Page 14 Deutsche Bank Securities Inc. Revenue ($ millions) YoY Change EFTA01415857 31 October 2017 Railroads Canadian Rails Profit and pricing discussion All Class I railroads are solidly profitable in North America, with the companies under our coverage universe (CP, CNI, CSX, NSC, UNP) reporting an average ebit margin of 36.5% in 2016. This makes sense to use given the consolidated nature of the Class I industry, price discipline, and capital intensity needed to maintain an efficient railroad. There is some variability in the performance, however, with the Canadian rails achieving superior profitability relative to the U.S. rails. Figure 24: Class I rail operating ratios (opex as a % of revenue) 2010-2019E 50% 55% 60% 65% 70% 75% 80% 85% CNI CSX NSC UNP Hunter Harrison's Precision Railroading model has helped CP and CNI achieve bestin-class operating ratios. 61.0% 66.3% 60.5% 55.6% 55.7% 2010 2011 2012 2013 2014 2015 2016 2017E 2018E 2019E CP Source: Deutsche Bank, Company reports One driver of profit variation is yield differentials (i.e. revenue per unit or carload). For example, if we look at CNI, which has the lowest operating ratio (OR), 34% of its total traffic in 2016 came from its highest yielding end markets, compared to just 30% for CSX, which had the highest operating ratio of the Class I's in 2016. As you can see below, the companies with a larger percentage of revenue coming from higher yielding carloads typically see a lower operating ratio. While EFTA01415858 this does not entirely account for margin differentials (carloads are not created equal from an incremental margin standpoint and average length of haul needs and # of touch points need to be considered, which we discuss later), higher yields do translate to higher fixed cost leverage. Deutsche Bank Securities Inc. Page 15 EFTA01415859 31 October 2017 Railroads Canadian Rails Figure 25: Rails that generate higher yields (revenue/carload) typically experience better fixed cost leverage Canadian Rails U.S. Rails CNI Forest products Automotive Metals and minerals Intermodal Coal Blended Operating ratio: CP Fertilizers & Sulphur Forest Products Grain Rev/Car % of vol 4,084 8.5% Petroleum and chemicals 3,629 11.5% Top half: Grain and fertilizers 3,485 11.6% 2,908 5.0% 1,303 2,176 55.9% 34% 66% UNP Agricultural Chemicals Industrial Products Automotive 1,509 15.5% Bottom Half Coal 1,316 41.6% 6.4% 100.0% Rev/Car % of vol 4,757 2.4% 34% Energy, chems, & plastics 3,408 9.9% Potash 2,904 4.6% Metals, minerals & consum2,880 7.8% 2,814 4.9% 1,990 12.1% 38.7% Autos Coal EFTA01415860 Intermodal Blended Operating ratio: 1,343 2,400 58.6% 100.0% Bottom Half 66% NSC Chemicals Paper/clay/forest 4,154 2.6% Top half: Ag/consumer/gov't 3,426 17.1% Automotive Intermodal Blended Operating ratio: CSX Chemicals Forest Products Metals & Equipment Ag & Food Products Automotive Coal Fertilizers Minerals Intermodal Blended Note: Canadian Rev/Carload is in C$ Source: Deutsche Bank, Company filings Operating ratio: Intermodal Blended Operating ratio: Rev/Car % of vol 3,699 11.6% Top half: 3,235 12.7% 3,052 13.0% 37% 2,317 10.2% Bottom Half 2,093 13.8% 38.6% 63% 1,139 2,203 63.5% 100.0% Rev/Car % of vol 3,464 6.6% 2,620 3.9% Top half: EFTA01415861 2,575 8.3% 2,213 6.1% 22% Metals and construction 1,847 9.4% Bottom Half Coal 1,649 12.4% 53.3% 573 1,362 68.9% 100.0% Rev/Car % of vol 3,130 10.9% 2,821 4.2% Top half: 2,718 4.0% 2,696 7.4% 2,616 7.5% 2,187 13.0% 1,543 4.7% 1,497 4.8% 43.6% 30% 614 1,716 70.4% 100.0% 78% Bottom Half 70% Against the aforementioned yield characteristics, we note that operating costs per carload averaged about $1,150 in 2016 and were about 15% higher for U.S. rails than Canadian rails. On the revenue side, however, the Canadian rails actually generated an additional $25 per carload, or 1% more than the U.S. rails. In our view, this highlights greater efficiency and longer lengths of haul as well as the aforementioned fixed cost leverage which translates to better margin performance for CP and CNI. Page 16 Deutsche Bank Securities Inc. EFTA01415862 31 October 2017 Railroads Canadian Rails Figure 26: Canadian Rails generated 1% more revenue per carload than the U.S. rails on avg. in 2016... 1,500 1,550 1,600 1,650 1,700 1,750 1,800 1,850 1,900 1,839 1,813 Figure 27: ...and it cost them 15% less to move it 1,000 1,050 1,100 1,150 1,200 1,250 800 850 900 950 U.S. Rails Source: Deutsche Bank, Company filings Canadian Rails 1,216 1,054 2,500 U.S. Rails Source: Deutsche Bank, Company filings Canadian Rails In order to better measure pricing and productivity we compare revenue and cost metrics on a revenue and gross ton mile basis (i.e. RTM and GTM), respectively. Revenue ton miles (RTMs) incorporate both the total weight and distance that goods are transported (vs. carloads which do not include distance or weight) while gross ton miles (GTMs) measure the movement of one ton of freight and/- or rail equipment as well as the distance transported. In our view, these items better reflect the overall workload of a railroad. Generally speaking, the rails with lower operating ratios (better margins), generate less revenue per RTM (likely reflecting longer length of haul and/or heavier mix of cargos) but incur even lower EFTA01415863 costs per GTM on a relative basis. Figure 28: The Canadian rails revenue per RTM is roughly 20% below the U.S. rails 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.052 0.043 0.036 0.045 0.052 Figure 29: But cost per GTM (measure of total workload) is roughly 30% lower than U.S. peers 0.000 0.005 0.010 0.015 0.020 0.025 CNI Source: Deutsche Bank, Company filings CP UNP NSC CSX 0.018 0.015 0.012 0.012 0.019 CNI Source: Deutsche Bank, Company filings CP UNP NSC CSX As shown below, rails with longer lengths of haul (CNI, CP, and UNP) enjoy a more productive workforce as the linehaul portion of transportation typically requires less manpower than the loading/unloading aspect. In addition, this helps improve fuel efficiency as well. Deutsche Bank Securities Inc. Page 17 Revenue/RTM (cents) Opex/GTM (cents) EFTA01415864 31 October 2017 Railroads Canadian Rails Figure 30: Employee productivity, as measured by GTMs per employee (in M) 10 15 20 25 0 5 CNI CP Source: Deutsche Bank, Company filings UNP NSC CSX 19.0 20.1 20.0 14.6 13.3 500 600 700 800 900 CNI Source: Deutsche Bank, Company filings CP Figure 31: Fuel efficiency, as measured by GTMs per gallon of diesel consumed 1,000 1,100 1,200 1,061 1,035 919 880 801 UNP NSC CSX Compensation/labor related costs are the largest component of operating expenses for a railroad - accounting for over 35% of expenses on average. In our view, this highlights the importance of employee productivity as it provides the largest opportunity for margin improvement. When we compare different cost contributions across our coverage universe, this is where CP and CNI see the largest difference relative to the other rails. We believe a large driver behind this is the Precision Railroading model implemented by Hunter Harrison at both CP and CNI (we go into more details in company-specific sections). However, we EFTA01415865 note that Canadian rails benefit from pension differences which optically reduces comp/labor expenses by roughly 3-4% points. We note that this will change on January 1, 2018 as these gains will flow through other income. Figure 32: Employee productivity is the largest source of margin differential for the industry (2016) NSC 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% 5% Comp/Labor Purchased services, materials, other Source: Deutsche Bank, Company reports Fuel D&A Equipment & Other Rents Profit CNI CP UNP CSX Page 18 Deutsche Bank Securities Inc. % of revenue GTM/Employee 17.6% 28.5% 16.2% 19.7% 8.7% 6.4% GTM/gallon of fuel 10.2% 11.8% 3.1% 4.0% 44.1% 29.6% EFTA01415866 31 October 2017 Railroads Canadian Rails In addition to yield, mix, and cost considerations, overall network fluidity also impacts profitability. To this point we note that the major railroads report performance measures weekly such as average train speed and terminal dwell. ■ Speed/Velocity: Train speed measures the linehaul movement between terminals. The average speed is calculated by dividing train-miles by total hours operated. In general the higher the average train speed the better the network is being run, with high frequency and stop duration translating to lower average speeds and in turn less efficiency (all else equal). ■ Dwell: Terminal dwell is the average time a car resides at the specified terminal location expressed in hours. Dwell has averaged 24.5 hours YTD industry-wide, which is slightly above historical levels (23.9 hours in 2014). The underlying geographical characteristics of a rail footprint can have a major impact on the aforementioned profitability and service elements. Both CNI and CP are transcontinental railroads accounting for 75% of Canada's railway tracks, while the United States railroad system is more fragmented, with over 610 total freight railroads operating across roughly 140,000 miles of track. The U.S. is also more densely populated on the East Coast with seven of the largest metropolitan areas located east of the Mississippi River. This not only results in a shorter length of haul for both NSC and CSX, but introduces truck as an intermodal competitor as well. Conversely, Canada has a lower population density (4 people per square kilometer vs. 35 in the United States according the world bank) and major population centers that are more spread out on both the east and west coast. While the distance between major population hubs may improve length of haul for certain commodities for the Canadian rails, there are certain other geographical challenges imposed by the unique terrain and climate of Canada compared to the U.S. The Canadian prairies, which stretch across much of the three of the Western provinces (Alberta, Saskatchewan, and Manitoba), which make up a significant portion of the transcontinental route, are relatively flat and as such have been converted into cropland. However, there are still pockets of difficult route portions, such as Field Hill and formerly Big Hill for CP, which has EFTA01415867 now been converted into the Spiral Tunnels. For railroads, a wider variation in gradient and curvature of tracks can result in higher maintenance and fuel costs. To this point, a study by Oliver Wyman in 2012 found that Canadian Pacific's network has steeper grades and more track curvature than CNI, which the consulting group concluded would require an additional 203 main line AC locomotives than CNI, resulting in increased depreciation, fuel, and maintenance costs (though this did not actually turn out to be the case). Cash flow and balance sheet discussion The very strong return profile allows railroad companies to generate significant operating cash flows. A disproportionate amount of this needed to be allocated to significant capex investments for network upgrades and expansion, new locomotives to meet stricter emission standards, as well as other regulations (particular in the U.S.). To these points we note that the Class I railroads under our coverage universe generated nearly $170B in operating cash flow cumulatively over the last ten years (2006-2016), compared to $235B in ebitda over the same period. Free cash flow totaled $65B (62% of net income), as capex totaled $103B and averaged 17.2% of sales. Capex as a % of sales has averaged closer to 20% of Deutsche Bank Securities Inc. Page 19 EFTA01415868 31 October 2017 Railroads Canadian Rails revenue over the last five years, however, driving weaker net income conversion during this time. The vast majority of free cash flow has been used to repurchase stock, with average share counts down about 15% since 2010 (3% CAGR). Figure 33: Cash Flow Metrics - U.S. Rails (CSX, NSC, UNP cumulative) 2005-16 10,000 12,000 14,000 16,000 2,000 4,000 6,000 8,000 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 U.S. Operating Cash Flow U.S. Free Cash Flow Source: Deutsche Bank, Company filings FCF as a % of NI 0% 20% 40% 60% 80% 100% 120% Figure 34: Cash Flow Metrics - Canadian Rails (CNI & CP cumulative) 2005-16 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Operating Cash Flow Free Cash Flow Source: Deutsche Bank, Company filings FCF as a % of NI 0% 10% 20% 30% EFTA01415869 40% 50% 60% 70% 80% 90% Partly driving the significant capex investment cycle was heavy investment in new locomotives to handle growing energy-related volumes (prior to the decline in oil prices) and Tier 4 emissions requirements which came into effect in 2015. Significant investments have also been made