1、DISCLOSURE APPENDIX AT THE BACK OF THIS REPORT CONTAINS IMPORTANT DISCLOSURES,ANALYST CERTIFICATIONS,LEGAL ENTITY DISCLOSURE AND THE STATUS OF NON-US ANALYSTS.US Disclosure:Credit Suisse does and seeks to do business with companies covered in its research reports.As a result,investors should be awar
2、e 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.20 June 2019Americas/United StatesEquity ResearchRailroads Revising Rail Estimates SECTOR REVIEWResearch
3、AnalystsAllison M.Landry212 325 3716allison.landrycredit-Samantha Yellen212 325 7451samantha.yellencredit-Brian Wright212 538 1855brian.wrightcredit-Were Not Totally Worried.Yet;Buying Oppty in UNP,Network Cutover Risk at NSLowering Ests for U.S.Rails by 2%on Weaker Vols:We have cut our Q2 and 19-21
4、 estimates by about 2%for U.S.rails on average,as the negative volume delta vs.our initial forecasts currently stands at 4%.Should We Be Worried?Two-Year Stack Tells Us Not to Panic Yet:On a two-year stack the picture doesnt look as bad as the y/y numbers suggest;that said,it is undeniable that the
5、macro has weakened,with risks clearly to the downside for the broader freight market.Truck supply,and of course coal(particularly export),are also obvious concerns.PSR,Pricing Offsets:In the meantime,there are plenty of reasons to view the short term set up as favorable.PSR is driving faster headcou
6、nt declines and the removal of redundant assets.Pricing remains a positive,and we point out a few structural considerations:1)NS has plenty of room to simply mark-to-market;2)worsening congestion at BN(independent of weather)provides runway for a later cyclical acceleration in pricing at UNP;and 3)s
7、hare gains from truck in CSX merch biz(underscoring its improved service offering)should allow it to charge a premium.The Cdn rail pricing envt remains firm and is the strongest we have witnessed(ever).Stks Should Work,So Long As We Dont Flirt with MSD Core Vol Decline:The second derivative trends i
8、n Q3 will be telling;assuming the flooding impact at least starts to subside(will it ever?)we should be able to see a clearer picture of the real underlying trends.If the core number is minus 1-2%,the stocks should work as there is plenty of scope for op profit growth in the MSD+range in 19(which sh
9、ould translate into HSD earnings growthand if nothing else,should make for a pretty favorable setup in 2020).However,core carload declines in the 4-5%+range have historically coincided with sector underperformance(albeit this has only happened 3 times on an annual basis over the last 15 yrs).Sentime
10、nt Still Positive:At this point we think the market has already discounted numbers coming in a bit;it doesnt feel like investors are yet willing to throw in the towel on rails,and overall sentiment remains positive.The refrain we continue to hear is“yeah man,the structural story is just so good;cape
11、x has been coming down;and FCF Conversion is near 100%.Where else can I find that in industrials?Theyre not THAT expensive considering these factors.And I hated trimming my position in UNP last week,but I had to.”Short Term Trades to Consider:We think the relative underperformance at UNP is potentia
12、lly misplaced(PSR execution upside and pricing upside),while the risk at NS related to its network cutover in the next couple of weeks may not be fully discounted.At the same time,we point out that the roughly 1-month underperformance in CN relative to CP(though not fundamentally unwarranted)creates
13、 a potential opportunity.Our rank order preference on a 12-month view is:CSX,UNP,CP,CNI,KSU,NSC.20 June 2019Revising Rail Estimates2Key PointsRail Volumes Continue to Slide;Two-Year Stack Not So BadNorth American rail volumes have been decelerating since 2Q18,turning negative y/y in Q1(Figure 1).We
14、are most of the way through the second quarter,and the y/y decline has worsened to 2.5%y/y from-1.1%in Q1.The comps are unusually difficult,and flooding in Missouri,Kansas and Arkansas has contributed to the year-over-year shortfall.Having said that,it is undeniable that the macro has weakened,with
15、risks clearly to the downside for the broader freight market.Truck supply,and of course coal(particularly export)are also obvious concerns from both a volume and pricing standpoint.On a 2-year stacked basis,however,the picture is not as bleak and we think this is an instructive way to view the curre
16、nt volume trends(Figure 2).Compared with 2Q17,the current quarters carloads are up 1.3%a step up from the+0.6%in Q1 on this same basis.Still,the downward trend is evident.Figure 1:North American Carloads(y/y%chg)Figure 2:North American Carloads(2019 vs 2017)-4.0%-2.0%0.0%2.0%4.0%6.0%8.0%1Q172Q173Q17
17、4Q171Q182Q183Q184Q181Q192Q19Year-over-Year(%Change)-4.0%-2.0%0.0%2.0%4.0%6.0%8.0%10.0%12.0%1Q172Q173Q174Q171Q182Q183Q184Q181Q192Q19Year-over-Year(%Change)Source:AAR dataSource:AAR dataIn Figures 3-6,we show rail volumes on a two-year stack basis for aggregate industry carloads;intermodal;coal;and to
18、tal less intermodal,coal and Ag(e.g.a proxy for core industrial volumes).Figure 3:Total North American Carloads(000s)Figure 4:Intermodal Carloads(000s)8,0008,2008,4008,6008,8009,0009,2009,4001Q172Q173Q174Q171Q182Q183Q184Q181Q192Q193,9004,0004,1004,2004,3004,4004,5004,6004,7001Q172Q173Q174Q171Q182Q18
19、3Q184Q181Q192Q19Source:AAR dataSource:AAR data20 June 2019Revising Rail Estimates3Figure 5:Coal Carloads(000s)Figure 6:Total excl Intermodal,Coal and Ag(000s)1,0001,0501,1001,1501,2001,2501,3001Q172Q173Q174Q171Q182Q183Q184Q181Q192Q196,0006,1006,2006,3006,4006,5006,6006,7006,8006,9007,0007,1001Q172Q1
20、73Q174Q171Q182Q183Q184Q181Q192Q19Source:AAR dataSource:AAR dataCP Tracking Above Our Prior Q2 Estimate;KSU Had Largest Negative DeltaPrior to revising our estimates,CP had been tracking 1%above our Q2 estimate.Meanwhile,the U.S.rails were tracking below our forecasts.KSU had the largest negative del
21、ta(5%),followed by UNP at 5%.CSX and NSC 2QTD carloads were both 3%below our prior estimates.Figure 7:Delta Between 2QTD Carloads and CS Estimates-6%-5%-4%-3%-2%-1%0%1%2%KSUUNPNSCCSXCNICPSource:Company data,Credit Suisse estimatesHeadcount Declines Accelerating from PSRGiven PSR initiatives at the p
22、ublic Class I rails(CSX,UNP,NSC),headcount has been trending downward(see Figure 8 through Figure 11).UNPs headcount declined meaningfully in May(-6.2%y/y),followed by CSX at-5.2%y/y and NSC at-4.4%y/y.Importantly,headcount reductions have been accelerating at both NSC and UNP.At NSC,headcount was d
23、own by 4.4%y/y in May,following-2.0%y/y in April and+0.6%y/y in March.Similarly at UNP,headcount came down by 6.2%y/y in May,following-4.8%y/y in April and-3.6%y/y in March.20 June 2019Revising Rail Estimates4Figure 8:CSX Monthly Headcount Y/YFigure 9:NSC Monthly Headcount Y/Y(16.0%)(14.0%)(12.0%)(1
24、0.0%)(8.0%)(6.0%)(4.0%)(2.0%)0.0%Jan-18Feb-18Mar-18Apr-18May-18Jun-18Jul-18Aug-18Sep-18Oct-18Nov-18Dec-18Jan-19Feb-19Mar-19Apr-19May-19(5.0%)(4.0%)(3.0%)(2.0%)(1.0%)0.0%1.0%2.0%Jan-18Feb-18Mar-18Apr-18May-18Jun-18Jul-18Aug-18Sep-18Oct-18Nov-18Dec-18Jan-19Feb-19Mar-19Apr-19May-19Source:Surface Transp
25、ortation Board dataSource:Surface Transportation Board dataFigure 10:UNP Monthly Headcount Y/YFigure 11:BNSF Monthly Headcount Y/Y(7.0%)(6.0%)(5.0%)(4.0%)(3.0%)(2.0%)(1.0%)0.0%1.0%2.0%Jan-18Feb-18Mar-18Apr-18May-18Jun-18Jul-18Aug-18Sep-18Oct-18Nov-18Dec-18Jan-19Feb-19Mar-19Apr-19May-19(1.0%)0.0%1.0%
26、2.0%3.0%4.0%5.0%6.0%7.0%8.0%9.0%Jan-18Feb-18Mar-18Apr-18May-18Jun-18Jul-18Aug-18Sep-18Oct-18Nov-18Dec-18Jan-19Feb-19Mar-19Apr-19May-19Source:Surface Transportation Board dataSource:Surface Transportation Board data20 June 2019Revising Rail Estimates5Worsening BN Congestion a Positive Indicator for U
27、NP PricingWhile severe winter weather and flooding in Q2 have significantly impacted both UNP and BNSFs network performance metrics(and in particular,train speed),we think a clear pattern has emerged at BN whereby service has degraded over the last several quarters(see Figure 12 through Figure 17).I
28、ndeed,trends in both terminal dwell and cars online both important metrics for network fluidity stand in stark contrast to the improvement we have witnessed at UNP since it announced the implementation of PSR in 3Q18.What this tells us is that after having been aggressive with respect to market shar
29、e gains over the course of the last 4-5 years(following substantial investments in network capacity),BNSFs network is likely full and congested.Figure 12:UNP Train SpeedFigure 13:BNSF Train Speed-8.0%-7.0%-6.0%-5.0%-4.0%-3.0%-2.0%-1.0%0.0%1Q172Q173Q174Q171Q182Q183Q184Q181Q192QTDHigher is better-14.0
30、%-12.0%-10.0%-8.0%-6.0%-4.0%-2.0%0.0%2.0%4.0%1Q172Q173Q174Q171Q182Q183Q184Q181Q192QTDHigher is betterSource:Company data,Credit Suisse estimatesSource:Company data,Credit Suisse estimatesFigure 14:UNP Terminal DwellFigure 15:BNSF Terminal Dwell-25.0%-20.0%-15.0%-10.0%-5.0%0.0%5.0%10.0%15.0%1Q172Q173
31、Q174Q171Q182Q183Q184Q181Q192QTDLower is better-10.0%-5.0%0.0%5.0%10.0%15.0%1Q172Q173Q174Q171Q182Q183Q184Q181Q192QTDLower is betterSource:Company data,Credit Suisse estimatesSource:Company data,Credit Suisse estimates20 June 2019Revising Rail Estimates6Figure 16:UNP Cars OnlineFigure 17:BNSF Cars Onl
32、ine-4.0%-3.0%-2.0%-1.0%0.0%1.0%2.0%3.0%4.0%5.0%1Q172Q173Q174Q171Q182Q183Q184Q181Q192QTDLower is better-4.0%-2.0%0.0%2.0%4.0%6.0%8.0%10.0%12.0%1Q172Q173Q174Q171Q182Q183Q184Q181Q192QTDLower is betterSource:Company data,Credit Suisse estimatesSource:Company data,Credit Suisse estimatesNSC Network Cutov
33、erLittle Room for ErrorIn early July,NSC will initiate a major network cutover as part of its Top21 implementation.While the company has already started to rollout the new strategy on a much smaller scale,the forthcoming change of operations is material.As such,we think this embeds a degree of execu
34、tion risk that the market might not be fully discounting and this is particularly true given the lack of executives with PSR experience.Specifically,we are concerned about the potential for service/customer disruptions;and while we wouldnt expect anything on par with what we saw at CSX in the summer
35、 of 2017,we think its fair to say that the STB is on notice.Therefore,even minimal shipper discomfort could provoke regulatory scrutiny.And the timing of this is critically important,considering the now active Board slate and robust agenda(see our notes:Highlights from NARS and STB Demurrage Hearing
36、).CSX Merchandise Segment Growth Outpacing NSCAs we can see in Figure 18,below,merchandise volumes are growing at a faster rate at CSX(+1.8%)in Q2 than at NSC(-3.1%).Merchandise traffic has increased y/y for 4 consecutive quarters for CSX which should be favorable for mix.Meanwhile,at NSC,merchandis
37、e carloads have been declining y/y since the beginning of 2019.In Q2,NSC merchandise carloads worsened to-3.1%y/y from-1.3%y/y in Q1 and flat y/y in Q4.20 June 2019Revising Rail Estimates7Figure 18:Difference in Merchandise Carload Growth Y/Y Between CSX and NSC-15.0%-10.0%-5.0%0.0%5.0%10.0%1Q152Q15
38、3Q154Q151Q162Q163Q164Q161Q172Q173Q174Q171Q182Q183Q184Q181Q192Q19Source:Company data,Credit Suisse estimatesCP,NSC Have Performed Best Among Rail GroupSince the beginning of 2019,NSC and CP have reflected the strongest performance among the rail group.Meanwhile,KSU has lagged its Class I peers.Figure
39、 19:Rail YTD Stock Performance;NSC Continues to Outperform U.S.RailsFigure 20:CP has Outperformed CN Significantly in last 1-2 months 9510010511011512012513013514031-Dec7-Jan14-Jan21-Jan28-Jan4-Feb11-Feb18-Feb25-Feb4-Mar11-Mar18-Mar25-Mar1-Apr8-Apr15-Apr22-Apr29-Apr6-May13-May20-May27-May3-Jun10-Jun
40、17-JunNSCCSXUNPKSUCPCNI9294969810010210410629-Apr6-May13-May20-May27-May3-Jun10-Jun17-JunCPCNISource:Company data,Credit Suisse estimatesSource:Company data,Credit Suisse estimatesCSX merchandisegrowth outpacing NSC20 June 2019Revising Rail Estimates8Service PerformanceIn Figure 21 and Figure 22,we
41、show the 2QTD train speed and terminal dwell for each of the Class I rails.NSCs train speed has improved the most among the rail group(+19.1%y/y),but we note that NSC also has the easiest comps as velocity fell 14.3%y/y in 2Q18(vs.the rail avg.excl.NSC of-3%y/y).This is followed by CSX with+14.8%y/y
42、 improvement in 2QTD despite challenging comps y/y(velocity improved 7.4%y/y in 2Q18).Meanwhile,BNSF and UNPs train speed fell the most in 2QTD by 5.4%y/y and 6.8%y/y,respectively.We note that severe flooding in the Midwest likely contributed to the service deterioration.In terms of terminal dwell,N
43、SC improved by 35.7%y/y,which was the best performance within the group in Q2.Notably,this was followed by UNP with improvement of 12.4%y/y,which may signal improvements related to PSR implementation.This is further underscored by the deterioration in dwell at BNSF by 12.1%y/y meaning that UNP was a
44、ble to keep its assets moving despite weather headwinds,while BNSFs service continued to get worse.Figure 21:Train Speed(Y/Y%Change)Figure 22:Terminal Dwell(Y/Y%Change)-10.0%-5.0%0.0%5.0%10.0%15.0%20.0%25.0%UNPBNSFCNCPKSUCSXNSCHigher is better -40.0%-30.0%-20.0%-10.0%0.0%10.0%20.0%BNSFCPCSXKSUCNUNPN
45、SCLower is betterSource:Company data,Credit Suisse estimatesSource:Company data,Credit Suisse estimatesTrain SpeedIn Figure 23 through Figure 29 we show train speed on a 4-week moving average for each of the Class I carriers.Note that operating efficiencies improve as train speed increases.Figure 23
46、:CN Train SpeedFigure 24:CP Train Speed0.05.010.015.020.025.01471013161922252831343740434649522018201912.014.016.018.020.022.024.026.028.0147101316192225283134374043464952201420152016201720182019Source:Company data,Credit Suisse estimatesSource:Company data,Credit Suisse estimates20 June 2019Revisin
47、g Rail Estimates9Figure 25:CSX Train Speed Figure 26:NSC Train Speed10.012.014.016.018.020.022.024.0147101316192225283134374043464952201620172018201917.018.019.020.021.022.023.024.025.026.0147101316192225283134374043464952201420152016201720182019Source:Company data,Credit Suisse estimatesSource:Comp
48、any data,Credit Suisse estimatesFigure 27:UNP Train SpeedFigure 28:BNSF Train Speed20.021.022.023.024.025.026.027.028.029.030.014710131619222528313437404346495220142015201620172018201920.021.022.023.024.025.026.027.028.029.030.0147101316192225283134374043464952201420152016201720182019Source:Company
49、data,Credit Suisse estimatesSource:Company data,Credit Suisse estimatesFigure 29:KSU Train Speed8.09.010.011.012.013.014.015.0147101316192225283134374043464952201720182019Source:Company data,Credit Suisse estimatesTerminal DwellIn Figure 30 through Figure 36,we show terminal dwell on a 4-week moving
50、 average for each of the Class I carriers.Note that operating efficiencies improve as terminal dwell decreases.20 June 2019Revising Rail Estimates10Figure 30:CN Terminal DwellFigure 31:CP Terminal Dwell0.02.04.06.08.010.012.014.0147101316192225283134374043464952201820194.05.06.07.08.09.010.011.012.0
