Oh, boy. Not a Big Boy. Just an "Oh, boy". UP wants to buy NS.

Not this

This

+

UP wants to pay Norfolk Southern shareholders a total of $85B for a company that's worth $62B at the moment.  

You KNOW I have an opinion on this.  Here it is. 

That $23B that's going to NS shareholders could be spent on actually improving the railroad.  Most of the merger benefits are doable without the merger, although it would take some hard work.

Net result?  At best a delay and distraction from the real work of railroading.  At worst, an integration nightmare that will take years and years just to get back to the starting point.

Here's why.

UP and NS aren't stupid, so why are they doing it?  Here's some of the claims.

Smoothing interchanges.

This is a real benefit.  Lots of east-west traffic doesn't move on the most efficient route.  It moves to get the originating road a prime share of the revenue split.  For example, chemical traffic from Texas to the northeast mostly moves through interchanges to NS and CSX in Illinois. A shorter and faster route though New Orleans exists, but would "short haul" UP and BNSF.  Do you need a merger to do it?  No.  UP and NS could negotiate routes and revenue split to get the "win-win".

Chicago congestion (this is a subset of better routing/blocking/faster merchandise trips)

This is a paper tiger.  Lots of Chicago interchange traffic could find a route to avoid Chicago if two things were true.  

One, the railroads did blocking optimization between roads to find the best routes, some of which would avoid Chicago, some of which would create more through blocks and some of which would reduce single car classification to block swaps in Chicago.  

Two, the alternate routes can handle the traffic.  Just because you can find routes through St. Louis and Kansas City, for example, it doesn't mean those routes are up to the increased traffic.  It might take a good bit of infrastructure and crew hiring to accomplish.  After all, these very assets have been "skinnied down"with PSR over the past decade.  Also, changing how traffic is handled in Chicago can change flow and require changes in yard use.  A hump yard is not a block swap yard.

Once again, this is doable without a merger.  It only takes the railroads to negotiate and do joint planning.  Not easy.  But, doesn't require merger.

Better routing/blocking/faster merchandise trips

Right now, roads tend to optimize their blocking and train plan for their internal trips and routes and only do minimal "quid pro quo" through blocking with each other on the really low hanging fruit. A much better blocking plan is theoretically possible.  A "clean sheet" network redesign using optimization tools (that NS has, at least, or a consultant like Oliver Wyman can supply) could result in meaningful and measurable improvement in the carload merchandise traffic handling. 

Again, this can all be done without a merger.  All it takes is some negotiation and planning.

Reduction in staff /back office personnel

This is real, but with each successive merger, it's diminishing returns.  Exactly how many people work in Finance and HR et. al.?  Not that many.  $23B would pay 500 people for 200 years.

Now, let's check out the other side of the ledger.  The cost.  Or, more precisely the cost of "What could go wrong?"

Every major railroad merger since 1990 has had major issues with service interruptions and congestion.  Sources of the trouble include:

• - IT issues
•     Systems from two roads won't communicate properly or on a timely basis
•     Cutting over new territory from one railroad's system to another fails
•     IT system capacity to handle more data improperly estimated
•     IT systems complexity increases risk when integrating and cutting over.  (i.e. PTC data is reliant on car reporting and crew call data)
•     Complexity and fragility of IT systems is increasing, some due to system age, some is IT evolution
• -Traffic issues
•     Incomplete understanding of traffic flow changes
•     Lack of resources - crew, track, power - to handle changes in traffic flow
•     Incomplete and under resourced understanding of transient effects of traffic flow changes
• -Management
•     Institutional culture changes
•     Insufficient transfer of institutional knowledge
•     Poor employee moral  

The UP-SP merger and CSX-NS Conrail split were examples of where just about everything on this list went wrong.

Every new merger has claimed that "they have learned from the past", yet even the recent, small, simple CP - KCS merger is having problems with system integration.

Digging out from congestion is very difficult for railroads.  Why?  Resources - crew, locomotive  and track.  It takes the least amount of resources to operate smoothly.  To operate in a congested state, requires more resources.  To dig out from congestion, even more.  

A highway provides a simple analogy.  A highway is flowing freely at 60 mph, with 1000 vehicles per lane per hour, Rush hour starts, more vehicles enter the highway.  Traffic builds until it gets to capacity - 2000 vehicles per lane per hour. But, everything is still moving along. Then, a squirrel runs across the road.  A few people brake hard for a few seconds, dropping their speed.  A chain reaction ensues and now there is stop and go traffic with an average speed of 30 mph.  At this speed, capacity is only 1000 vehicles per lane per hour.  But, "demand" is 2000, so the back-up traffic jam continues to grow to the rear and congestion will continue to grow until "demand" gets down below 1000 cars per lane per hour, sometime after rush hour is over.

Highways usually get a good "reset" after rush hour is over.  Airlines get a good reset every day, overnight. Railroads have steady flow 24/7.  There is no "reset" for them.  Instead, they have to have extra crews available, more locomotives and more places for trains to sit in order to survive and recover from congestion.  All of these resources have really long lead times.  It's a year to train a locomotive engineer, a year or more to buy new locomotives and many months to years to add more track.

Railroads used to have quite a bit of extra capacity.  Before deregulation, the process to abandon or sell rail lines was long and complicated and often a "no."  Fewer, smaller railroads provide alternate routes for traffic.  Large variability in service meant keeping a good buffer of crew personnel available.

Deregulation, mergers and better planning tools led to a great reduction in excess capacity.  The ultimate expression of this is Precision Scheduled Railroading, which aims to tune service to a very low resource, steady state.  Over the past 20 years, there have been several service congestion issues on the large railroads due to the inability to keep up with business cycle traffic fluctuations or other service interruptions with the necessary resources.  That issue is not fully resolved.  NS is still recovering from post-pandemic traffic recovery and several house-cleanings at mid and upper management in the past decade.  CSX still not up to speed with disruption from planned tunnel outage in Baltimore, for example.

So, is there any reason to think UP and NS can merge operations with minimal problem?  No.  Absolutely no.  Particularly these two railroads.

1. There is less excess capacity-crew, locomotives, yards - than pre-PSR days.
2. Neither NS nor UP have managed big merger integration without massive congestion.
3. Neither road is any better equipped to manage integration than they were 25 years ago. In fact, they may be worse since data integration is tighter and more complicated in their internal systems...and there are fewer folk to manage it.  NS ground to a complete halt on at least one occasion when a connection between PTC and the UTCS dispatching system got clogged and dragged down the entire IT system.

Okay, so what should UP spend $23B on?  Key factors:

1. Nearly all the operational benefits are available without merger thru marketing deals/alliances.
2. The key to making more money is velocity. Not max speed. Avg speed. Electrify and smart braking system are the way. How about 400 mile crew districts? A trucker can do 500.

Electrification is about $4M per route mile.  $23B gets you 5700 route miles. (using Amtrak's New Haven to Boston cost adjusted for inflation - $4M/route mile)  That's 80% of what's needed to do all of UP's heavy mainlines.  Chicago to Ogden Utah and on to LA and Seattle plus LA to Houston to Chicago (7100 miles).  

• Electrification allows: 
• much greater energy efficiency as power is not restricted to generation by heat engines
• regenerative braking
• much greater HP per locomotive, which allows for faster train acceleration, braking and ability to maintain track speed for longer time.

• A smart, electronic braking and intelligent freight car system could be developed to allow:
•     Faster "restricting" speed as low speed braking distances are cut in half
•     Closer train spacing as braking is based on adhesion and not "worst case" wheel sliding on empty cars
•     Near elimination of defect caused train delays using on-board detection.
•     Near elimination of train handling issues including derailments
•     Near elimination of delays and time to add/drop cars from trains due brake system requirements (pumping air, no "bottling", air brake testing)

Railroad dabbled in electronic braking (ECP) 30 years ago.  The braking performance has been proven.  Now is the time to do a complete redesign using new technology and get a workable system designed, tested, hardened and implemented.

These two technologies can greatly increase velocity and service reliably which increase the value and market reach of the product, and reduce cost as resource requirements and energy consumption go down.

Railroads need to improve their product in order to complete and fit into the 21st Century logistics network that is going to require more velocity, flexibility and lower costs.  The alternative is a glorified "going out of business" sale.  

It's past time to start.  That's what $23B should be doing.