Amtrak Cuts Operating Losses to Lowest Level in Decades

[jis]

Neroden, you are absolutely right about GAAP. I would also point out that the audited financial statement of a company is not exactly the document that one looks at to make tactical and even some strategic decision. You require finer details than what you will ever find in a GAAP financial statement, to effectively run a company.

Amtrak is also stuck in this peculiar position that they have to report stuff in a particular way specified by the FRA, and FRA is much worse at specifying how to report finances than it is in specifying safety standards apparently.

 

[TiBike]

GAAP is also just about financial accounting – reporting results and status in a standardised way for investors, creditors and tax purposes primarily. There's a whole other realm called managerial accounting, which is not strictly defined. There are standard analytical methods in managerial accounting, but it is a much more flexible discipline that's intended to give managers the insight and tools they need to run a company. GAAP and financial accounting is for external audiences, managerial accounting is for internal purposes, although companies often talk about it publicly.

Much of the criticism of Amtrak's accounting that I've read is about their managerial accounting – how the company tracks and analyses its business results. The operating decisions that Amtrak, or any other company, makes are driven in the first instance by those kinds of analyses. Most of which we never see. But it's completely proper to do so.

It is a mistake to assume that what Amtrak publishes is 1. all the analysis they do, and 2. the basis for any particular decision.

 

[GBNorman]

OK Neroden, I respect your credentials as a financial professional, but take it from this CPA who has held such credential since 1974 (in the railroad industry 70-81; private practice 82-03, post retirement practice 03-17) that any Responsibility Accounting system in practice offends more internal departments than it favors. It simply is the "nature of the beast".

While much, if not the majority, around here enjoy riding LD trains, they naturally would like to have the Amtrak Responsibility Locations (RESLOCS) favor such. Should LD'S simply have "above the rails" costs charged against them in that the underlying Act stated there would be a National System, so the costs of paying the roads to access their Systems, and even that of Station Employees, be charged to a RESLOC identified as Legislation Costs and not against the trains themselves? I guess that's what Andrew ("they make money") Seldon holds. 

But not addressed at this topic are those that hold the Corridor is charged with disproportionate costs. After all, a ride on a Regional is simply "a ride". I guess some still hold an "experiential factor" for same on an LD.

So let's just say that Responsibility Accounting in any industry is going to arrive at a point of "What do you WANT to see, Big Daddy"? It is simply an accounting fact of life.

 

[neroden]

TiBike: yes, I'm talking about managerial accounting.  Which is also what competent investors want to look at.

GN: The correct way to do accounting for Amtrak for business decision purposes is to do NO allocations at ALL. 

If you can actually track an expense to a particular source (as with the "vending machines" for parts at Beech Grove, tracking when a part is used for a particular car), that's great. 

But if you can't, don't pretend that you can.  Fake accounting is worse than no accounting.  The overhead should be separated out from the directly assigned costs.  This is the ONLY way to have decent information for business decision purposes.

This would reveal just how much of running a railroad is, in actual fact, fixed overhead -- it would make the economies of scale obvious.  Since the economies of scale are a major part of the business economics of a railroad, this is very valuable.

And if you have something which isn't really a fixed cost but where you can't allocate it to a particular route... that means you're not tracking your materials or labor use properly.  Go fix your underlying tracking system, so that you know whether those ties were used on the NEC or in Michigan.  Don't "allocate" the cost of the ties between different lines on a fake, artificial basis -- that way lies bad business decisions.

There's no rational way to allocate the costs of Penn Station between different train services -- it's needed if any one service runs.  So don't allocate it. Put it on a "Penn Station operations" line and leave it at that.  Same with every other station.

 

[Trogdor]

TiBike: yes, I'm talking about managerial accounting.  Which is also what competent investors want to look at.

GN: The correct way to do accounting for Amtrak for business decision purposes is to do NO allocations at ALL. 

If you can actually track an expense to a particular source (as with the "vending machines" for parts at Beech Grove, tracking when a part is used for a particular car), that's great. 

But if you can't, don't pretend that you can.  Fake accounting is worse than no accounting.  The overhead should be separated out from the directly assigned costs.  This is the ONLY way to have decent information for business decision purposes.

This would reveal just how much of running a railroad is, in actual fact, fixed overhead -- it would make the economies of scale obvious.  Since the economies of scale are a major part of the business economics of a railroad, this is very valuable.

And if you have something which isn't really a fixed cost but where you can't allocate it to a particular route... that means you're not tracking your materials or labor use properly.  Go fix your underlying tracking system, so that you know whether those ties were used on the NEC or in Michigan.  Don't "allocate" the cost of the ties between different lines on a fake, artificial basis -- that way lies bad business decisions.

There's no rational way to allocate the costs of Penn Station between different train services -- it's needed if any one service runs.  So don't allocate it. Put it on a "Penn Station operations" line and leave it at that.  Same with every other station.

And this is why I’ve never been able to take anything you say seriously.

If you really believe NYP would cost the same to operate no matter if Amtrak ran one train per day through there or 500, then I’ve got a bridge to sell you (feel free to account for the cost of that bridge however you prefer).

 

[ehbowen]

And this is why I’ve never been able to take anything you say seriously.

If you really believe NYP would cost the same to operate no matter if Amtrak ran one train per day through there or 500, then I’ve got a bridge to sell you (feel free to account for the cost of that bridge however you prefer).

I agree that you need to build in the cost of NYP...but after you have a hard accounting of direct costs. By not breaking out that cost where all can see it, we don't know for sure if you're honestly spreading that cost equitably across all trains or if you're playing shell games to make some trains look better and others look worse.

 

[Devil's Advocate]

If you really believe NYP would cost the same to operate no matter if Amtrak ran one train per day through there or 500, then I’ve got a bridge to sell you (feel free to account for the cost of that bridge however you prefer).

Obviously there is a major difference between operating one single train versus hundreds of trains, but since you cannot operate a real locomotive down a partially completed track or employ a fraction of an employee there will be various points where the station cost plateaus despite adding or removing more services.  At this point NYP traffic is so far beyond the original design goals I'm not sure how you can rationally divide the cost between individual trains in a meaningful manner.

 

[jis]

I think confusion arises from trying to define one magic number that serves several, sometimes contradicting purposes:

1. Tracking actual cost of individual items and revenues derived from individual activities. Often incrementals is a more appropriate measure of such things
   
2. Use to decide what to keep and what to dump
   
3. Report GAAP compliant financial accounts
   

For accurate tracking you need numbers that have as little arbitrary allocations included in the rolling up process as possible.

For deciding what to keep and what to dump it is better to pay more attention to incremental costs and revenues instead of trying artificially lump together random numbers through allocations.

The only purpose served by allocations is to assign costs to essentially arbitrary small set of account line items for reporting rolled up financial results.

The geniuses at Congress and FRA confused these various purposes of numbers and came to the erroneous conclusion that if you could create a magic rolled up cost number for each individual train and compare it with the actual measurable revenue from that train, then you have an artificial decision making mechanism for which trains to axe, and say with a straight face that a scientific method was used, never mind that it was not. The entire purpose of it is obfuscation to serve a heinous purpose, and Congress actually managed to get the supine spineless blokes at FRA to require Amtrak to report these artificial number in the way they are reported. No amount of firing Anderson, or the entire Amtrak management team will fix this. It comes from the top.

 

[neroden]

And this is why I’ve never been able to take anything you say seriously.

If you really believe NYP would cost the same to operate no matter if Amtrak ran one train per day through there or 500, then I’ve got a bridge to sell you (feel free to account for the cost of that bridge however you prefer).

Obviously *part* of the costs would be variable, but 99% of them are basically fixed.

Property taxes are fixed, for example.

The heating costs for Penn would be, actually, substantially higher if there was one train per day than if there were 500.  (Fewer people giving off body heat.)

 

[neroden]

I think confusion arises from trying to define one magic number that serves several, sometimes contradicting purposes:

1. Tracking actual cost of individual items and revenues derived from individual activities. Often incrementals is a more appropriate measure of such things
   
2. Use to decide what to keep and what to dump
   
3. Report GAAP compliant financial accounts
   

For accurate tracking you need numbers that have as little arbitrary allocations included in the rolling up process as possible.

For deciding what to keep and what to dump it is better to pay more attention to incremental costs and revenues instead of trying artificially lump together random numbers through allocations.

The only purpose served by allocations is to assign costs to essentially arbitrary small set of account line items for reporting rolled up financial results.

The geniuses at Congress and FRA confused these various purposes of numbers and came to the erroneous conclusion that if you could create a magic rolled up cost number for each individual train and compare it with the actual measurable revenue from that train, then you have an artificial decision making mechanism for which trains to axe, and say with a straight face that a scientific method was used, never mind that it was not. The entire purpose of it is obfuscation to serve a heinous purpose, and Congress actually managed to get the supine spineless blokes at FRA to require Amtrak to report these artificial number in the way they are reported. No amount of firing Anderson, or the entire Amtrak management team will fix this. It comes from the top.

I disagree.  Congress ALSO required Amtrak to report incremental costs; it's right there in PRIIA.  Amtrak has been illegally refusing to report incremental costs since 2008.  Each FRA report has the lines for incremental costs, they're blank, and they say "Not developed yet".

This scofflaw behavior by Anderson and Gardner, (and Boardman before them), totally ignoring a Congressional mandate, is all on them.  We can go back to Congress and ask them to fire Anderson for not doing what Congress ordered him to, of course.

 

[jis]

The question though is, will that change anything? Afterall, it is the Board that is supposed to ensure that the company obeys all laws. I think the fundamental problem is that Amtrak's very continued existence has depended critically on ignoring some laws (some even pretty ridiculous ones - remember the law to become profitable by year xyz, that got Warrington on the glide path?), and everyone knows it. So the argument boils down to which laws we will ignore today, and whether my ignored laws are more important than yours.

It is just an unholy mess when it comes to laws and Amtrak at that level.

 

[cpotisch]

The heating costs for Penn would be, actually, substantially higher if there was one train per day than if there were 500.  (Fewer people giving off body heat.)

What about the cooling costs in the summer, though?

 

[Mystic River Dragon]

Trains Magazine (January 2019) has an article by Bob Johnston called "Amtrak's Money Mystery." Bob (and Trains in general) is often, in my mind, almost too nice toward Amtrak. But this article is a scathing commentary on management, the accounting system, making cuts being what gets you bonuses as a manager, and just general idiocy (money for snow removal in Miami). (It makes my snarky remarks about NJT look like polite conversation at an afternoon tea party in comparison.)

I did wonder if Trains is super ticked off that Amtrak questioned the accuracy of its reporting recently (journalists and editors hate that) and, if so, the gloves are off and we're in for the journalistic equivalent of a food fight between them?

Anyone else read it? What did you think?

 

[drakejings]

I read the article and was very pleased.  I had breakfast this past spring with two contributors to Trains' articles and featured stories.  Like Mystic, I expressed the point of view that the magazine was less than aggressive in its' judgements on Anderson.  I was told that Trains was stepping lightly, hoping for cooperation from Anderson; that burning bridges would not be the most prudent choice.

Now it appears the gloves are off.  Mr. Anderson has been aloof to the media in general, to politicians, and to the rail press in particular.  His being evasive is counter productive, regardless of policy.

 

[neroden]

The question though is, will that change anything? Afterall, it is the Board that is supposed to ensure that the company obeys all laws. I think the fundamental problem is that Amtrak's very continued existence has depended critically on ignoring some laws (some even pretty ridiculous ones - remember the law to become profitable by year xyz, that got Warrington on the glide path?), and everyone knows it. So the argument boils down to which laws we will ignore today, and whether my ignored laws are more important than yours.

I  believe you're correct.  But when picking and choosing which laws to follow -- I can't think of ANY rational reason to ignore the law telling them to do their avoidable-cost accounting correctly, however.  It seems like flat-out incompetent behavior -- the sort of accounting which drove the "Milwaukee Road" to its unnecessary bankruptcy.  I guess I could see a reason for it back when a lot of the trains did lose significantly money on an avoidable-cost basis, in the 1980s, but that *simply isn't true any more*.  Nearly all the trains are either profitable or cost peanuts on an avoidable-cost basis, and there's absolutely no reason not to make that clear.  Boardman even put it in a presentation to Congress, so he understood this.