Railroad Forums 

  • High Speed Rail HSR (Houston - DFW Dallas Fort Worth) (FKA Texas Central )

  • Discussion related to Amtrak also known as the National Railroad Passenger Corp.
Discussion related to Amtrak also known as the National Railroad Passenger Corp.

Moderators: GirlOnTheTrain, mtuandrew, Tadman

 #1496035  by Arlington
 
I'm looking to estimate the effect of TC's entry into DAL-HOU on the air carriers.
To do this, I'm going to use the Acela's entrance into WAS-NYC

The "air" market DCA-LGA was down by half in the decade between 2007 and 2017
Using: https://www.transtats.bts.gov/airports.asp?pn=1" onclick="window.open(this.href);return false;
2006 688k
2007 680k
2008 546k (recession begins)
2009 451k
2010 403k
2011 381k
2012 366k
2013 425k
2014 413k
2015 360k
2016 374k
2017 305k

I'm going to say that that's what happens to a short distance city pair when rail gets its act together.
Moreso in D/FW - Houston because, unlike the NEC, TC will be an entirely new entrant, will have higher average speeds, and a smaller gate-to-gate disadvantage vs Air.

Today the annual air market (from BTS) blending 2016 & 2017 numbers
HOU-DAL = 620k on WN
IAH-DFW = 555k on UA+AA

The entry of TC will add frequencies equal to WNs, but at triple WN's capacity (train =400 seats per departure)
That's the equivalent of 3 NEW SOUTHWESTS being added to the market.

Swag that at 8000 seats per day 220 days per year and 4000 seats on the other 145 days.
That's 2.3m NEW seats per year in a market that, today, has 1.2m passengers.

And like the Chunnel, once the tracks are laid, even if the rails go bankrupt (as the Chunnel did, unable to pay its debt) the rail operators are never going away (as Eurostar has stayed, and basically driven AF/VS/BA out of the LON-PAR local market forever.

My forecast for "Year 3" of TC Operations
HOU-DAL = 300k on WN
IAH-DFW = 200k on UA+AA
HOU-DAL = 1,500k on TC (roughly 1/3 taken by taking half of air's passengers, 2/3 taken from driving)
 #1496055  by eolesen
 
At least you recognize it's heading for bankruptcy at the get-go....

This will be a $20B project by the time it's done. It's already grown from $12B to $15B before the first shovel of dirt has moved... Assuming 1% interest over a 30 year bond, that's a $64M monthly (not annual...) debt payment.

That's over $2M per day that the route would need to generate just to service the debt, and over 8000 seats, that means every seat has to generate $250/day...

The walk-up fare on DAL-HOU hovers around $100 per seat.

You can expect some revenue premium for convenience, but not double the going rate...
Last edited by eolesen on Tue Jan 08, 2019 8:03 am, edited 1 time in total.
 #1496056  by Arlington
 
I keep thinking that College Station isn't going to "happen" until the train station and the University start growing together, kind of the way that it wasn't until DC grew out along the Dulles Toll Road that Dulles became sustainable as a domestic hub. I'm sure that many employers will like the idea of:
- Hire Texas A&M Grads
- Consolidate ops in one office (or back office) at the HSR station
- Have them live and work in Transit Oriented Development around the HSR station
- Let them visit clients via HSR, or "visit home" on weekends

I'm also wondering if they'll do a "backwards h" or sideways T shaped network that ran:
San Antonio-Austin-College Station
 #1496078  by eolesen
 
The moment I graduated from university, the last thing I wanted was to remain on campus. Same thing for my nephew who graduated from A&M...

That's only a sample size of two, but I can't imagine there's a huge population of A&M grads who want to stay in College Station to start their career. What's the incentive to stay behind?
 #1496106  by electricron
 
eolesen wrote:At least you recognize it's heading for bankruptcy at the get-go....

This will be a $20B project by the time it's done. It's already grown from $12B to $15B before the first shovel of dirt has moved... Assuming 1% interest over a 30 year bond, that's a $64M annual debt payment.

That's over $2M per day that the route would need to generate just to service the debt, and over 8000 seats, that means every seat has to generate $250/day...

The walk-up fare on DAL-HOU hovers around $100 per seat.

You can expect some revenue premium for convenience, but not double the going rate...
How did you reach $2 million /day, where is the math?
$2 million x 365 days/year = $730 million/year, significantly more tha $64 million, at more than 1,100%
$730/$64 x 100 = 1,140%

Assuming your estimation that it will cost $20 billion (including interest) to build it, and the bonds will be paid off in 30 instead of 50 years.....
$20,000 million / 30 years = $667 million / year
$20,000 million / 50 years = $400 million / year

But suppose they actually build it for $15 billion (including interest)?
$15,000 million / 30 years = $500 million / year
15,000 million / 50 years = $300 million / year

As anyone should be able to see, the amount of money needed per year varies by total amount borrowed and by over how many years.
Another thing to keep in mind is the infrastructure; land, rails, wayside equipment, and trains will still be around 30, 50, 80 years from now, which can be used for second and third mortgages. As long as the trains are earning and operating profit refinancing later on will be easier. They should not need $500 million / year to run the trains at an operating profit.

Even you can refinance your home and stretch out repaying your debt of home ownership to mortgage banks far longer than 30 years.
 #1496107  by Arlington
 
Airlines go bankrupt when variable costs (fuel) gets too high
Railroads go bankrupt when fixed costs (rails) were too high.
Of the two, the rails are going to be much more naturally territorial. In bankruptcy, whole hubs can be closed and carted/flown away. Once built, the TC is going to be a great business for SOMEBODY, just like the Eurostar was happy to float separately from the Chunnel.

So once the planes and the rails exist, the incentive is to use them, and the bankruptcies are essentially irrelevant to the competitive economics. No travelers boycotted Eurostar to protest the Chunnel's zeroing out of original shareholders or debt-to-equity treatment of its original lenders.

I'd be happy if eolesen would address himself to the offered analogies for how HSR in a sweet spot can drive airlines from the market, and whether/how Southwest would be immune from something like either the halving suffered DCA-LGA or the near-zeroing suffered PAR-LON? Dallas and Houston (in particular) owe much to and stayed loyal to their airlines despite operations in/through bankruptcy.

1) The apparent halving of DCA-LGA air market between 2007 and 2017 (or, if he has figures back to 2003 ish, that's a better "Acela got real" date), and that this came as Amtrak was doubling its market share.
- doesn't this show that a train can win with even apparently-inferior gate-to-gate times
- DCA and LGA both had plenty of FF addicts, and it didn't keep annual volumes from halving.
- isn't the situation in DAL-HOU actually better for the train than Acela's situation?

2) The essentially-complete triumph of Eurostar over AF/VS/BA in LON-PAR being another great example of two huge hubs that devoted substantial equipment between two huge metros that, in the end, declined enough that [many] CDG & LHR slots & gates could be redeployed on Alliance/Open Skies flying. (BA and AF still fly 7x and 5x respectively, basically hub feed, not local, just as I'd expect AA, UA, and WN to do) Eurostar runs 13x/day despite the bankruptcy of the Chunnel. It would be very Japanese, indeed, to throw too much money into rails for too little return as a way of nonetheless boosting train vehicle sales and operations. [Eurostar was 3% of Air-Rail share of PAR-LON in 1994 and was 70% by 1997 and 78% by 2007). PAR-LON was 4m/yr in 1994 and 2m/yr-ish by 2007.
[https://www.ecologique-solidaire.gouv.f ... 008_uk.pdf]

3) Generally, it is also my sense that France used to devote a large number of small aircraft to France's second tier of cities. When I flew to Toulouse and Agen and similar places in the late 1980s and early 90s, they all had flights to Paris, and that this has all dried up (and really dried up before the CDG2 TGV station opened TGV had sucked so many PAR-xxx locals out the market--along each TGV spoke as it opened--that flights to Paris became unsustainable. It is AMS that still has the feed from French regional cities, not CDG.
 #1496139  by eolesen
 
Sorry, Electricon. $64M is debt service per month, not per year. That’s $768M annually, $2.1M daily.

Sure, you could stretch it out to 50 years, and at $500M you drop to $1.37M daily, and $171 per seat.

That’s still without operating costs, and also without taxes (this is private sector, so presumably sales taxes would be applied). All-in, average fare still needs to be over $200 per seat. A realistic load factor of 60% means an average purchased fare of over $330. That. Won’t. Happen.

I’m not going to bother with your comparisons elsewhere, Arlington. The biggest reasons for drops in shorthaul Europe travel aren’t due to Eurostar. It’s taxes and telecommuting. Why does AMS have more shorthaul? They don’t have the APD or it’s equivalent for DE/FR. Nor does it suffer from the delays and gate shortages at LHR or CDG.

Your comparisons to LGADCA are great marketing material, but that can’t cover up the economics. Tell me how this makes money and covers it’s costs. Starting a project by planning to declare it insolvent within a few years isn’t going to generate a lot of interest in the bond market....

The math is challenging to overcome, and that may be a bigger problem going after construction funding. Nobody wants to own a bunch of railroad RoW that can’t be redeployed to commuter or freight. That’s why these boondoggles have always chased government money where accountability and sustainability don’t exist as underwriting criteria.....

Brightline is doing this right. If they fail, the ROW continues to run freight, and the equipment gets sold off.

If TCHSR fails, the RoW and equipment are orphans. N700’s are narrow gauge. You can’t ship them off to Amtrak or Via...
 #1496141  by Arlington
 
eolesen wrote:It's possible some of that state traffic could fall off with HSR, but for corporate travel, I don't know if there's enough incentive to divert traffic away from a 30 minute flight to a 90 minute train ride.
...
My guess is that there would still be an hourly service pattern on DAL-HOU if for no other reason to help feed the international flights out of HOU.
I value your insights into Texas airlines, and would like you to revise/defend the above in light of:
1)"62minute flight to a 82 minute train ride" (or any other fair comparison, like curb-to-curb, door-to-door)

2) Predicting an hourly pattern for WN is a prediction that they'll shrink just 10% (from 19x to 17x). That'd be unprecedented under almost any forecasting. Whether you call it:
- "like" DCA-LGA or PAR-LON
- what happens when the equivalent of 3 new Southwests enter a market,
 #1496146  by electricron
 
Where did 8,000 seats a day come from?
Well, from the number of trains per day and the number of seats in each train.
Have those numbers been published?

Let's guess at the number of trains.
One train departure every half hour between 6 am and 11 pm, is 17 hours x 2 = 34 trains. Double that again because trains are traveling in both directions, we're up to 68 trains a day.
More math follows:
8,000 seats (stated earlier in this thread) / 68 trains = 117 seats per train.
I'm thinking there is going to be 200 to 300 seats per train.
200 seats x 68 trains/day = 13,600 seats/day
250 seats x 68 trains/day = 17,000 seats/day
300 seats x 68 trains/day = 20,400 seats/day

Let's reduce the number of hours per day they run the trains from 17 hours to 14 hours each day, to 6 am to 8 pm. We now have 56 trains per day.
14 x 2 = 28, 28 x 2 = 56
200 seats x 56 trains = 11,200 seats/day
250 seats x 56 trains = 14,000 seats/day
300 seats x 56 trains = 16,800 seats/day

They will be running 8 cars trains, at 40 seats per car that is 320 seats per train.
Let's assume they only run 6 cars trains initially, at 40 seats per car that is 240 seats per train. I might add 40 seats per car is fairly low, they can squeeze more into a car if they wanted to.

Come on, that "8,000 seats per day" number is way off. There's going to be at least two to three times as many seats available per day because each train has more seats.
But availability doesn't mean seats sold.

Texas Central can always increase the number of hours they run the trains, increase the size of the trains by adding cars to each train later, or lower the headways between trains from once every half hour to once every quarter hour without too much difficulty.
 #1496158  by Arlington
 
it was 8000 *round trip* seats per day. 8000 pdew.
TC has said its trains will have 400 seats (Google it)
TC has said it will run every half hour at peak times (just like Southwest) = 20x round trips vs Southwest's 19 round trips
20 x 400 = 8000 round trip seats per day each way.

Don't get hung up rivet-counting. The real competitive numbers are even simpler. WN runs "about 20" round trips on craft with "about 133 seats" implies that whenever you need to make a capacity comparison you can take Southwest and multiply by 3. Southwest can fly 143 or 175 seat planes or throw in another :30 flight (which'd push the radio below 3), but TC could lay on more :30 trips (and push the ratio toward 4). In the end "3" remains a good swag at how many more seats TC can offer than WN.

Meanwhile, I was working on a revenue forecast.
Arlington wrote: Today the annual air market (from BTS) blending 2016 & 2017 numbers
HOU-DAL = 620k on WN
IAH-DFW = 555k on UA+AA
At $100 per ticket, you'd say that today 1.2m passengers spend $120m on tickets

METHOD 1: $ SHARE = SEAT SHARE (STATIC MARKET SIZE)
Arlington wrote:The entry of TC will add frequencies equal to WNs, but at triple WN's capacity (train =400 seats per departure), [NOTE: as long as those ratios hold, it doesn't actually matter how many seats]
That's the equivalent of 3 NEW SOUTHWESTS being added to the market.
On a straight "seats share = passenger share = dollar share" basis, you'd add up all the seats and divvy by seat share. The seat share would look something like
TC = 3
WN = 1
AA = 0.5
UA = 0.5
The simple model is:
TC has 3/5 of seats, so they'd get .6x1.2m = 720,000 passengers
TC has 3/5 of seats, so they'd get .6x120 = $72m at $100 per

Similar methods would say TC has 1/3 of frequencies, so they'd get .33 x120 = $40m
Those are probably good upper and lower bounds on how much TC can take from its air competitors.

METHOD 2: "NEW ENTRANT EFFECT"
The Southwest Effect applies here. TC has the potential to radically lower average fare, grow frequencies by 50% and seats by 150%, and to vastly improve service quality for center-to-center trips (call it the 610-635 market)

The old rule of thumb was fares fall 50% and traffic triples. Under this you'd predict
3.6m annual passengers
$50 avg fares
180m annual dollars

But still allocate these by frequency share or seat share, so:
60% seat share of 3.6m pax = 2.1m pax
33% freq share of 3.6m pax = 1.2m pax

60% seat share of $180m = $90m
33% freq share of $180m = $60m

METHOD 3: SECOND ORDER EFFECTS
With such low prices, but TC being unable to feed a hub, it seems logical to me that:
All 3 Airlines would focus on connecting traffic, kind of like AA's pattern in PHL-DCA (4x) or PHL-LGA (4x) Is that enough to support hourly on WN? Presumably they have connections today riding alongside their 600k, but in the face of a loss of 300k passengers, I'd expect them to join AA & UA in cutting back on local flying

That looks like
TC = 3
WN = .5 (10x)
AA = 0.25 (4x)
UA = 0.25 (4x)

And I'd say that TC would then have
75% seat share of 3.6m pax = 2.7m pax
50% freq share of 3.6m pax = 1.8m pax

75% seat share of $180m = $135m
50% freq share of $180m = $90m

Something like that. TC captures between 1.5m and 3m passengers per year and between $90m and $135m in revenue per year.
 #1496159  by eolesen
 
$135M in revenue, $500 to $700M in debt service, and a minimum of 20-30% on top of that for operating costs (crew, trainsets, maintenance, electricity)...

Sounds like a compelling business case if your intent is going out of business.
 #1496163  by Arlington
 
eolesen wrote:$135M in revenue, $500 to $700M in debt service, and a minimum of 20-30% on top of that for operating costs (crew, trainsets, maintenance, electricity)...
Sounds like a compelling business case if your intent is going out of business.
Consider the rails to be a gift from Japanese state-sponsored investors, just like all of Airbus (and the A380 in particular) is a gift from European taxpayers and somebody, not Eurostar, paid for the Chunnel. Once this big metal exists, deal with its competitive impact. It does not pencil out to say WN would operate hourlies at the same time that TC is operating.
 #1496186  by electricron
 
Again, you are comparing Texas Central vs airlines with airline passengers numbers. You have completely ignored the 300 - 400% more potential passengers that drive between Houston and Dallas every day. Passengers Texas Central can afford to attract better than any airline. Without those potential passengers attracted from the highways included into the mix, your numbers mean NOTHING!

If the trains were to have 400 passengers per train as you suggest, 8,000 round trip passengers is only 20 round trips per day, which means you suggest they only run trains for 10 hours each day, not 12, not 14, or not up to 16 hours a day.
I would expect similar number of hours Amtrak runs on the NEC, the first southbound Acela train departs Boston at 5:05 am while the last southbound Acela train departs Boston at 5:20 pm. The first northbound Acela train departs D.C. at 5:00 am while the last northbound Acela train departs D.C. at 4:00 pm. These trains take almost 7 hours to travel the entire length of the NEC, not Texas Central's 82 minutes. The last southbound Acela train arrives in D.C. at 11:58 pm, the last northbound Acela train arrives in Boston at 10:51 pm.

So Acela trains are running somewhere on the NEC between 5:00 am and 11:58 pm, almost 19 hours every day. Do you really believe Texas Central will run just 20 round trip trains a day 10, 20, or 30 years from now?

I am not even including MBTA, MTA North, NJT, LIRR, SEPTA, or MARC commuter trains, nor Amtrak's regional trains.
 #1496189  by Arlington
 
I'm pretty comfortable staking out a position between electricron's "all your drivers belong to us" and eolesen's "FFs will be unmoved" ;-)

When the "Southwest Effect" /"New Entrant Effect" happens, the usual hypothesis is that those passengers have been recruited to the air market from the vastly-larger "must drive" market, not that the trips were spontaneous fun stimulated from nothing. But the Southwest Effect has rules of thumb precisely because the "before" air market is roughly predictive of what the "after" market will look like, and I think this has been borne out in looking at Acela, TGV, and Eurostar all of which seem to have at least halved air passenger volumes while growing air-rail volume overall, and sometimes drives formerly primo markets down to one-per-hub-bank kinds of frequencies, as if they were a "regular" spoke (Paris-London, Paris-Lyon)

So when I estimate in different ways how the combined air-rail market will grow 2x or 3x compared to today's air market, the 1.2m to 2.4m "new people" in air/rail are essentially a number for "recruited from driving" to air/rail. Another way to think of this is to say that today's air market tells you something about the total number of people willing to cede control of their trip to a common carrier.

EOlesen directs us to the very salient fact that today's average airfare DAL-HOU is low (I assume he's right in pegging it at about only $100 (each way)), which lessens the threat & opportunity of "cutting average fares in half" in order to attract people to a fast-and-fancy scheduled carrier. Cutting airfares in half is simply two times more lucrative when they start at $200 instead of $100.

Air and Rail both have the problem that they aren't point-to-point and can impose a lot of annoying terminal time* compared to point-to-point driving. By 2025--or certainly within the trains bond-repayment period-- I'd also assume that the road warriors will have fully-automated highway cruise for handling the long stretch of I-35, even if it isn't doing neighborhood streets by then.

It'd also help if I-35 bogged down and were paralleled by a toll road and that parking at destinations in DAL and HOU would be expensive. These are obviously great help in framing the value of taking a bus/train WAS-NYC compared to MD-DE-NJ-NY tolls + downtown parking.

*TC will do really well in terminal time, though. With just 10 trains available, It is highly likely that they'd come close to *always* having a train ready for boarding, and that there will be no need for a waiting area: As soon as the train on one side of the platform closes its doors, the cleaners will be finished with the one on the other side will open.

00 First doors open for boarding (maybe a rolling 1 car every 3 minutes as they're cleaned?)
30 Departure
60 College Station
90 (in transit)
120 Arrived 8 minutes ago and now emptied

With 8 trains the 120 = 00 and you need a perfect 38 minute turn all day to sustain :30 frequencies. WIth 10 trains, you've got a spare at both ends and/or 68 Minute turns.
  • 1
  • 8
  • 9
  • 10
  • 11
  • 12
  • 21