Books on airline ticket pricing
Thread Starter
Books on airline ticket pricing
Apart from Google and reading the odd paper in a journal, are there any good recommended texts on the economics / maths behind airline ticket pricing ? Many years ago, I did 2 Master's degrees - one in economics and one in maths, so I'm not afraid of a little bit of calculus.
I've seen a few books, but they seem to be from before the time the EU liberalised the passenger air transport market, so are hopelessly out of date. I'm particularly looking for something that is reasonably current - if it doesn't acknowledge LCCs in a significant way, I'd be rather dubious. I know various people (e.g. Simon Calder) have written books about the human angle, and are a fairly gentle read and probably used to sell quite well to the general public, but I'm after something that talks about proper pricing with some maths and, I'm guessing, differential equations.
I've seen a few books, but they seem to be from before the time the EU liberalised the passenger air transport market, so are hopelessly out of date. I'm particularly looking for something that is reasonably current - if it doesn't acknowledge LCCs in a significant way, I'd be rather dubious. I know various people (e.g. Simon Calder) have written books about the human angle, and are a fairly gentle read and probably used to sell quite well to the general public, but I'm after something that talks about proper pricing with some maths and, I'm guessing, differential equations.
Join Date: Sep 2004
Location: Coventry
Age: 48
Posts: 1,946
Likes: 0
Received 0 Likes
on
0 Posts
Would that not be a pretty heavy technical equation, primarily of interest to a few airline managers, but few others.
I can give you this one for air v rail market share, but it doesn't take yield management into account (also see thread in JB on this:
I'd love to see a formula for yield management, I guess based on:
Price p =
£ (base cost + PSC + APD or local eqv) +
a/b X maxmimum fare,
Where a = tickets sold so far and b = tickets available x days remaining.
Even this would need to factor in sales from the previous year, then day of week, known events and other variables.
I hope you find something - if general books don't come to mind, there are books / bios on various airlines and their bosses. In addition to the obvious UK / European ones, Neelemen (Jet Blue) and Herb Kelleher (Southwest) are two fascinating characters, as of course were the post-war pioneers before them. Would possibly add Bethune from Continental in recent times?
I can give you this one for air v rail market share, but it doesn't take yield management into account (also see thread in JB on this:
According to Peter Jorritsma, the rail market share s, as compared to planes, can be computed approximately as a function of the travelling time in minutes t by the formula[37]
s = {1 \over 0.031 \times 1.016^t + 1}
s = {1 \over 0.031 \times 1.016^t + 1}
Price p =
£ (base cost + PSC + APD or local eqv) +
a/b X maxmimum fare,
Where a = tickets sold so far and b = tickets available x days remaining.
Even this would need to factor in sales from the previous year, then day of week, known events and other variables.
I hope you find something - if general books don't come to mind, there are books / bios on various airlines and their bosses. In addition to the obvious UK / European ones, Neelemen (Jet Blue) and Herb Kelleher (Southwest) are two fascinating characters, as of course were the post-war pioneers before them. Would possibly add Bethune from Continental in recent times?
Thread Starter
If there's sufficient demand for an academic journal, About the journal : Journal of Revenue & Pricing Management to exist, then someone with some half decent ability at (non) linear optimisation must have wanted to write a book about it when applied to airlines which doesn't already require a lot of pre-existing knowledge
Join Date: Aug 2007
Location: Derby UK
Age: 59
Posts: 112
Likes: 0
Received 0 Likes
on
0 Posts
The problem is that each pricing model is unique to the airline its created for.
the model is created based on a few fundamental rules which are created first or copied first which determine the principals of the way the fares are calculated.
for example the modern LCC model is based on each seat on each flight being a commodity for sale and that commodity having a cost.
using those basics a model based on booking time before flight can be created with a simple break even load factor incorporated.
This will be unique to each route and unique to the carrier and the all the factors that drive the cost including that horrible nasty fuel price.
so what i'm trying to say is that the info you seek is available as its not specific to airlines, but to any business, and the model can be created which is then specific to the airline based on those principals.
the maths is basic spreadsheet level and even i can do it, don't get confused by jargon such as ASM etc they are only derivatives of your basic cost and availability of commodity data and designed to confuse
.
The economic model is simply supply and demand driven, you want to fly to New York for the cheapest possible fare in the most comfort you can get for the price you are willing to pay on a transport with a finite number of seats having a cost to get you there less than you will pay if shared out amongst everyone on board etc etc.
So the reason its not a highly written about subject is because there really is nothing to it worth going into that isn't in other business pricing texts.
hope this helps.
the model is created based on a few fundamental rules which are created first or copied first which determine the principals of the way the fares are calculated.
for example the modern LCC model is based on each seat on each flight being a commodity for sale and that commodity having a cost.
using those basics a model based on booking time before flight can be created with a simple break even load factor incorporated.
This will be unique to each route and unique to the carrier and the all the factors that drive the cost including that horrible nasty fuel price.
so what i'm trying to say is that the info you seek is available as its not specific to airlines, but to any business, and the model can be created which is then specific to the airline based on those principals.
the maths is basic spreadsheet level and even i can do it, don't get confused by jargon such as ASM etc they are only derivatives of your basic cost and availability of commodity data and designed to confuse
.The economic model is simply supply and demand driven, you want to fly to New York for the cheapest possible fare in the most comfort you can get for the price you are willing to pay on a transport with a finite number of seats having a cost to get you there less than you will pay if shared out amongst everyone on board etc etc.
So the reason its not a highly written about subject is because there really is nothing to it worth going into that isn't in other business pricing texts.
hope this helps.
Join Date: Dec 2012
Location: leeds
Age: 77
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
The book market has changed as much as the air travel market over the last decade. There is the Simon Calder sort of book which finds a general market. Then there is the Rigas Doganis sort of textbook currently version 4 which is mainly aimed at students at Cranfield etc. That is a very good book of its kind. There is no book market any more for the tough stuff, at least not at sector level.
If you want the techie stuff which is not just about pricing itself but its relationship with numbers of carriers, direct v indirect routeing, frequency and capacity, I think your best bet is the journal mentioned above plus Journal of Air Transport Management, Journal of Transport Econ and Policy. Even a very few articles will give you reference lists at the back.
If you view airline pricing as an exercise in Price Differentiation then Louis Phlips's book Price Discrimination is a classic text covering much relevant theory. But it is not specifically about the air travel market.
If you want the techie stuff which is not just about pricing itself but its relationship with numbers of carriers, direct v indirect routeing, frequency and capacity, I think your best bet is the journal mentioned above plus Journal of Air Transport Management, Journal of Transport Econ and Policy. Even a very few articles will give you reference lists at the back.
If you view airline pricing as an exercise in Price Differentiation then Louis Phlips's book Price Discrimination is a classic text covering much relevant theory. But it is not specifically about the air travel market.
Join Date: Dec 2012
Location: leeds
Age: 77
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
Airline pricing would look a bit different if there was an off-peak regulated fare set by the CAA and a route revenue pooling and allocation system for certain ticket types but not others. Those features introduce massive complexity relative to the airline pricing problem which is fundamentally simple but brilliantly executed. Not defending the bearded wonder's rail division BTW, just saying....
Do I remember rightly that routes like BA/AF London to Paris were pooled revenue routes once?
Do I remember rightly that routes like BA/AF London to Paris were pooled revenue routes once?
Join Date: Sep 2004
Location: Coventry
Age: 48
Posts: 1,946
Likes: 0
Received 0 Likes
on
0 Posts
Those features introduce massive complexity relative to the airline pricing problem which is fundamentally simple but brilliantly executed. Not defending the bearded wonder's rail division BTW, just saying....
Hence any book on ticket pricing as a whole should answer the question I have yet to get an answer to, namely why isn't there more yield management in the rail industry? (The walk up fares I can just about get)
Join Date: Dec 2012
Location: leeds
Age: 77
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
I think you need to separate out two different things.
1. Is rail inter-city fare regulation sensible and should ORCATS (the revenua allocation system) be maintained or abolished? Let's just agree (or not) that abolishing regulated fares and seasons would be highly politically controversial. No Tory Govt would do it because of where their heartland is. No Labour Govt would do it because they wouldn't believe in it and Bob Crow would throw a wobbly.
2. Given the existence of some regulated fares, how good a job do individual operators do in setting prices given that this is to some extent a walk up and go sector? I think most of the evidence is that operators price pretty successfully to maximise their revenue within the above constraints. A consequence of that is that the 1730 out of euston may be 50% full and the 1903 120% full, but that is not the operator's fault, it is a consequence of the system and if people don't like that they should be pointed to the ORR. But the pricing fence has to be kept pretty high at the peak times to protect the high value revenue stream. Numerous vox pop TV programmes show the PBI doing their best to manage that on the ground.
Conclusion 1-- you can't use load factors alone to judge performance; yield per pax times load is king.
Conclusion 2--- rail and air cost and demand structures are not the same so comparing 50% LF on one with 70% on the other is fraught with danger.
Given the existence of a range of ticket types covering advance purchase, operator only, time of day/week, across to turn up and go, I wonder what innovative pricing method the operators are missing within the existing regulatory constraints.
1. Is rail inter-city fare regulation sensible and should ORCATS (the revenua allocation system) be maintained or abolished? Let's just agree (or not) that abolishing regulated fares and seasons would be highly politically controversial. No Tory Govt would do it because of where their heartland is. No Labour Govt would do it because they wouldn't believe in it and Bob Crow would throw a wobbly.
2. Given the existence of some regulated fares, how good a job do individual operators do in setting prices given that this is to some extent a walk up and go sector? I think most of the evidence is that operators price pretty successfully to maximise their revenue within the above constraints. A consequence of that is that the 1730 out of euston may be 50% full and the 1903 120% full, but that is not the operator's fault, it is a consequence of the system and if people don't like that they should be pointed to the ORR. But the pricing fence has to be kept pretty high at the peak times to protect the high value revenue stream. Numerous vox pop TV programmes show the PBI doing their best to manage that on the ground.
Conclusion 1-- you can't use load factors alone to judge performance; yield per pax times load is king.
Conclusion 2--- rail and air cost and demand structures are not the same so comparing 50% LF on one with 70% on the other is fraught with danger.
Given the existence of a range of ticket types covering advance purchase, operator only, time of day/week, across to turn up and go, I wonder what innovative pricing method the operators are missing within the existing regulatory constraints.
Join Date: Sep 2004
Location: Coventry
Age: 48
Posts: 1,946
Likes: 0
Received 0 Likes
on
0 Posts
anothertyke - thanks, an excellent answer which gets closer than anyone I've asked within the rail industry ever has!
However,
IIRC, the 19:03 would be full as it is the first of the W Mids off peaks, and yes, Virgin have CHOSEN to extend the peak period by at least an hour in the mornings, pretty sure similar amount in the evenings.
I'm lost now - PBI? Permanent barrier inspectors?
In this respect, why is, as per "Price Discrimination", high yield at risk by selling other tickets on the same service for a lower price, as long as the terms (ie adv purchase) are different?
Agreed - but, but I find it odd that LFs are so low at what is supposed to be the peak times - even with the exclusion of a large amount of the Milton Keynes bunch, who tend to be season ticket holders, hence lower yield!
Again, agreed if you are trying to do like with like, but my point was one is around 70% full, the other typically 70% empty. Yet the TGV, which operates an all booking policy, but over longer distances, has LFs very similar to airlines.
Well in the whole Beardie v First spat, we got a number of suggestions that they were prepared to do, including:
1) Lower peak fares by ~15-20% - so clearly they saw increase yields due to higher LFs.
2) Much stronger use of yield management - which takes us back to the whole question of learning from one of aviation's strongest cards.
3) Conversion of some 1st to 2nd class seats - which again makes better use of under-utilised space, with minimal impact on 1st yields.
So sorry for drifting the thread from air to rail, but I hope it gives some insights into how a similar question is dealt with across different industries, in the lack of an obvious bible on the subject!
However,
A consequence of that is that the 1730 out of euston may be 50% full and the 1903 120% full, but that is not the operator's fault
Numerous vox pop TV programmes show the PBI doing their best to manage that on the ground
In this respect, why is, as per "Price Discrimination", high yield at risk by selling other tickets on the same service for a lower price, as long as the terms (ie adv purchase) are different?
Conclusion 1-- you can't use load factors alone to judge performance; yield per pax times load is king.
Conclusion 2--- rail and air cost and demand structures are not the same so comparing 50% LF on one with 70% on the other is fraught with danger.
I wonder what innovative pricing method the operators are missing within the existing regulatory constraints
1) Lower peak fares by ~15-20% - so clearly they saw increase yields due to higher LFs.
2) Much stronger use of yield management - which takes us back to the whole question of learning from one of aviation's strongest cards.
3) Conversion of some 1st to 2nd class seats - which again makes better use of under-utilised space, with minimal impact on 1st yields.
So sorry for drifting the thread from air to rail, but I hope it gives some insights into how a similar question is dealt with across different industries, in the lack of an obvious bible on the subject!
Join Date: Dec 2012
Location: leeds
Age: 77
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
Thanks for generous reply Jabird.
PBI = poor b****y infantry. Not sure what the PPRUNE rules are! Difficult to board a plane with the wrong ticket. Not difficult to board a train.....
I'm sure if some of the constraints in the problem were relaxed the fare and load profile would change.
PBI = poor b****y infantry. Not sure what the PPRUNE rules are! Difficult to board a plane with the wrong ticket. Not difficult to board a train.....
I'm sure if some of the constraints in the problem were relaxed the fare and load profile would change.
Join Date: Sep 2004
Location: Coventry
Age: 48
Posts: 1,946
Likes: 0
Received 0 Likes
on
0 Posts
Ah yes - I know a guy who works at Paddington and he doesn't take any crap, but pax will try all sorts of things. Lots of room for grey area, esp with delays caused by other TOCs, tube etc.
You could try An Introduction to Air Transport Economics From Theory to Applications (2nd Ed) by Bijan Vasigh, Ken Fleming & Thomas Tacker (Ashgate ISBN 978-1-4094-5487-8). Its written from a US perspective and whilst overall I wouldn't rate it too highly, Chapter 11 on Dynamic Pricing Policy & Revenue Management is very good. Parts of it are fairly mathematical but from what you say that shouldn't be a problem.
http://www.ashgate.com/pdf/catalogue...s-2ed-2013.pdf
http://www.ashgate.com/pdf/catalogue...s-2ed-2013.pdf
Join Date: Sep 2004
Location: Coventry
Age: 48
Posts: 1,946
Likes: 0
Received 0 Likes
on
0 Posts
I saw Herb Kelleher give a talk in London once, and I got to ask him about his pricing policy, and why they didn't "bottom out" the starting price, yet maximum prices for a short sector tended not to go above, iirc (06 prices ish) say $150.
I particularly suggested that, given how much the European LCA's have copied from his model, was there anything he could copy from them?
His response was interesting - firstly, he said his passengers expected to pay a certain price, so charging them too much might get a bum on a seat, but it would depress loyalty.
He then made a comment in response to Ryanair along the lines of "I wouldn't go as far as Michael".
So I think key thing is that even if yield management is a computer system, the airline bosses still get to choose the parameters.
I particularly suggested that, given how much the European LCA's have copied from his model, was there anything he could copy from them?
His response was interesting - firstly, he said his passengers expected to pay a certain price, so charging them too much might get a bum on a seat, but it would depress loyalty.
He then made a comment in response to Ryanair along the lines of "I wouldn't go as far as Michael".
So I think key thing is that even if yield management is a computer system, the airline bosses still get to choose the parameters.
