Bjorn’s Corner: Aircraft programs

By Bjorn Fehrm

By Bjorn Fehrm

30 October 2015, ©. Leeham Co: There has been dramatic news this week around Bombardier’s (BBD) CSeries program. I wrote a subscribers article about what to expect in terms of the cash flow problem that the BBD management has been wrestling with. The announcements yesterday and the following earnings call confirmed the financial modelling I did with our aircraft modelling tool.

Having watched experienced Wall Street analysts being hard pressed to understand what has happened with the CSeries, I thought I could use this week’s corner to explain the overall economical flow of an aircraft program like the CSeries (there will be details in a follow up subscriber article). I will also put it in context with how it affects a company like BBD and what one must think about when it comes to timing of such projects.

To give the timing aspect more colour, I will also compare with Embraer and their E-Jet E2 project and Boeing’s 787 program. The three programs are very different and they demonstrate in an illustrative way the challenges of making a new civil airliner and that one must adapt the project to the company’s position and its strength and weaknesses.

The dynamics of a new airliner project

An aircraft project has traditionally been divided into two phases, development and production. To understand the project from an economical perspective, this is too simple. We need to divide it into more phases. We take a project the size of the CSeries as example for the money flows involved.

Here a description of the typical phases of such a program:

  1. Initial design or the concept and overall design phase. This is a rather cheap phase. There are “only” several hundreds of highly qualified people working at the OEM and selected sub-suppliers (typically engine suppliers) on defining the overall technical and program aspects of the project. The programs technical and economical performance is developed and the overall business plan made. Exit criteria is design freeze of the aircraft’s overall technical specification with dimensions and loads. There are also other technical and economical dimensioning data generated, but it all lacks detail. You burn “only” in the order of one or several $100m chunks.
  2. Detail design. Now every tiny part of the aircraft and its production apparatus is designed at the OEM and its partners. And we are talking about many parts, around 4 million in a modern airliner. The work is based on the overall design data from the initial design. This requires tens of thousands of people to work at the OEM and its sub suppliers and now the real cash burn starts. Typically up to 1 billion per year for a project the size of CSeries. The exit criteria are the release of production drawings to the manufacturing organizations.
  3. Production of test aircraft and certification. The parts and first aircraft gets produced on the program’s preliminary tooling. Certification gets done while aircraft production gets going in earnest. An expensive phase, a lot of people involved, now also including factory people who develop, test and produce production methods and with them the first aircraft. You burn close to a billion a year.
  4. Aircraft certified and first deliveries start. One would think that the cost would diminish now that most of the development engineers are winding down their work, test rigs go quiet and the focus is on incremental improvement to get the in service reliability above 99%. The cash burn does not go down; it remains on a high level for years to come.
  5. Production ramp. The problem is the cost of making the first aircraft. These are three to five times more expensive than the eventual target production cost and the cost only goes down slowly from these high values. For the aircraft industry, the cost reduces typically 15% every time the produced number of aircraft doubles (an 85% learning curve); i.e. #2 is 15% cheaper to produce and then again #4, #8 and so on. It takes hundreds of aircraft before the costs are to the level where the net price from the customer covers the production cost. All the aircraft produced until then burns company cash.
  6. Cash positive production. When the production costs are lower than the customer net prices, the program starts to generate positive cash flow to the company. This is what should happen for the Airbus A380 this year and for the Boeing 787 end of next year. For the CSeries, the guidance in today’s call was around 2020.

During the Initial phase of the CSeries project (and any other as well), a business plan has been made how to take the program to profitability. Typically this can be at over 500 aircraft and when that many aircraft have been delivered, there shall be an overall positive program result. For the CSeries, the plan was to burn $3.5bn for development, then probably close to the same amount for production and the whole program should break even with what the customers pay somewhere around 400 to 500 aircraft (my guess, I don’t have the number).

The payment of Phases 1-5 is with excess company cash and any additional loans that are taken. Typically existing programs that are in Phase 6 shall contribute as well.  When a program gets delayed, the expensive Phases two and three typically get extended, as it has for CSeries (and 787, A380, A350 and so on, so BBD has good company). Phase 4 will now finish with $6bn on the books in 2016 for the CSeries instead of $3.5bn in 2014 (includes CS300 development).

The problem for BBD was that CSeries was now (because of the delays) one of three big simultaneous development projects. Learjet 85 and Global 7000/8000 were also going through Phases 2-3 over the last several years. The company had increased their loan side from $4.5bn to $9bn to pay for this all and the three times larger Business jet side with their highly profitable Global 5000/6000 contributing something like $30m a copy. At 80 of these per year, the cash flow should have been ample.

Soft business aircraft market and too many cash drains

The problem for BBD was not only that the CSeries (and Learjet 85) cost more, they also slid into a time where the business jet market went soft and BBD needed to reduce the Global 5000/6000 production to 50 a year instead of 80. With parts not stopping coming as fast as customer progress payments, there was suddenly $1bn less cash from the Globals during 2015 than planned.

This was unexpected and coincided with the first CSeries customers not accepting their first aircraft at the end of 2015 (Malmö Aviation) but spring 2016 (Swiss), money from customers would take another 6 months. All-in-all, there were billions of cash missing and BBD management needed to act. They had too many programs stressing the company for cash at the same time and a weakness in those that should have been in their cash producing phase. The long-in-the-tooth CRJ keep the market going but was no cash producer, nor wasn’t the Q-400.

In all this it should be remembered nothing of this was caused by the CSeries being a worse aircraft than planned. If anything, it lives up to what has been promised and a tad more. The problem was that the aircraft’s software side took much longer to stabilize than planned and everything got badly timed in the process.

A project which had a better timing was Boeing’s 787. It more than doubled its development cost and production has clearly been more expensive than planned; yet the Boeing Company has taken it on the chin. Program accounting has helped present nice figures over the years but the real crunch cannot be masked by program accounting, missing cash flows. Yet there has been no cash crisis; why? Because the 737 and 777 have acted like BBD Globals, only much stronger and with much higher numbers.

These have been producing a lot of positive margin when it was needed and now when the 777 morphs into the 777X then the 737 MAX and 787 are asked to take over the show (the MAX is a rather modest change and will be cash positive quickly; the 787 shall produce cash from 2017 if things go to plan).

Embraer also presented their 3Q results this week. Here, one can see that the less challenging E-Jet E2 program is well supported by an E-Jet that is still selling well and being produced with ample margins. It supports both E2 and a build up of their business jet side. The company is only wrestling with the effects of Brazil’s currency, the Real, loosing half its value against the dollar. It makes production work in Brazil cheaper but creates problems with currency effects in many other corners. Overall the result is: “Steady as she goes.”


It is challenging to put a new civil airliner in the market. When one has reached certification, one is only half way through the economical marathon; there is about as much left in cash drain on the company before the project stops consuming company cash.

Before it has paid back all the money it consumed and generated an overall profit, there are many more years needed. In all, a project will be overall positive somewhere close to half its economical life. Only the clever, prudent and strong survive in this business.

19 Comments on “Bjorn’s Corner: Aircraft programs

  1. This is a terrific piece; very enlightening. May I contribute with a fragment of information that is not connected to development/production, but affects program cash flow quite significantly. In an Airbus Global Investor Forum 2014 presentation, Harald Wilhelm, CFO, showed a graphic that illustrates the contracted prices of aircraft sold during the phases of initial design, certification and production (p. 10 of the pdf document with link below). In the illustration, the current prices of the A350XWB is roughly twice the contracted prices during initial design. It may well be a pretty accurate description. A similar general pattern of the 787 program was reported by Jon Ostrower during its ramp-up difficulties (“FlightBlogger”). I believe Greg Smith also showed an informative graphic at the last conference call. In addition, the pricing pattern may explain why there are often order sprees at initial design, and then a dearth of orders during ramp-up.

    This particular syndrome of sorts also affects Bombardier, I believe, and probably not very positively. Though I wish them nothing but the best, needless to say.

  2. Bjorn thank you for this comment but I question two aspects. The first: to read all analyzes on the subject, is to believe that only Bombardier must pay the entire bill developing the CSeries. And how is it treated in outsourcing? Contracts awarded to suppliers under a risk-sharing scheme? The second: What is, in your model, the dynamics of additional economic value of a subsequent development of the CSeries CS500 and CS700 with, for example?

    • I am not sure BBD pays the full bill, I would assume that PW takes development and ramp up costs of the GTF to some extent. For other subsuppliers, can’t see which one it should be, perhaps AVIC pays for its development of the fuselage as a risk sharing partner. Shorts who does the wing is part of BBD.

      The value of further derivatives are not included in the initial business plan. It is non valued potential in a design that has made it to EIS and who has margin for further stretches (which the CSseries has).

  3. Thanks Bjorn. Fabulous as usual.

    One aspect that I thought I have heard is not to underestimate, but of quite some importance, but which you have not specifically mentioned:

    during production uptake, the inventory is growing rapidly and requires cash to be built up. Is this with regard to the other positions neglectable or are these costs nowadays loaded on the suppliers?

    • Hi Chris,

      this is part of what requires cash, the inventory will have to be increased and paid for, this is what ties up capital.

  4. “The long-in-the-tooth CRJ keep the market going but was no cash producer, nor wasn’t the Q-400”

    So Bjorn, you are saying that Bombardier is currently not making money on its established commercial aircrafts?

    Is this because of the small backlog and resulting low production rate, because of competitive price pressure from Embraer and ATR… or both?

    • This is primary because of competitive pressure from Embraer and ATR. The E-Jet has a better cabin cross section to a slightly higher CASM, while the CRJ is Ok for regional haul, passengers finds the E-Jets more comfortable. Re ATR it is at most the same in comfort as a Q400 but cheaper to operate.

  5. Sorry Bjorn , but reading your article, I fail to find something new about the aviation business. Taking Risks is part of every business and Cseries is a wonderful aircraft which the market definitely needs and in my opinion managed to decrease the Business.

    Not every Airline in the world will operate A320/B737-800 or MAX8. Please take into account the fact that A320 is a design of 1988 and Airbus was heavily financed by Europe.

    It is a shame that so many analyst are against Cseries , a plane trying to introduce new technology in the short-haul market …

    • Hi Georgi,

      I don’t understand how you can interpret what I have written as against the CSeries, you will have to do your research better. I have gone on record in several articles that I think the CSeries is a very well composed aircraft, check them during the years. What I have written about now is an explanation of what actually has happened. BBD CEO has said on Canadian TV yesterday that the program was in serious cash trouble and it has therefore acted. I have just explained what is behind these words and how BBD got there. BBD has now described how they plan to get out of the squeeze.

      • Agreed. Well laid out.

        For the C series the irony is that the technical issues are all worked out, its the cash issue that is a real cruncher from here on in.

        Hopefully it gets through all that. I too think its a good bird with a good future if they can get over that cash hump.

  6. Some OEM’s demand all components designed for the aircraft approx. 3000-6000ea on a new commercial airliner free of charge until certification. Many components are derivatives of already existing components and hence are not that very costly for the subs: Honeywell, Moog, Parker, Liebherr, Hamilton Sundstrand etc. They make up for it with during serial production, spare sales and propriatory repairs and licences. As many are classified as LRU’s their reliability has not been like highly analyzed engine LLP’s as you quickly can change them. AW week show how more profitable their businesses are compared to the aircraft OEM’s.

    • In a sense almost all if not all compontents are der4ivatives of someitng preivous.

      However, the latest 3 new aircraft have huge new systems, 787 being the most prominent with its far more electrical system than previous.

      Still they are composite aircraft, engines are all new.

      While poor planning was the base the lack of those pieces seriously impeded the 787 early on.

      737MAx and the E jet are exceptions as they are make overs with new engines so the big new item is the engine. Rest is the same or as noted a minor change.

      777X looks to be an awful lot of new as well. Wing, engines and the fuselage possibly.

  7. The government of the province and also the government of canada won’t let BBD dying there are too much jobs in jeopardy even the minister of enterprise of UK came in montreal. A lot of jobs at Belfast northern ireland have to be protected also.

  8. “For the CSeries, the guidance in today’s call was around 2020.”

    Realistic? With the pricing pressures being applied by Boeing, Airbus and Embraer? It would be interesting to have a peek at the “spreadsheet” where they worked up the 2020 target: production rates, model breakdown, production costs, support costs, and expected selling prices.

    • Its difficult to say for sure when Cseries will be cash +. I suspect the management also only have a rough idea. I believe what we were told is a worst case scenario and what they felt they could justify based on industry experience. I think the hope is to do much better. But with 1B of cash burn not repeating in 2016 (global rate cut) and some very expensive parts of the C series development completed, I can see BBD cash burn fall at a pretty surprising rate. Maybe even cash +…..not by much but + in 2016.

      • Hi Mark,

        this is unrealistic. You should read the pieces about aircraft production and learning curves that I and others have publicized. A good start is the article that you just commented, it looks at an aircraft program in a holistic way with the CSeries as an example. I could just as well have picked Airbus A350, no difference, just the sums involved. Others you find at Javier’s Blog or the IISL, where they look at the 787.

        The science around learning in the industry is around 85 years old and has been proven in every aircraft program since, there is no-one of the OEMs that dispute these principles. They just don’t want to give away their particular numbers.

        BBD even told how much will be needed before the CSeries will be cash positive, $3bn. BBD know the net revenue of each of these aircraft as this is inside what it has already sold, it is all in the contracts with price escalation formulas etc. BBD also knows the production cost within +- 5%, it has sub-supplier contracts on each part with price rules that stretch that far and they all use the learning as part of the formula. It is industry standard.

  9. How much of 15% improvement is relevant to modern manufacturing methods? I faintly remember that this was observed pre-WWII. I guess this aircraft is designed to be manufactured by machines and improvements would be fast initially and slower later so costs should lower to start with and fall much more rapidly. Engine is almost common with 320 so engine cost should be inline with 320? As mentioned in other article, BBD is predicting less that 30 million loss for each of first 50 aircraft, still a lot of money. For A&B NB are cash cows and they are burning cash elsewhere, so I guess it is not easy to sell aircraft cheap even for them just to price BBD out and hopefully a mature CS would be 10% cheaper to operate. And finally a large A&B backlog favors Cseries. Lets see how the things turn out to be.

    • The 787 should not follow the usual 15% improvement, as the structure was made with mostly tape machines things should converge faster. It did not, it is at the 15% curve like the 777 was and so many other aircraft. We are still waiting for the exception from the rule. As CSeries (cleverly) is using mostly known manufacturing methods there is not reason to not apply the 15% rule.

      The engine is not the same one as for A320A (this is larger), it was the first one developed and therefore carry the full learning curve, it shares the core with the one for the MRJ (but not the fan, fan casing and so on).

Leave a Reply

Your email address will not be published. Required fields are marked *