registered: Jan. 2006
Econ2018B Rolls Out
We have recently finished and implemented an upgrade to our Econ 2018 model. The first “for the record” application of the new model will be to a second group of Juneau Hub flights. These flights are mostly bush flights with small payloads and will exercise the extension of the model to small aircraft. A new Econ Hub for larger aircraft will be forthcoming. To distinguish the new model from its predecessor, we are referring to it as Econ 2018B. The objectives of the modification were twofold: (1) to extend its applicability to all classes of aircraft and (2) to incorporate a known cost driver which was not included in the original (2018A) model.
Extend to all classes of aircraft. The original model was designed to apply to Cat I aircraft, such as those that would be appropriate to the flights of the Juneau Group 1 assignments. The operating costs provided by the model were based on publicly available data appropriate to aircraft below roughly 30,000 lbs. Maximum Take Off Weight. Likewise, the passenger and cargo revenues were set to be appropriate to a small regional carrier that might be operating light/medium aircraft in a low volume market (e.g., Juneau). It was never intended that this model be applied to aircraft or flights outside of this intended market. Instead, it was a kind of bare bones initial proof of concept model. The underlying methodology of Econ 2018A was simple, straightforward and intuitive. Operating costs (exclusive of fuel) were proportional to aircraft MTOW and operating hours, and revenues were proportional to passenger (or cargo ton) miles. To our surprise, the resulting relationships produced positive margins for well planned and executed flights and returned values which, while not representative of any particular airline’s operations, were generally reasonable, when applied to Cat I aircraft.
The same could not be said for larger (or significantly smaller) aircraft. When applied to small bush aircraft, the model generally failed to produce positive margins. When applied to larger aircraft, Econ 2018A returned unrealistically high operating costs and astronomically high revenues. We could see from the available data that operating costs for larger aircraft did not increase as rapidly with MTOW as they did for the Cat I aircraft, and it was apparent that the passenger revenues predicted for heavier aircraft were uncompetitive. As an example of the latter, the (roughly) $1 per passenger mile we were charging our Juneau customers for a flight to Ketchikan, would make a roundtrip ticket from KJFK to EGLL cost $6000, and that’s for a coach seat. We could charge them that, but we couldn’t begin to fill even a small aircraft capable of trans-Atlantic flight.
Include other known costs. The predicted operating costs for 2018A were derived from data that included all costs directly related to the actual operation of the aircraft: crew, ticketing, gate operations, airport use taxes, and all maintenance. We segregated fuel costs and accounted for them separately, primarily because fuel is something over which our pilots have more direct control. This was fine, as far as it went. However, an airline cannot be run simply by paying its direct aircraft operating costs, even if those include all taxes. An airline needs aircraft. Whether these are purchased new, purchased used, or leased, they constitute a significant cost. So, unless John’s rich uncle Ned died and left us a stable of aircraft not subject to inheritance taxes, we have to provide the aircraft and pay for them out of our recurring revenues. So the new model had to account for these costs and provide the additional revenue required, if even our original Cat I operation were to generate a positive margin and keep John and I employed in our lucrative VA jobs.
The new model. Solving all these problems was not easy; that’s why it has taken about four months to get it done. The new model is substantially more complex than its predecessor and we are not yet 100% certain that we won’t have to make additional minor modifications as issues are encountered. However, we have accomplished the three tasks set out above: (1) modified the operating cost algorithm to make it applicable to aircraft from a Piper Cub to a B747, (2) included the cost of aircraft ownership in the cost terms for every flight and (3) devised a proprietary SPA ticketing algorithm which should produce a competitive ticket (or cargo) price for every flight and still allow us to meet our fuel costs, operating costs, and aircraft ownership costs. This revenue algorithm will not guarantee a positive margin for every flight. Why is this so? Because the revenue (ticket price) is not based directly on operating cost. In other words, we don’t compute the cost of each flight and then price the tickets to ensure a positive margin. If you want to play that game, I suggest you purchase a barrel, a dozen aquarium fish and a shotgun. Instead, we are trying to construct a very simple, yet basically realistic model. In what we laughingly refer to as the RW, not every flight generates a profit. In fact, large passenger airline flights usually generate surprising narrow margins, characteristically less than 5% of revenue. With these thin margins, it makes sense some flights are losers. Why is this so? Chiefly because not every flight is full (of passengers or cargo). An airline operates on a schedule. They schedule flights, program aircraft and crew to these flights and try to fill the flight with both passengers and cargo. You can’t make a profit flying partially empty aircraft, because, if you charge the higher ticket prices required to generate a profit, your competitors will undercut you. At the same time, you can’t stay in business very long if you routinely cancel flights because they didn’t fill. So, you fly some flights at a loss.
I won’t bore you with how all of this works in mathematical terms, but I can give you a sense of the approach. Let’s take the cost of aircraft ownership. A little internet research reveals tables of the cost of new aircraft by aircraft model. Another source provides the MTOW of these various aircraft. If you plot new purchase price vs. MTOW you get a scatter graph which is reasonably well fit by a straight-line which has a slope and an intercept. Not every point lies on this line, but most are pretty close. So, to a reasonable approximation, we know what our various aircraft cost to purchase new. But, you object, we won’t purchase our aircraft new, we’ll lease them. Yes, but somebody had to purchase them new, and that cost has to be reflected in the lease cost. But, you say, I’ll only lease old aircraft and they’ll be cheaper. Yes, but they will cost more to maintain and your passengers won’t like them and you’ll sell fewer tickets and fly more empty seats. Clearly, in a simple model it doesn’t make sense to try to incorporate all of this extraneous detail; it just clutters up the underlying economics without adding anything worthwhile. So, our “approach” was to purchase our aircraft new (back when I joined SPA) on borrowed money and to pay them off on 20-year financing at 5%. To recoup that expense, we straight line amortize the loan and distribute the annual cost over an assumed 1200 hours of annual operation. So, if you make a 12-hour flight in a SPA B779, you will be charged 1% of the annual cost of amortizing the purchase cost of 243.6 M$ at 5%, about $16,078. Now, I can hear our big iron pilots moaning about having to pay off this huge loan, while our bush pilots only have to shell out a few bucks. But remember two things: first your airplane has more seats and is hauling them a longer way, and passenger revenue is proportional to the number of passengers and the distance of the flight. So, you “make it up in volume”, as the expression goes. Seen from the point of view of the individual SPA pilot, the “cost of ownership” term is the rent you pay SPA for the use of its aircraft.
So, finally. When you are selecting an aircraft for an Econ Hub flight assignment, look closely at the maximum seat and cargo (i.e. payload) capacities of your aircraft as they appear in SPAACARS. If these significantly exceed the assigned number of passengers and cargo weight of the selected flight, you are likely to lose money. How much less can they be? I don’t know, because we don’t base the ticket price on the operating and ownership costs, so there is no fixed relationship. If you are attached to a particular aircraft, fly some Econ test flights with various payload fill fractions in that aircraft and see what it takes.
And a last word. Please bear with us here. This took a lot of time and work on our part. It isn’t and never will be perfect. If you see something that doesn’t look right, let us know (politely) and we’ll think about it and, maybe, fix it. This was a challenging project and we both enjoy this kind of work, but in the end it was to give SPA pilots more interesting options. If you don’t like it, that’s fine, just don’t fly the Econ flights. Otherwise, we hope you find the application of the new Econ model interesting, challenging and fun.
John and Mike