Sure…no problem


They are average of both directions…often the numbers are relatively close both directions, but not always.


Yup, you’re on the right track. These are January 2012, and Q1 is Jan/Feb/Mar. But even if we compare these to the actual January-only load factor (I think it was just above 80%) doing a simple average of these won’t get you to the same number. There are a couple of reasons for this:

(1) The overall reported Load Factor is a weighted average
The stats I show are simply the percentage of seats occupied on each route. The overall load factor they report is based on total revenue passenger miles divided by total available seat miles. Let’s say they only flew two markets: Denver to San Francisco and Sioux Falls. Let’s say DEN-SFO was 80% and DEN-FSD was 60%. One might think the overall load factor would be 70% since that’s the average of 60% and 80%. But DEN-SFO is twice as long as DEN-FSD. So if you factor that in, the overall load factor would be about 73.3%. Then remember than DEN-FSD is only once daily with a 99-seat plane, while DEN-SFO might be 5x/day with 138-seat planes for 690 daily seats. That also makes DEN-SFO “weigh” much more than DEN-FSD – factoring that in, the load factor for our simple airline would be 78.7%.

(2) The overall reported Load Factor includes international flights, which are not reported here.

(3) These numbers don’t differentiate between mainline and EJet, and I’m not positive if the reported “Frontier” numbers for load factor include the Republic EJets or not.

(4) These numbers may or may not include non-revenue passengers onboard. (I’ve tried to nail that fact down over the years with stats from various airlines and have never found it clear and consistent.) Usually the load factor reported by airlines in press releases is (supposed to be) for revenue passengers only.


That’s the $64,000 question, and not an easy one to answer.

It really does rely a great deal on fare and traffic composition. Really that boils down to “yield”, but I like to point out both fare and traffic composition because both are highly important, and people tend to think of “yield” as fare…perhaps the fare the see advertised, or the anecdotal fare information they happen to get from shopping. Traffic composition is important because it tells us not to just rely on what we (think) we know about the local fare. Let me illustrate it this way. Okay, so you see a $79 DEN-PHX fare advertised and you think “yield is horrible”. Or conversely you remember paying less than $100 for DEN-PHX in prior years, and this year the best fare you find for your upcoming vacation is $169, so “yields are strong”. Well, that kind of anecdotal stuff is always dangerous to draw conclusions from. But these days you can sometimes find better data showing *average* fare, which is a lot better than just hunt-and-peck price tallying. Average fare is a much better yardstick, but it’s still missing a big component. Let’s say average fare DEN-PHX is $150 this year, and the average DEN-PHX Airbus has 110 people onboard. Some of those people flew DEN-PHX locally, while others were connecting in on trips like MDW-DEN-PHX, MSP-DEN-PHX, etc. A connecting passenger usually brings quite a bit less revenue than a local passenger (for a few different reasons) and so the financial performance of a DEN-PHX flight booked to 110 with ten local passengers @ $150 is far, far worse than an identical DEN-PHX flight booked to 110 with ninety local passengers @ $150. That’s traffic composition.

So what are profitable load factors? Well, in theory a flight with one person aboard could be profitable if the fare was high enough. Heck, once in awhile I’ve heard people claim that completely empty flights make money because of the high cargo revenue, but I’m not inclined to believe either of those cases happen in the real world on any meaningful basis. In the real world, many (not all) markets have relatively depressed fares because of heavy competition, and that’s certainly true for Frontier.

Markets which *may* have stronger yields include:
--Those with no nonstop competitors
--Those without much low-cost-carrier presence
--Those with mostly local traffic

Markets which *may* have weaker yields include:
--Those which are new
--Those with heavy LCC competition
--Those which rely heavily on connecting traffic

Years back I would have put business markets in the strong section and leisure markets in the weaker section. But the distinction is far less than it once was, especially for an airline like Frontier. Yes, someone like Delta might get $300 average fares for DTW-ROC but $122 average fares for DTW-MCO (those are real Q3 numbers) because Rochester is a mostly-business monopoly route and Orlando is a high volume leisure route with LCC competition. But those distinctions are less than they once were, and much less for an airline like Frontier who can only dream of $300 average fares. And….as I pointed out earlier…the traffic composition is key here as well. If Delta only sells a few seats a day to local Detroit-Rochester passengers at that golden $300 average fare, but they fill the rest of the plane with $99 LAS-DTW-ROC passengers, it’s a bloodbath in spite of that remarkable fare.

Because of fare depression and unknown traffic composition, it’s very hard to tell if a full flight is making money or not. Undoubtedly some of the fullest markets still lost money, and sometimes a high-load market can be a *bad* sign for profitability. Sometimes it signifies that the airline can’t get any pricing traction whatsoever, and is using the $50-is-better-than-$0 theory of revenue management. So it’s no shoe-in that the top markets in loads are the most profitable.

On the other end of the spectrum, however, it’s difficult to see how airlines can make money on the low end of the scale. Yes, in theory somebody could be breaking even with 50% or even fewer of their capacity filled. But there definitely need to be special circumstances for that to be plausible – circumstances that are rare, and perhaps nonexistent with an LCC like Frontier. So there’s a bias to the downside with onboard load stats like this when it comes to guessing financial results. Loads in the 30’s, 40’s, 50’s…you can probably be pretty sure they are money losers. But it’s hard to be sure that *any* route is profitable just from load factor, even at 95% load or higher. The best you can do from these sorts of numbers is say that when they get higher, you can be less confident that they lost money. To me, loads of about 75% and up are less interesting and meaningful when it comes to guessing profits because they are in the broad “maybe / maybe not” range. Loads lower than that are more interesting because you can somewhat more confidently judge a loss when a market is only flying half full or worse.

Loads are not meaningless, but they mean more in context than in simply applying a pass/fail threshold for profitability because you just can’t do that with any confidence.

I enjoy playing with these sorts of stats when I have the time. But they key reason I don’t post these more than I do (especially to more busy boards like airliners.net) is that it’s very easy to draw unsupported conclusions from them, and there are throngs of people willing and able to do just that if given the chance. The discussions tend to devolve pretty quickly. On FT there are fewer participants who hold strong orthodoxy and ax-grinding tendencies, especially on a slower board like F9. So it’s a more manageable discussion.