MilesBuzz! - CPA's/Cost Accountants-Help.

Forgetting FASB(do they still call it that) and SEC statements, please enlighten us on your analysis of the actual discounted cash flow marginal cost (of the airline) to provide an award seat. IMHO FTers generally overstate this amount.

IMHO, the marginal cost is the cost of fuel to carry the extra weight, the catering costs (I've seen some of you drink in F), plus any fees the airline picks up...like airport taxes.

Of course, that is the cost and does not include revenue that might have been lost if the seat could have been sold.

Now, I am only a CMA with an MBA, so your mileage may vary.

According to Delta, it's $21.37 per trip. Longer trips cost more in fuel, but some of that $21.37 is fixed per trip, such as check-in and checked luggage transfer. Food cost varies somewhat with trip length.


<< Of course, that is the cost and does not include revenue that might have been lost if the seat could have been sold. >> (Comicwoman)

Yes, Delta's $21.37 excludes opportunity cost. Excluding opportunity cost not only makes it much easier to calculate, you can assume it's practically zero with good yield management.

Southwest Airlines has no capacity controls on award seats, so there is certainly some opportunity cost there, probably more than marginal cost.

Correct me if I'm wrong, but it strikes me that a plane that's going anyway is going to carry a substantially similar fuel load whether it's half full or full. This means that the marginal fuel cost is close to, if not zero. Even if there is some variance in fuel load based on passenger load, that should be somewhat less than that allocable to changes in weather/temperature/cargo load (difference of 100 passengers is probably about 20,000 lbs max against a MTOW of 272,500 for a 757-300 as an example).*

Opportunity cost should also be about zero. That's the purpose of capacity control: the airline doesn't issue award tix on full flights, and it does where there's surplus seat inventory (WN notwithstanding).

This means that we're down to the marginal cost of meals and service (FAs). Most flights, I believe, are going to be staffed about the same even if half full**, so we're down to meal service, which I expect is why Delta's reported number is so low. Does it really cost $21 for that lousy meal? Maybe, if you add a "per service" charge which accounts for some portion of the FAs' time.

Disclaimer. I am not any kind of financial professional, so my observations are simply common sense.

Mike

*I'm not a pilot, so this is just a guess.
**pre-September 11, anyway.

You also have to factor in staffing at the ticket counter and other places within the airport. At first, it may seem that there's no marginal cost for an extra passenger, as those people are going to be there either way. But, that's not really true. Staffing at the airport is based on the expected passenger load for that day. For a hub airport, there could easily be 1000's of award ticket flyers per day. That's certainly going to equate to several extra people needing to be at the ticket counter to check bags.

It's not so much the amount of fuel loaded based on passenger count, but how much of it is burned. More passengers = more fuel burned = more fuel to buy next time. There's no separate fuel gauge for award tickets, but it was burned nonetheless.

As far as labor is concerned, there is still a labor cost even if a particular station isn't staffed extra just for award tickets. Ticket/gate agents are paid by the hour, so more passengers equates to more hours of labor.

Airline yield management is a statistical game. They can shoot for a 95 percent probability that an award seat won't be needed by a last-minute revenue passenger, 99 percent, 99.9 percent or anything else - but it's never 100 percent (absolute certainty.) Especially on popular routes where they allocate a minimum number of award tickets that are snapped up many months in advance, there's a chance they could have sold the seat.

If we assume a 99 percent probability of a seat not being wanted for revenue, the statistical opportunity cost of an award ticket is 1 percent of the lost fare. Depending on the route and all sorts of other things, that's probably in the range of $5-20 for economy awards, higher in the front cabin(s).

Still doesn't matter. Most seats on the plane that could cause the loss of a full fare are bought as discounted round trips, not awards. If the plane does sell out, you can't directly blame an award or a discount purchase for occupying the seat. You could allocate the opportunity cost in many ways; one method is to assign the loss of the full fare in proportion to the award/discount seats given away/sold. BTW "sells out" means all seat inventory sells out, not just an oversale at the gate. If you can buy a full fare, the plane hasn't sold out yet.

So, let's say that 10% of award+discount seats are awards, and 99% of the time, full fare is available at departure. Then you could allocate 10% of 1% of a full fare to the opportunity cost of an award seat; 90% of 1% of a full fare is the opportunity cost of a discount seat.

Let's use $1000 one way for a full fare (i.e., repeat the game for the return trip). Opportunity cost of the discount fare is $1000 * 0.01 * 0.90 = $9 ($18 roundtrip), and the opportunity cost of the award seat is $1000 * 0.01 * 0.10 = $1 ($2 roundtrip). Not very much!

If yield management avoids sell-outs 95% of the time, the opportunity cost of a discount/award is $45 and $5 one way, respectively.

Of course, you could always add sh**loads of seats to inventory and avoid the sell-out problem, but then you invite oversales, with the resulting voluntary bumps (vouchers) and involuntary bumps (cash, angry bumpees and DOT numbers).

Well here is my two cents...

An airline has fixed costs and variable costs in order to operate. As has been mentioned above there are many ways to determine which are fixed and which are variable. It would also depend on how many seats the airline expects to sell to break even.

I would say the real answer is it costs $X to fly the plane from point A to point B. If they are selling lots of discounted seats, they need to sell more seats to break even, if they sell seats at higher fares, the need less seats, so the answer is "it depends".

About 10 years ago I heard a number that said it cost the airliens 7 cents per seat per mile to fly the plane. I am not sure how valid that number is, but there it is.

Of course, it also depends on the flight. If it is a flight that usually is full, such as the United SFO-HNL flight which always looks for volunteers to give us their seats, then it costs a lot to add another free seat to the flight.

So there it is....

Let's look at it this way: Mongolian Airways has been isolated from the world for a long time, and they haven't heard about award travel at all. Some foreign traveller lets the management in on this hot new thing in the west, and soon the airline decides to implement it.

For the sake of simplicity, let's assume that immediately after the announcement, there was an award traveller on 7% of all seats. Let's not consider the acrual of miles or the lengths people go to earn them.

This means that Mongolian Airlines' load factors suddenly rise by 7.5%. Let's consider how this affects the airline's costs.

For fixed landside costs, the effects are small. Unless already at maximum capacity, no significant amount of extra ground personnel are needed.

Because of capacity management and award ticket restrictions, we can assume that load factors rise the most on routes that have excess capacity. There might be great demand for award tickets on tight routes, but their availability won't be good. This means that the airline doesn't need to introduce new routes or services to cater for award travel.

In essence, award travel doesn't affect the number of planes the airline needs to operate -- or the total amount of flight hours for the fleet.

If we go further, this means that extra personnel for flight operations isn't probably needed either. Part time flight attendants might be needed if the airline uses just-in-time staffing of cabins (based on actual load).

This leaves the actual fuel cost of hauling 50-150kg of weight per award seat, catering costs, per transaction reservation costs and airport taxes, if included in the award.

The actual fixed costs of the frequent flyer program are quite possibly a very significant part of the total cost. The frequent flyer program, however, probably causes revenue travel as well, so its costs per award seat are hard to calculate.

if a flt costs the a/l $20 incremently, & you pay 20k mi [half of rt], then it costs the a/l 1/10 cent/ff mi.

The original question was the marginal cost of an award ticket, and as such, I would guess that the answer would depend on route loads. For routes where you usually have empty seats, the posts above cover it. On extremely high load routes, the airline is missing the opportunity to sell a ticket, maybe 70% on the low side, 90+ on the high side (think Hawaii peak season, or TLV over the New Year or Passover/Easter).

If you had the info to break it down that way, my guess is the marginal cost would be much higher than the $20 to $50 range noted above. I think that's why the airlines were required to list the miles as a liability on the balance sheet - I have absolutely no clue what rate they're using but it could answer the question.

There is surely and unknow cost if the flyer would have paid if not for the award he has sitting there. I personally use my tickets instead of paying, with some, but not many for flights that I would not have made. Therefore, it does cost them money, in most cases $300-$600 that I might have paid.

But - who cares anyway and where is this going?

Amen, NoStressHere. Topster, what are you trying to do with this number you're looking for? I can't ever recall seeing the term "actual discounted cash flow marginal cost" in b-school or on the CPA exam. It's sounds like you opened a finance 101 book, pulled a couple of buzz phrases out of it and joined them together. Help us out here...

cubsfan10, say if I have 100,000 mi that I plan to use as follows: one 25,000 mi award seat for the next 4 yrs. The airline has a marginal cost to provide the service, and since there is a time line of 0-4 yrs, discounting the cost is proper. Really, my interest is in gaining a better understanding of the airline's cost to provide this product. Not there yet, but learning thanks to you guys.
For sure, I can appreciate that ff miles are an outstanding example of a value-added product, since the value we consumers place to each mile is far,far greater than the cost to provide the product.

Reading this thread has made me realize that my B.S. in Accountancy is just that!

Chexfan...I hope thate you just got the degree. As a holder of the same one, it took me about 4 months of working to figure that out. http://www.flyertalk.com/forum/smile.gif

Back to my question - who cares what it costs the airline?

To me, 25000 miles are worth around $400, though there have been times where it was worth $800 due to short notice flights.

FWIW, if you accept that there is a close relationship between the cost to the airline of issued miles and the willingmess of the airline to be generous in issuing more miles (double/triple through promotions, extending Elite status, etc..)then it seems to me that knowing the cost provides a basis for understanding the airlines' ff strategies, especially during this period of turmoil.

"actual" = excludes opportunity cost

"discounted cash flow" = present value of future cash flow

"marginal cost" = first difference of cost of award seats = (cost of n+1 award seats) minus (cost of n award seats)


[This message has been edited by JS (edited 10-02-2001).]

the miles are also a liability on the books of the airline. "if all the ff miles were redeamed in the same year...." this is something an audit also should evaluate.

Here's my $0.02.

The only cost to calculate would be the incremental cost. Opportunity cost would not be a quantifiable factor as award seats are on flights that are typically not full. Therefore, the same fixed costs per flight would apply (depreciation, landing fees, labor costs, etc.) regardless of passenger load. Sure, there might be one less flight attendant with a significant reduction in passenger load, but this scenario is based on a single award seat.

So what are the incremental costs that can be identified? Paper luggage tags, boarding passes, meal (if applicable), beverages, cups, pretzels/snack mix, napkins, etc. Fuel costs would also have to be included as fuel is calculated based on the passenger load. Otherwise, I can't think of anything else.

Therefore, the incremental cost of one additional passenger is negligible. The present value of that is immaterial.

Having said all of this, I'm sure that the airlines have a formula to calculate a "cost" for unawarded FF miles. This would be on their balance sheet as a Contingent Liability. It is "contingent" because the miles do have an expiration date. Bear in mind that FF miles can be used for awards other than flight tickets.

I'll wager my CMA certificate that somewhere in the accounting office of each airline lurks a giant spreadsheet. On this spreadsheet would be an alogrithym showing how many miles are outstanding in the program, the typical redemption rate, the typical redemption prize, and an assumed cost per mile. This is then given to the most mole-like accountant where he proceeds to pencil-whip this continuously each day until the end of the fiscal quarter.

Heaven help anyone in the room when he gives a presentation of how he calculated his accrual!!!!

But, again. This is just my $0.02. And, FWIW, I am a CMA with almost 20 years of doing things just like this!

Makes sense to me.

A rough calculation of fuel cost is: 5 percent per hour.

A 200lb person on a 10 hour flight costs 100lbs of fuel.

I beg to differ about ignoring the cost of the flight attendant.

If there is one FA per 40 pax, each passenger (paid or award) carries 1/40 the cost of the FA. On the average, one of every 40 award pax will cause another FA to be added to a flight. Given the large numbers involved, the total cost will be nearly exact even if the chance of a given award passenger causing that to happen is small.

Each FA on a coast-to-coast round trip costs the airline about $1,200, considering how much they're paid, additional overhead costs, and how much of their monthly workload one such trip comprises.

Therefore, each award ticket costs about $30 for the FA - less on shorter trips, of course, as is true for other cost factors as well.

If the airline cuts FA staffing to 1:60, now that they're not serving as much food, the FA cost drops to about $20.

Pilots or F/A's correct me if I'm wrong, but it's my understanding that F/A staffing is dependent upon number of seats on the ship, empty or full. Therefore F/A staffing is a fixed cost.

You are correct, Topster. FAA minimum is one FA for every 50 seats.

A few more thoughts on the overall cost of FF programmes.....
1:Factor in the large amount of FF miles that are never redeemed (exclusing FlyerTalkers of course!) and that reduces the cost of FF miles to the airline.
2:Most of those miles earned other than by flying are sold by the airline to marketing partners...not sure of the rate they sell them at, but the profits from that also defray the cost of FF miles.
3:Use of miles for upgrades or airline club memberships is also a nice one for the airlines. For example, 50k miles for a transatlantic upgrade on effectively a space available basis is 50k miles redeemed at the marginal cost of the difference in meal service and amenities of one class of service. Club membership is even better for the airline, normally valuing miles at a very low figure (<1c per mile).

None of these relate directly to marginal cost or DCF calcs but without the presence of a FF programme the airline couldn't make profits from these types of deals.

As regards DCF, all depends on the average period of time between earning and redemption. My own guess system wide is about two years on average between earning and redemption, so a compounded DCF of <20% on true cost.

JW2CWFWIW (and I'm just a lowly CA)

<< As regards DCF, all depends on the average period of time between earning and redemption. My own guess system wide is about two years on average between earning and redemption, so a compounded DCF of <20% on true cost. >>

Much less than 20%. Assuming interest of 6% and inflation of 3%, that's effectively interest of 3%. A duration of approx. two years is only a 6% discount on today's cost.

BTW Delta already figured in the proportion of outstanding miles actually used for award travel to arrive at total liability = expected trips times $21.37 per trip. I don't know if they discounted expected value of future award tickets for interest, or if they discounted future cost to get $21.37.