NOT QUITE: Financial Analyst Claims “Airlines Make More Money Selling Miles Than Seats”

Justin Bachman has a Bloomberg piece on the business of airlines selling miles. He’s basically covering the financial analyst Joseph DeNardi’s thesis that airlines are undervalued because their mileage programs have huge margins and represent a stable source of income not nearly as cyclical as air transportation.

The airline-miles business, formally known as loyalty programs, has become a high-margin enterprise that’s grown in size and value amid airline consolidation, with carriers keen to expand credit-card rolls and see loyalty members spend more. This year, Alaska Airlines began tying a small percentage of its 19,000 employees’ performance pay to the market growth of its card with Bank of America Corp.

The piece is titled “Airlines Make More Money Selling Miles Than Seats” but while airline mileage sales have higher margins than transportation, and mileage programs at times have been the only profitable piece of airlines, it’s not unambiguously clear that airlines have been making more money selling miles than selling flights in recent years because as the price of fuel has fallen the transportation component of gotten more profitable.

  • American Airlines earned $7.6 billion in 2015, while total revenue (not profit) from mileage sales was $2.25 billion. Selling miles wasn’t more than one-fifth of American’s profits that year.

  • American Airlines earned $2.7 billion in 2016. That’s profit. American’s mileage sales revenue in 2016 was $2.5 billion. American doesn’t disclose their margins, but they booked $750 million as the transportation cost associated with these miles. They may have made $1.75 billion on the mileage program (though it’s likely less, the transportation component of mileage sales isn’t the full cost of the AAdvantage program). It’s possible that AAdvantage was more than half of American’s profit in 2016.

  • While American’s profit fell in 2016, Delta still earned $6.1 billion. Total revenue from SkyMiles didn’t exceed half of the airline’s profit, so the mileage program certainly wasn’t more than half the airline’s profit.

I worked through the financial analyst argument about the value of airlines a week ago, suggesting it’s somewhat overclaimed — especially failing to consider the risk that each airline has basically one major customer for miles sales (their co-brand bank partner) and the future of this revenue stream is highly dependent on maintaining current levels of credit card interchange.

The Bloomberg piece also overstates the price that airlines are selling miles at — 1.5 to 2.5 cents apiece. In fact, taking numbers from American Airlines SEC 10-K filings we can see that on average it’s less than that. Up until the current 10-K, where American stopped breaking out useful information about the AAdvantage program, we could take the disclosed:

  • Total miles issued for the year
  • Percentage of those miles sold to third parties
  • Total revenue from mileage sales

And calculate the average price at which they were selling miles. Total revenue from mileage sales/(Total miles issued * Percentage sold to third parties) = Average price per mile.

That price has ranged from 1.23 to 1.32 cents per mile. Now, most partners are spending a whole lot more than this. However since the largest chunk of miles have historically been sold to Citibank, and Citi is clearly spending less than this, they bring down the average substantially.

Other airlines haven’t been historically as forthcoming about their mileage program as American — except that it’s also clear American has booked the lowest liability for future redemptions of the major US legacy carrier frequent flyer programs. So there’s a reasonable chance with new GAAP accounting rules for loyalty programs being adopted by Delta, United, and American starting in January 2018 American may see the biggest increase in liability as they move away from ‘incremental cost’ accounting to a fair value method of estimating redemption costs.

DeNardi thinks the AAdvantage program is worth over $35 billion — about 3/4ths more than the entire airline is valued by the market including the mileage program. That seems crazy.

  1. That would value AAdvantage at about $350 per member.
  2. No airline frequent flyer program spinoff has valued a program at over $200 a member, and that was Aeroplan in 2005 (which hasn’t performed to expectations)
  3. The Canadian market was less competitive than the US market, and that per-member valuation was achieved with a member base only 5% the size of American’s
  4. More recent valuations have ranged from $60 – $120 per member.

There’s no doubt AAdvantage is a valuable asset (as are MileagePlus and SkyMiles), and their value represents a substantial portion of what their affiliated airlines are worth. AAdvantage may be worth $5 – $10 billion. But $35 billion isn’t in the realm of reality.

In general analysts likely undervalue the free cash flow provided by loyalty programs, because they don’t understand the programs, but it sure seems the cheerleading goes well beyond anything supported by the numbers.

And it may well be that the market understands something that the Stifel analyst doesn’t — precisely that the airlines have been killing their golden goose. Rarely in history have companies had marketing programs which aren’t costs, but profit centers. And while they’re very profitable now, parent airline valuations which are lower than a financial analyst thinks mileage programs ought to be worth based on current performance suggest that the mileage program profits may not be sustainable over time.

Put another way, Wall Street may be punishing the airlines for devaluing their frequent flyer programs.

About Gary Leff

Gary Leff is one of the foremost experts in the field of miles, points, and frequent business travel - a topic he has covered since 2002. Co-founder of frequent flyer community, emcee of the Freddie Awards, and named one of the "World's Top Travel Experts" by Conde' Nast Traveler (2010-Present) Gary has been a guest on most major news media, profiled in several top print publications, and published broadly on the topic of consumer loyalty. More About Gary »

More articles by Gary Leff »


  1. […] NOT QUITE: Financial Analyst Claims “Airlines Make More Money Selling Miles Than Seats”. – Very interesting analysis of how selling miles and points affects the bottom line of airlines.  I doubt the charges that airlines are attributing (aka cost allocation) are as accurate as they claim.  Many companies are notoriously bad at allocating true costs to their source of revenues. […]

  2. […] with the estimates – supposedly some sort of wall street conspiracy against the airlines. NOT QUITE: Financial Analyst Claims "Airlines Make More Money Selling Miles Than Seats" – … Let's Not Forget How Important Frequent Flyers are to Airlines (And Maybe They'll Remember, Too.) […]


  1. If airlines really are receiving 1.25 cents per mile sold, there’s no excuse for the major devaluations we’ve seen in recent years. Airlines that try to keep too much of that 1.25 cents as profit will end up losing it all. That’s how markets work.

  2. Interesting take, but I’m with the analyst on this one. The analyst said the loyalty program represents ~50% of EBIT in his research. There is a very big difference between the numbers Gary is using and EBIT. While yes, AA reported 7.6 billion of profit in 2015, 3 billion of that was a tax benefit from the years of accumulated losses the airlines have stockpiled. Professional analysts and investors ignore one-time, non-core events like this and base their modeling and valuation work on “normalized” earning power, as one should. We call that ignoring the “pig in the python.” (One time item pushing through earnings is like a pig swallowed whole pushing through the python).

    Normalizing for the tax benefit in 2015, they earned $4.6 billion in 2015 and the analyst’s view is entirely plausible, especially if you are looking at the earnings of the business smoothed out over a decade when airline operations traverse the entirety of the business cycle. Not fair to be cherry picking 2015 peak earnings for the airline side of the business where fuel prices were low and then fudge your analysis even more by using the after tax number to support your argument. Us wall street types call that “window dressing” your analysis.

    Also Gary, it’s not fair to look at a puny deal for a Canadian loyalty program and value the largest loyalty program in the world based on that comp. It’s like saying a farm in rural Idaho sold for $1,000 an acre, so Times Square should be worth $2,000. And that would be called “juking the comps.”

  3. @Rob there have been several spinoffs, that was the largest and most successful, and in a less competitive market. My argument is you’re never going to get a premium over the per-member valuation achieved a dozen years ago by Aeroplan especially since the financial performance of Aimia has lagged.

    I state quite clearly that there was a time when mileage programs were the only profitable part of airlines, so over time it’s not a crazy claim but in recent years it’s a stretch. The claim is being made that the programs are more than half of profit NOW, so it’s more than reasonable to look at the most recent two years.

  4. @Gary thanks, that was interesting, but when I read the Bloomberg article originally, I always thought that headline was not literal, but hyberbole, as many headlines are

    @nsx I disagree. The relevant metric is the % of flown seats being given away to award bookings. That number has been climbing for years. Any one could figure that the airlines would devalue the rewards, not perpetually turn away more and more paying customers to fill more award seats.

  5. “American has booked the lowest liability for future redemptions of the major US legacy carrier frequent flyer programs…”

    Does this have any relationship to why American stopped releasing Saver availabilty?

  6. @Gary, I hear ya on the half of profit now argument, and I agree that may not be true every year. But the analyst is trying to value the company as a going concern, which is how the street is going to value it, and to say half of the earnings are coming from the loyalty program is very likely a true statement.

    I do disagree with the assertion that American will never achieve a premium per-member valuation to one achieved a dozen years ago. That thought ignores scarcity value (there are fewer of these programs to acquire than 12 years ago), control premium (a controlling stake gives the buyer the power to devalue the points to zero and make a windfall if it feels like it), as well as the most important thing…a strategic premium (i.e. how much would Delta overpay for it to keep United from controlling the number 1 market share?)

    I’m not smart enough to answer all of these questions but I would point out Warren Buffett just made an enormous bet on AA after decades of calling all airline stocks a bad idea. He is famously the one who said “the best way to become a millionaire is to invest a billion dollars in an airline.” Something changed his mind. Don’t know if he’s right, but there is something to the notion that the loyalty program could be massively undervalued.

  7. @Rob There are 2 separate things going on here.

    1. The popularized notion of miles being (currently) a majority of profit. I’ve pointed out in the post that at one point they were ALL the profit, but currently (now) they are not in fact a majority although they’re a significant source of profit.

    2. Just how valuable the programs are, and what that means for the value of airline stocks. Here I suggest that the analyst substantially overstates the likelihood that the revenue is stable. There’s very little room for revenue growth, and quite a lot of risk. There’s one major customer for each program, and that the future of that revenue is highly sensitive to maintaining current levels of interchange rates. Those are far more likely to fall than to rise, there’s regulatory risk.. bargaining risk… and most of all technological risk, the banks even are competing to offer new payment technologies that process at lower cost.

  8. @ Gary, Fair enough. As a career stock analyst, I can tell you the risks you cite are the reasons why the stock isn’t trading at the analyst’s price target right now. Part of the exercise of equity research is to paint a picture of how you think it will play out and postulate the opportunity that exists if you are right and the rest of the world is wrong or pricing in too much risk. I really took notice when Buffett bought in, because he is quite consistently risk averse, which suggests to me that he must be seeing a major disconnect in market valuation vs intrinsic value he sees. While no company’s revenue outlook is dependably stable, I think you may not be giving enough credit to the demand from banks ready to take the place of the big three, should they ever decide to pull back from their relationship with AA. Barclay’s is a good example.

  9. The problem with trying to determine the “value” of a company is that it implies you’re look for a “rational” valuation. That rarely exists on Wall Street. I agree with Buffett that airline stocks are significantly undervalued here, but I’m not certain that in the very short term that will change. The collapse of oil prices in 2014 ushered in the era of Mega Profits for the airlines, but Wall Street assumed they were just temporary. They weren’t temporary, but those conditions weren’t replicable either, as airfares did start to decline a bit as the industry settled into a world of lower fuel prices.

    So we now have a situation where airlines are wildly profitable, but those profits aren’t growing, and Wall Street tends to like stocks where there is high growth potential (even if it isn’t always realized). I suspect that when airline profits do start to grow again, their market capitalizations will also start to make more logical sense.

  10. @Rob – the issue is what you think will happen to interchange, we’ve already seen it fall in Europe and Australia and the credit card co-brand contracts in the Mideast build in contingencies if the same thing happens there. Lower fees to banks for processing cards means it doesn’t make sense to spend as much per dollar of transaction to incentivize use. In much of the world it’s regulatory change that’s the driver but ultimately technology has to drive down the cost of processing.

    That doesn’t mean the end of cobrand cards, the way the practical elimination of transaction fees on debit cards has eliminated those from the mileage-earning space. Credit cards don’t earn only per-transaction but also earn finance charges. In the French market Air France and American Express fund (less generous than the US) mileage-earning by splitting the APR. It’s not as lucrative a business for Air France as the cobrand deals in the US, one reason why Air France is so anxious to partner with all 3 of the major bank transferrable points currencies.

    Point is there’s real risk in the business from the largest banks, and that’s the single largest source of revenue for the programs. So you don’t just chart it out with growth into the future. And remember that the underlying airlines haven’t been growing as fast as miles are printed, so the program either have to disappoint more and more customers or their costs have to go up (and margins down) to provide reasonable fulfillment to keep the game going.

    The programs ARE valuable, but they’re not as valuable as this analyst thinks. And there’s nothing that suggests this analyst’s view is what’s driving Buffet’s investments. I’m not making a claim here about the overall value of airline stocks, or any individual airline’s stock…

  11. @ Gary. All very valid points, but I think you are focusing WAY too much on interchange fees in drawing your conclusions and it is steering you down the wrong path.

    Incrementally, sure interchange matters somewhat, but the driver of the credit card business is and always has been net interest income. The reason rewards went away on debit cards is exactly because of this point. Debit card transactions only generate interchange fees, there are no loans created, so when fees went away, so did the ability to pay anything out or the incentive to do so either.

    Interchange fees are one-third of the revenue and a much smaller share of the profit from the bank’s perspective. Don’t get confused by the fact that they are incentivizing new card accounts to spend $3,000 in the first 3 months as meaning they profit from activity. They want you to spend because they are banking on some of that charging sticking as a balance…that is what they want. The interchange fees just cover the overhead of doing business.
    I know it is hard to come by real numbers because the banks work very hard to obfuscate the metrics for competitive reasons, but that is why folks like me have a job. Take a look at JPMorgan’s investor relations page and download their 4Q16 Earnings Supplement to see for yourself what analysts are looking at.

    On Pages 11-14 they give enough detail about the credit card division to see that the interchange story is a sideshow compared to the net interest income. They disclose that the “card services revenue rate” (2016 full year) is 11.29% of average loans (pg. 14). On average 2016 credit card loans of $131.2 Billion (pg.12) that is JP Morgan’s way of telling analysts with decoder ring powers that their credit card division put up $14.8 billion of revenue in 2016. And they do actually make it easy to find the interchange fees in 2016 (pg. 4 “card income”) in the “non-interest revenue” section of $4.779 billion. (There are other little fees in that number, but it is predominantly interchange). So for Chase/JP Morgan in 2016, only 32% of their card business revenue came from interchange fees and the rest was from net interest income.

    And that is revenue. Earnings will be more like 80% driven by net interest income because of the huge interest rate spread between the 16% they are charging balances and funding at fed funds rate of half a percent. Furthermore, times are good for the economy right now and the charge off rate in the card division is a tiny 2.6% so they are clearing more than 10% on that huge $131.2 Billion average loan number. That is why the profitability of the loan balances is so high, there isn’t much variable overhead when a customer carries a $10,000 balance versus a $2,000 balance so it is all incremental profit.

    So the very big conclusion all of this leads us to is that the value of the loyalty program essentially takes on the growth profile of the underlying earnings driver…which is the bank’s credit card businesses that are putting up ROE’s of 18% and growing 10% plus. This transforms the AA program into essentially a high growth, high profit, credit card business. I think you are worried about banks pulling out because you aren’t fully appreciative of how profitable this arrangement is to the banks.

  12. @Rob — Interesting analysis. Thanks for sharing it. Obviously, the usurious interest rates banks charge on credit card debt is highly profitable. But logic would suggest to me that rich people — the kind who get Titanium Platinum cards — might be better at managing their debt and spending than “ordinary” middle class folk. Which suggests to me that I might not want to target them with generous sign-up bonuses because they’d be likely to NOT run a balance on their credit cards.

    Am I wrong? Are wealthy people equally stupid with debt? Do they take credit cards and buy more than they can afford? Logic would at least suggest to me that wealthier people would have greater financial resources to refinance their debt and not pay usurious credit card interest rates.

  13. @Rob of course APR is a huge deal for the banks, however interchange matters — we’ve seen what happens with rewards cards in Europe and Australia when interchange falls, and if APR were enough Citi wouldn’t have tapped out in the deal with AA and given up exclusivity.

  14. @ iahphx. Thanks, yup wealthy people are a gold mine because they are often carrying balances on behalf of their business and write off the interest expense. I know it seems like stupidity, but that is one of the biases of being so engrossed in this hobby is that we assume everyone is motivated by the same things, but we are a very tiny minority…most people don’t pay any of the issues we care about a second thought.

    @ Gary, I would just say Europe and Australia are much less profitable in general for the banks, so it is more likely to be affected by interchange fees. I would caution against using that as a model of what you would expect to happen in the U.S. The consumer spending culture, disposable income and credit utilization of American consumers is unparalleled by large orders of magnitude. Something like 80% of Australians only have 1 credit card, so there is not a credit culture, giving banks more leverage to pass costs through to the consumer who isn’t going to notice anyway since there aren’t 15 banks sending him credit card offers every month.

  15. @Rob misunderstands the Australian credit reporting system. Anyone regularly applying for various credit cards, churning that is, is going to be refused pretty quickly despite their credit worthiness. So yes, most people have one or two cards not for any enticing rewards but for normal day to day use. The US airlines should cherish those loyal consumers as the day will come again when spending on their co-branded credit cards and all others take a dive as recessionary circumstances take a grip. Yep, hard times have not been banished, and the political instability being seen now is fertile ground for a new round to dwarf the 2007/8 recession, which devastated many businesses and households.

  16. @CoolHandLuke You are actually helping make my point. The australian banks can afford to be picky with rejections and churning for the same reasons why they can afford to cram interchange fee reductions back on their customers. My point was that US banks won’t likely do the same if interchange fees suffer a reduction. The U.S. competitive environment and value of the U.S. customer is much more likely to lead to the banks eating the majority of any interchange hit.

  17. Say what you want, but it’s ABSOLUTELY TRUE. Airlines would rather sell miles to banks than tickets to flyers. Once you understand this, everything else should make sense including why carriers routinely hold out for EXCESSIVE premiums on underselling flights. In reality, the money was already made on the mileage sale to the bank.

Comments are closed.