A Reasonable Redemption SNAFU

They’re barely out of the gate, but already there’s a problem with Reasonable Redemption Values.  Reasonable Redemption Values (RRVs) are estimates of the redemption value one might reasonably get from various types of miles and points.  In yesterday’s post, “Reasonable Redemption Values taking shape,” I presented the following table that shows the RRVs in cents per point for various airlines and cabins:

RRVs based on estimated ticket value:

  American Delta Southwest United
Domestic Economy 1.33 1.02 1.65 1.37
Domestic First 1.01 0.78   1.03
International Economy 2.12 1.70   2.44
International Business 2.58 1.86   2.56
International First 2.71     2.29


In response, the post received the following comments (among others):

The Delta Domestic First Class redemption is interesting, as cardholders should note that their RRV will be closer to 1 with the PWM option.

I am probably going to take a lot of heat for this, but if each point is really only worth about 2 Cents even for International Premium Airfares, why not just MS the heck out of the Barclays Card at 2.2cpm?

when there’s PWM option, the delta value should never be below 1.0. If someone is redeeming for below 1.0, I would say their redemption is NOT reasonable taking that out of the math.

RRVs should never be below 1 for Delta, because you can always get a penny a point for them if you have their credit card.

All of the above comments share the same incorrect root assumption.  Whether the commenters know it or not, each above comment makes the underlying assumption that RRVs are based on retail flight prices.  With economy cabins, that’s true.  With premium cabins, it is not true.  With premium cabins, I made the following value assumptions:

  • Domestic first class is “worth” 1.5 times economy
  • International business class is “worth” 2 times international economy
  • International first class is “worth” 2.5 times international economy

These assumptions of “worth” or “value” are intentionally less than observed ticket prices.  The thought is that even though premium cabins cost much more than economy, we don’t necessarily value the difference as much as the price indicates.

Here’s a way to think about this: if a first class international flight costs $10,000, but you value it at only $5,000, then even US currency pennies are not worth 1 cent each when redeemed for this flight. In this scenario, pennies would have an RRV of half a cent each.  Yes, its confusing.

I ran a number of scenarios to compare real world economy ticket prices to real world premium cabin prices and found the following averages with my small sample:

  • Domestic first class costs 2.8 times economy
  • International business costs 3.2 times international economy
  • International first class costs 6.5 times international economy

If I were to recalculate RRVs based on estimated retail prices (using these multipliers), we would see a very different result:

RRVs based on estimated ticket prices (rather than value)

  American Delta Southwest United
Domestic Economy 1.33 1.02 1.65 1.37
Domestic First 1.89 1.45   1.91
International Economy 2.12 1.70   2.44
International Business 4.44 3.03   4.14
International First 7.74     6.00

Note that by basing the chart on retail prices, you can see that Delta SkyMiles will result in more value with traditional awards in premium cabins (even domestic) than would pay with points options (which give a value of 1 cent per mile).  You can also see that AA, Delta, and United miles are much more valuable when used for international premium cabins.  You won’t get nearly as much value using a fixed value point program (such as Arrival) or fixed value miles (pay with miles) to purchase premium international cabins.

A fix is needed

Going into this project, I had thought that RRVs would be easier to understand and use than Fair Trading Prices.  The fact, though, that smart people immediately misunderstood the meaning of the RRVs in yesterday’s post suggests to me that the concept isn’t as easy to understand as I thought.  Even though I developed the RRVs, it took me a while after reading some of those comments to figure out the misunderstanding.  My first reaction, to be honest, was “yep, you’re right” before I finally caught on to the underlying issue.

I could continue on unabated in the direction I’ve taken so far and use long posts like this one to explain to each new reader what’s going on.  That seems like a bad idea to me.  Metrics are only useful if they are understood.

An easy fix would be to simply change the estimated value in the spreadsheet to be based on estimated ticket prices (as I did in the second table above).  The term “value”, then, would mean something like “the price you would likely have to pay for that flight and cabin combination.”  This approach has a few advantages:

  • The results would be easier to understand
  • Pennies would once again be worth 1 cent each
  • It would be easier to compare real world redemptions to these values.  For example, if you had the choice between redeeming 125,000 Delta miles and $100 in fees or paying $7,500 for a business class flight, you could calculate your real world redemption value as ($7,500 – $100) / 125,000 =  5.9 cents per mile.  Then, you could compare that result to the table and see that 5.9 cents per mile is better value that the RRV (3.03), so it is a good use of miles.

The one disadvantage I can think of is that many people truly do not value premium cabins (or premium hotel rooms, for that matter) as high as the corresponding retail prices.  That said, I still think the metric would be more useful if based on estimated prices even if those prices do not correspond with people’s views of “value”.

Reader input requested

Before I turn on a dime and change to a price based scheme, I’d love to hear your thoughts about this. Should I switch to price estimates even for premium cabins?  And, if I do, should I change the name to something like Reasonable Redemption Prices (RRPs)?  Or, would it be better to go forward with both metrics?  Should I present RRVs and RRPs side by side or would that only be even more confusing?  Or, do you have other solutions in mind altogether?  As always, please comment below (and sorry about the CAPTCHAs… I hate them too).

About Greg The Frequent Miler

Greg is the owner, founder, and primary author of the Frequent Miler. He earns millions of points and miles each year, mostly without flying, and dedicates this blog to teaching others how to do the same.

More articles by Greg The Frequent Miler »


  1. And for a mathematical encore, our host will mathematically ‘FROM INSIDE” a cardboard box demonstrate to us all the theoretical probability that a punch through the cardboard box may indeed lead to a never ending series of therapy sessions and doctor’s bills associated with a potential hemorrhage of the knuckle thereby rendering the subject completely locked within their own box.

  2. I don’t get the value angle. At some point you need to exchange currency for a ticket. Even if I’d never actually pay the cash price, I’d still value my points exchanged in those terms. The second chart is more useful to me. Thanks for your efforts.

  3. Your original estimates with RRV were itself higher than what I calculated for United/AA international redemptions (I value 1.2 for AA and 1.4 for United). RRP using retail prices of premium cabin will be totally flawed.

      • Dear FM – I priced several international tickets that were around $1200-$1400 for typical 85000 UA points from US to India. You have to then account for loss of frequent flyer miles that would not be earned with award ticket. And then add back the value of being able to cancel/change an award ticket for minor fees versus ticket bought on cash. Followed again by risk of miles getting devalued over time or restraints on being able to get multiple seats on a flight for a family of 4, plus limited days for availability of award seats. So modeling it for all this ends up with no scenario of exceeding 1.4 cents/mile for UA or AA miles.

  4. Well, I prefer the RRVs as they are.
    I find the RRPs interesting on the level of, “hmm, that’s interesting”, but not useful to me in any practical way. Other than to remind me that I’ll probably never get into a first class cabin if all I’m using are cash back cards.
    I do find it confusing that the first line in the chart is domestic first. Is that because you have them ordered by increasing value? That totally threw me off, because even though it says first, I was thinking economy and then also thinking, well, those values should be based on retail price, and therefore the arguments about PWM kind of make sense. Once it sunk in that those are domestic first class, then I understood your point. Does that make sense? I wonder if anyone else was thrown off track by the ordering of the lines.

    • Thanks Jamie. Yes, the tables were sorted by RRV so the top table had domestic first class on top and the second table had it second. I’ve changed it to be as expected and will change the sorting in yesterday’s post as well.

  5. No, no, no, don’t change!
    Now for the non-emotional reasoning: very few to none of your readers or likely readers will ever pay personally for retail priced international premium cabin tickets (and if you extend the analysis, high end luxury hotel rooms), so RRPs would be an irrelevant statistic for those cabins/rooms. They would just be a way to feel good about higher ‘value’ received for points/miles, and encourage people to pay silly prices to acquire those points/miles.

    RRVs are at least more relevant in that they attempt to get at what real cash price multiplier one might pay for a premium cabin/room. I’m fairly sure your audience would not pay the retail 3.2x intl J multiplier personally (e.g. $7500 in J to Australia), but they MIGHT pay something closer to your value estimate of 2x. Although I personally think that is too high as well, the point is that the premium cabin/room multiplier has to be an RRV based personal value estimate, because retail price multipliers are too high to expect this audience to pay.

    My sense is that the Delta 1cpm reaction is an emotional one based on people’s iron-clad beliefs of the fall back Pay With Miles option being a floor for the value of that currency. Buying an overpriced commodity often leads to value received being less than the price paid.

  6. I second Jig’s comment above. The Daily Getaway flops prove this to me. No one thought $84 for a Super 8 room was a good deal even though they said you would save over their inflated rack rates. Deep down people know the published 1st class fares aren’t worth the quoted price, they just want to feel good about their redemptions by saying how much “value” they got out of their miles.

  7. Frankly, I can’t stand when some blogs tout how you can get, for example, 10 cents per mile for an international premium flight just because the flight was $10,000. I like your approaches better.

  8. FrequentMiler,
    How about adding a footer to your charts when you publish them with the 3 lines you posted above about your assumptions of worth? This might help prevent problems in the future.

  9. I think the current calculation is better. There might need to be a footnote explaining why value is different than price. When the price is so high that the average person won’t buy it, it becomes irrelevant to them.

  10. Greg, the engineer loves the first method and the ability to adjust the values in the spreadsheet to suit your personal take on the worth of the value. however, the reality is that you’ll need both methods depending on the person using it.
    Think of it like the financial term “payback”. Anyone who knows basic finance knows that this term is not valid since it does not incorporate the time value of money, but payback is commonly used in business since it is easy to calculate and relate to. Real analysis is much more detailed. I see the same thing with you two methods.

  11. I would either show both RVV and RRP, or just stick with the original RVVs. Those cash ticket prices are meaningless to me, as I would never actually pay them. Since it had been awhile since your initial posts on this subject, people probably just forgot your valuation assumptions for the premium tickets (which by the way seem pretty reasonable to me, and I assume a user could customize in their own spreadsheet copy if desired).

    • What RRV and RRP really boils down to is each user’s perception of the value of premium versus economy. For those who would never pay the cash price, the RRP is irrelevant. For those who would pay the cash price, that means the valuation assumptions are wrong for that user. So the solution seems to be 1) design the spreadsheet so that each user can input his own valuation assumptions for premium tickets, and 2) use a (prominently disclosed) reasonable set of valuation assumptions appropriate for a “typical” user in the generic spreadsheet, which is what you have already done with the RRVs.

  12. Greg:
    I prefer the cash value. When I look at booking a ticket with points vs. cash, I divide the cost of the ticket by the number of points for the ticket to get the cents/point I am getting. I then look at what my points cost me and if the value of the points realized is greater than the cost of the points, I book the ticket.

  13. My recommendation is to stick with the RRV metric but clearly state your assumptions as a footnote to the table.

    I can’t find fault with your calculations, but the results are disheartening when taken collectively with your posts on fair trading values. I primarily deal in Skymiles and redeem for domestic coach travel with my family. You give the FTV of a Skymile at 1.29 cents and the RRV for domestic coach at 1.02 cents. While it’s easy to beat both of those and create an arbitrage opportunity, your face value FTV and RRV values make it look like I’m going backwards by playing the game. Am I missing something here?

  14. I agree with Jig, and others – the actual price of a premium int’l class ticket has little relation to how *I* value it. I’m much more in line with your assumptions. If you want to get fancy, you could make the multiplier for the various classes variable, and allow individual adjustment at the spreadsheet level.

    Thanks for your thoughtful approach to the valuation question – definitely food for thought.

  15. as one of the one’s that commented about 1.0 CPP with skymiles. I agree with this sentence “Metrics are only useful if they are understood”
    I would also say that metrics are not useful with the methods are fictitious. There’s confusion because assigning a random multiplier is fairly meaningless (results would be different if you added 0.1 to each multiplier).
    A fictitious assumption doesn’t result in correct information. Garbage in, garbage out.

    You have to assign values to each category based on some methodology for determining that value. Determining a methodology for one value than multiplying for others will result (and did) result in confusion.

    Regardless, there’s only one way to assign a reasonable value of less than 1.0 cpp for skymiles. That is if you incorporate discounts (i.e. skymiles) into the calculation.
    i.e. a $1,000, 10,000 mile ticket fare will yield. . . .30,000 miles (or whatever, as it depends on medallion status and class of service). Those are miles you don’t receive (you don’t get a discount) if you pay with miles, so the 1.0 cpp value would be dropped to ~0.72 cpp.

    That being said. Everything is worth different amounts to everyone, and you’re not going to make everyone happy. Which is why everyone should determine their own reasonable redemption values.

    • I would also add that hypothetical fair market values of a business class do not even relate to coach through a constant multiplier across all airlines, routes, distances, etc…
      So my vote is to abandon the multiplier method.

  16. I’m really enjoying the discussion about valuation. I assume that mile has the opportunity cost of about 2 cents, assuming that’s what an average set of cashback cards would provide for my mix of credit card usage. I think 2 cents is reasonable since you can get 5% in some categories but not all, and the Arrival Card is 2.2% on travel. So that means to me that I need to get at least 2 cents per mile to breakeven.

    Using this approach for cost: dollars spent generate points, I like using the actual prices, since I used the actual dollars spent to generate the miles.

    For me, it’s simple: only use the miles for international premium seats, and avoid fuel costs as much as possible.

  17. Stop. You are being completely dishonest.

    Your so-called “valuations” mean nothing – NOTHING!

    It is a complete fiction to assert ANY of your numbers. Why? Because you completely ignore the dirty little secret that is obvious to anyone who has ever tried to book an award ticket: the theoretical “cost” in points/miles for a given seat is NOT an indicator of how many points you need to throw at the airline to actually get that seat. That[s because, in many (most?) cases, THE PUBLISHED AWARD CHARTS ARE A LIE. Bloggers, who have a financial interest in perpetuating this lie, never cal out that lie – you are a a willing participant in this deception.

    IN THEORY, I might be able to fly from Portland to Paris in Business class on Delta for X miles. I might be able to fly from Seattle to London in Business class on American for Y miles. Well, as anyone who has ever tried to actually USE their miles, those X and Y numbers are simply not available. In some cases, they are not available EVER, period, in the real world – not 330 days ahead, not tomorrow. In other cases, there might be a single seat, maybe released 3 days before the flight leaves. maybe a few scattered days during the worst winter weather of the year – times when nobody wants (or can) go there. Bloggers, in their limitless greed for credit card referral fees, gloss over these facts with such platitudes as “availability is tight” but that is simply being dishonest. And you know it.

    The relative value nonsense is just a scam, clickbait for the suckers, a feel-good attempt to obfuscate the truth, despite all your disclaimers. A lie is a lie is a lie. Bloggers lie.

    • Award availability is a VERY well explored area. Looks like you’re already well informed, and so are lots of other travelers. This blog IMO excels more with new & original (sometimes unorthodox) content. I hope it continues that way 😉

    • You are mistaken and rather vehemently mistaken. I have 6 business class tickets together to fly to London and fly home from Munich in June. I also had 6 one ways from Canada in business included as part of this ticket. Not only do I have these tickets, but I also have been able to change them 4 times over the course of the 11 months I have had them, so there was availability for 6 tickets together in business for my dates at 4 different times over the last year. I also have 6 tickets to Costa Rica for Thanksgiving this year. Just because you are not willing to invest some time to get top value from your miles/points, doesn’t mean that there is not value to be found.

  18. I would keep your original calculation with one change:

    +Take into account of the miles one would have earned if they purchased a paid fare. For example, if you will gain 5k miles for a cross country flight in economy, that 5k should be translated to RRPs and then subtract from award redemption.

    This calc should indicate that sometimes it may not even make sense to redeem international economy flights because the miles foregone would have been greater than the $ saved by using miles instead of $.

  19. I guess the only way your premium numbers work would be if you were buying miles to redeem immediately…….but that still doesn’t calculate for me. A more accurate figure would be if I fouled the best sales price in that route then the ticket would cost X. For me realistically, it is when there are business seat wars on certain routes that really become the gauge of those premium seats value.

  20. I think that miles/points should be valued based on what each individual is willing to pay in cash for the service. By way of example, if I would be willing to pay $500 for a RT domestic ticket, then I value the miles by dividing the $500 by the miles expended (in this example 25,000 saver miles + 5,000 flown miles lost + 5,000 elite bonus miles lost) of 35,000 miles. This results in a value per mile of approximately $.01425 value per mile. On a simple international business fare I would substitute a cost value between $2,500 and $3,000 and adjust the miles as appropriate. Everyone should use their own values. Many people would prefer to fly coach rather than pay $2,500 for that business ticket. They should substitute the price that would cause them to choose business class. Different trips and elite status result in different mileage costs and should be appropriately adjusted. I believe this is the intent of the RVV calculation. We can delude ourselves by using ticket costs that we would never be willing to charge to our plastic, but that’s like buying something you neither need or want because it is on sale.

    • Further expansion is the idea that fuel surcharges have to be factored in at some point because they must be paid and DO affect the picture

  21. I vote for keeping the RRV – this is exactly how I valued miles/points intuitively already, but has more objective calculations behind it. The only subjective part based on opinion and feel is the value one assigns to premium cabins and you have very clearly laid them out, so that anyone who values it differently can change the assumption in one place and come up with their own RRV calculation. RRP would lose the ability to say how much would I really be willing to pay if I were paying cash for a premium cabin which is how most of us decide whether the redemption makes sense or paying cash makes sense. So I vote for RRV.

    FWIW, my intuitive multipliers for the premium cabins happened to be exactly what you put down as the default. The only part that I had not thought of was accounting for the loss of frequent flier points on redeemed tickets. Interesting twist… (@Mike/@Lin, thanks for raising that important point – on cheap long distance tickets that can make a big difference, especially if elite miles matter even more to someone).

    One minor quibble on RRV – I think the RRV term is not self-explanatory. I mean, it describes what it IS but perhaps it should be based on what it would be USED FOR by most of us. I still have to keep going back to your introductory posts to remind myself of what RRV was trying to achieve. So if anything, let’s come up with a more precise name but keep the calculations as they are. Hopefully that may make multiple blog posts and comment replies explaining why its not RRP redundant 🙂

    I can’t think of any specific replacement term at the moment but promise to put some thought into it.

    That’s my two cents… (even if it is worth less than that in redemption value 😉

  22. Agree with Ken S above. The purpose of this exercise seems to have been lost a little– my view is that these metrics should help the flyer understand 1) when to use miles vs cash, and 2) which miles/points to pursue. Thus each person needs to use their perceived value of each class of service for a particular itinerary to be able to effectively compare.

Leave a Reply

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