Over the weekend, a friend showed me her wallet full of carefully labelled credit cards. One card was labelled “gas”, another “rental cars”, and another “grocery”. She had applied these labels to remind her which card to use where. That was all great. Then I saw one card was labelled simply “2.73”. I asked what that was about. She answered that it was her Freedom Unlimited card using “your numbers”. She meant that the Freedom Unlimited card earns 1.5X everywhere, so with the Frequent Miler Reasonable Redemption Value (RRV) for Chase points at 1.82 (at the time), the Freedom Unlimited earns 1.5 x 1.82 = 2.73%.
The RRV problem
My friend was using the Reasonable Redemption Values (RRVs) not as intended. RRVs estimate how much travel you can reasonably expect to get from your points if you use those points wisely. I think that’s useful when trying to decide which points to use for an award. Will you get more value than the RRV? If so, you can call it a win. Similarly, I find RRVs useful for comparing credit card welcome bonuses. We use RRVs to estimate “first year value”. The idea is to be able to compare, across cards, how much travel value you might get from the first year of owning the card.
RRVs were not meant to tell you a good price to buy points. In fact, you should never buy points for the same price that you’ll redeem them for. You should always look to buy them for less. Here’s an example: suppose you can buy a flight to Europe straight-up for $1,000. And further suppose that it would cost $1,000 to buy enough points to book that same flight as an award. Would you buy the points? No, of course not. The only reason to go through the extra hassle of acquiring points is if you’ll either save money or get a better outcome such as first class instead of coach.
When you decide to use a points earning card over a cash back card, you are essentially buying points. Here’s an example: if you can choose between earning 2% cash back or earning 1.5 Chase Ultimate Rewards points per dollar, then picking Chase points costs you 2 cents for every 1.5 points you “buy”. In other words, by making that trade-off, you are buying Chase points for 1.33 cents each. Nick covered this topic in detail in his post: How much do you pay for your miles and points? And, I’d argue that 1.33 cents is a fair price to pay for Ultimate Rewards points, so that’s fine. The problem is that my friend was using the 1.82 RRV number to decide which card she should use. By labeling her Freedom Unlimited as “2.73%”, she would likely use this card instead of a card earning 2.5% cash back (if she had the Alliant card, for example). If she made that choice, she would be essentially buying Chase points for 2.5/1.5 = 1.67 cents per point. That’s not terrible, but it’s also far from a bargain, especially if she uses those Chase points to buy travel through the Chase portal for 1.5 cents per point value.
RRVs were not intended to be used for comparing credit card spend rewards. As I mentioned earlier in this post, I think that RRVs are a decent way to compare credit card welcome bonuses, but that’s because credit card welcome bonuses do not have the same type of opportunity cost as points earned through spend. There are opportunity costs involved in signing up for new cards, but they’re different in nature. That topic could easily make for an entire post or two. With points earned through spend (or from shopping portals for that matter), there are clear cash trade-offs which make it important to consider the opportunity cost of earning points rather than cash. If my friend doesn’t have any cash back cards to compare to, then RRVs are a fine way to compare the relative value of credit card rewards. But when cash rewards are part of the equation, I think a better metric would be an estimated cash value, rather than estimated redemption value. View from the Wing describes this concept as “the amount at which you are indifferent to holding miles versus cash.”
It’s my fault. Not long ago, I enhanced our credit card displays to show point earnings in the form of RRVs. For example, the page “Best Everywhere Else Rewards Cards,” previously showed this for the Freedom Unlimited:
See how it says that the Base earnings for this card are 1.5X which equals 2.73%? Yeah, that’s my bad. That display absolutely encourages the RRVs to be used the way my friend was using them.
Identifying the correct cash value is an impossible task. I described this problem at length in the 2013 post “Impossible point valuations and the joy of free.” There I discussed the fact that points are worth more or less based on how they’re used, how many you already have, how easy they are to use, whether or not they add value (such as free flight changes), whether or not you have elite status, your travel habits (e.g. airline miles are often worth much more on international premium cabin flights), and your subjective value of the “joy of free.”
It is important for us to establish estimates, even though getting it right is impossible. Most consumers have no idea how to estimate the value of points earned. For example, many consumers would assume that Hilton cards, which earn at least 3 points per dollar on all spend, must be better than the World of Hyatt card which earns only 1 point per dollar for most spend. In reality, most of us who understand the relative value of these points would probably choose 1 Hyatt point over 3 Hilton points (unless you frequently use Hilton points at category 1 and 2 hotels) if we didn’t have a brand preference. In other words, I strongly believe that consumers need some guidance as to how to value points and miles.
A better solution may be to estimate ranges based on use case scenarios. For example, “for those who use points to fly domestic economy, United miles are worth between XX and YY.” Ranges would be more accurate, but less useful. Ultimately our goal is to compare rewards earned from different cards, different promotions, different portals, etc. Point estimates work far better than ranges for those purposes. Point estimates based on use cases would be a better solution, but it would be much harder to create and maintain. I need to make a lot of decisions with this blog as to where to invest my time (and Nick and Stephen’s time too), and I’m not interested at this moment in investing in such a large project. The problem isn’t just creating the estimates, but rather in displaying them in a useful way. Our credit card database and displays are setup right now to expect a single point value estimate. Anything different would be a lot of work.
The RRV fix
Over the weekend, Nick and I discussed this in detail. Rather than try to generate brand new cash value estimates, we decided to improve our RRV calculations in ways that would make the values more useful when comparing card spend rebates. Specifically, we made the following adjustments:
- All airline mile RRVs were adjusted downward by 7%. This is intended to approximate the loss in earned rewards for those flights. Previously we calculated that most airline Miles are worth 1.4 cents each. Now, we’ve reduced that 1.4 value by 7% to 1.3. Similarly we used previously calculated values for the rest of the airlines (see: What are oddball airline miles worth?) and reduced those values by 7%.
- We did not make a similar adjustment to hotel values because there are two counter-weighting factors with hotel RRVs which were based on observed hotel prices compared to observed point prices: Our RRVs are arguably too high because we don’t account for the fact that you don’t earn hotel points on award stays; and our RRVs are arguably too low because we don’t account for the fact that you usually do not pay taxes and fees on hotel award stays. A simplifying decision is to assume that those two factors cancel each other out.
- We changed our adjustments for transferable points programs. With transferable points currencies there is a much larger pool of high value awards one can pick from. So, we assume that a reasonable award value for informed consumers will be higher for transferable points. Previously we accounted for this by simply increasing the standard airline RRV at the time (1.4) by 30% for all transferable points programs. There were two problems with this: 1) 30% was arguably too big of a bump; and 2) we treated all transferable points programs as being equal. Now, we use the adjusted airline mile RRV (1.3) and adjust upward differently based on our subjective assessment of each transferable points program. We then rounded each one to the nearest .05 as follows:
- We lowered the Hyatt RRV to 1.5. We no longer have a data driven source for Hyatt point values, so we instead lowered this value to match the Chase Ultimate Rewards point value. This makes sense to us because Chase Ultimate Rewards points are transferable to Hyatt 1 to 1.
While none of the above RRV changes alter the nature of RRVs, the RRV values were reduced to where Nick and I feel reasonably good about them being used for the purpose of comparing point rewards to cash rewards. No, it’s not perfect, but nothing is.
With the above changes in place, we now see that earning 1.5X Ultimate Rewards points equals 2.25% rewards based on RRVs. That seems about right.