Save your points or use them now? A rule for one balance

The trade in one paragraph

Saving points has two costs and one benefit. The costs are erosion, when a program you would have used raises its prices or a bank cuts its ratio, and opportunity, when a booking you could have made today is gone tomorrow. The benefit is optionality: points at the bank have not committed to any program, so a change at one partner does not trap them. Spending points now locks in today's price and ends the exposure, but only for the booking you make; the rest of the balance stays exposed.

That is why the useful question is never whether points in general will lose value. It is whether this balance, held for a specific future trip, is more likely to buy that trip later than a booking in hand would buy now. A balance with no named trip has nothing to compare against, and tends to sit until a change decides for it.

None of this is an argument for transferring early. Points moved to a program are no longer optional; Chase, Amex, Capital One, Bilt and Citi each say a transfer cannot be reversed or returned (Citi in its 2022 Premier Card terms summary, the latest Citi transfer terms we could open). The stop signals before any transfer still apply. The choice here is between a booking now and points left at the bank.

Dated changes, read on the official pages

Program changes are the reason saving has a cost, so it helps to see them as dated facts rather than as a general mood. Each row below was re-read on its official page on October 28, 2026, and the date column says what kind of date it is: an announcement, an effective date, or the date a page was last updated. The record is not complete; it covers only changes that touch programs US bank points reach and that their own pages state.

Read each row as what it did to someone holding points for that program, because that is the only reading that matters on this page. The ANA revision of June 24, 2025 changed the required mileage for international awards, except certain classes and seasons. A balance sized for an ANA itinerary before that date had been sized against prices that no longer applied. The same date added one-way awards, which cut the other way: a balance short of a round trip could now book one direction and wait on the other.

The Hyatt announcement of February 25, 2026 changed what an unbooked balance could expect; what it means for stays already booked is covered further down. Once a hotel moves some nights into the Upper or Top levels, points saved for that hotel at its old price can cover fewer nights, and the announcement did not say in advance which hotels or which nights would move.

The Chase row is the one that reached all the Hyatt-bound points on the affected cards at once, and the next section prices it. The Bilt row is a different kind of fact, with no price involved, and the section on partners leaving the list below deals with it. Price rises cost a named trip some points; a closed door costs it the route, unless a second program had already been priced.

What the 4:3 Hyatt change does to a saved balance

The Chase change is the clearest example of erosion at the bank rather than at the program, because nothing about the hotel changed. Chase's June 15, 2026 announcement says Ultimate Rewards points transfer to World of Hyatt at 4:3 for Sapphire Preferred, new and existing Ink Business Preferred, and existing Ink Plus and Corporate Flex cardmembers, so 1,000 Ultimate Rewards points become 750 Hyatt points. For Sapphire Preferred cardmembers who applied before June 15, 2026, the change took effect October 1, 2026; for new applicants from June 15 it was immediate. For Ink Business Preferred, and for existing Ink Plus and Corporate Flex cardmembers, it took effect October 1, 2026. For a single Hyatt night, the 4 to 3 break-even grid shows the cash rate each award price has to replace.

The arithmetic for a saved balance is direct. A Hyatt night priced at 15,000 points needs 15,000 bank points at a 1:1 ratio, the ratio Bilt prints for World of Hyatt through December 31, 2026 (4 to 3 from January 1, 2027, per Bilt's announcement). From those Chase cards at 4:3 it needs 20,000. Chase's release does not print the rate those cards had before the change, so the 1:1 column below is a comparison ratio, not a quoted Chase rate. Against it, every Hyatt booking from those cards costs a third more in Chase points, on the same nights at the same hotels. A holder who saved 60,000 points for four such nights now has enough for three.

Two lessons follow. First, a change at the bank hits all the points on the affected cards that were aimed at that partner at once, while a change at a program hits only that program's prices. Second, a booking made before the change kept its price: the Hyatt points were already in the hotel program, or the stay was already booked. The Chase issuer guide covers how the card tier now decides the Hyatt rate.

Why a balance at the bank is still the safer place to wait

If saving has a cost, waiting inside a program has a bigger one. Points at the bank can still go to any partner the bank lists, so a change at one program redirects them rather than devaluing them. The same points moved to a program early take on that program's risk alone, and cannot leave. Capital One says transferred miles become subject to the receiving program and cannot be transferred back; Chase says transfers are final; Amex says a transfer cannot be reversed.

Hyatt's own announcement shows the difference from the other side. It says the five-level chart takes effect gradually, with limited hotels moving a limited number of nights into the Upper and Top levels in 2026, and that when categories change, existing reservations are honored as booked. That protects a booking, not a balance: points sitting in a Hyatt account with nothing booked get whatever price the night carries when they are used.

There is one more protection worth knowing, and it is narrow. Amex's terms say that if a partner loyalty program stops participating, Amex may limit how many points you can transfer to it before its participation ends. The October 2026 edition added that the limit for an airline program would be at least 100,000 points; the November 2026 edition does not repeat that floor, so no minimum is printed now. Either way it governs access during an exit and is not a promise of notice or value.

The keep-or-spend rule

The rule that follows from all this is short enough to apply at the search screen. It uses only things you can know on the day: the booking in front of you, the target you set, and whether a specific future trip exists.

• If a booking in front of you clears your cents per point target on the bank points it would use, and you would take the trip, book it. Waiting adds erosion risk and gains nothing for that booking.
• If it does not clear your target, keep the points at the bank. A weak booking now is a certain loss against the target; a later booking is only a possible one.
• Keep a balance beyond today's booking only for a trip you can name, with dates or at least a season and a destination, and an estimated points price.
• When you name the trip, price it now in every program your bank reaches. That price is the benchmark the saved balance has to beat later.
• Never move points to a program to wait. Transfer only when the booking is on screen.
• Recheck the named trip whenever a program or bank you would use announces a change. A change that raises its price is a reason to book sooner, if a seat exists at the old price.

Pricing the cost of waiting

The cost of waiting can be written as a simple expected value, with every input stated. This is an arithmetic illustration, not a forecast. A traveler holds 80,000 bank points and has found a booking today that replaces a $1,400 fare with $90 in surviving fees, using 70,000 points: ($1,400 minus $90) divided by 70,000, or 1.87 cents per point. The alternative is to save the points for a named trip next year, estimated today at 80,000 points against a $1,500 fare with $120 in fees: ($1,500 minus $120) divided by 80,000, or 1.73 cents per point.

On those numbers the booking today is better per point. Now add the risk of change. Suppose the traveler judges that the program for next year's trip has a one in four chance of raising that award by a quarter before it can be booked, to 100,000 points, which would be more than the balance. The saved balance then has a three in four chance of 1.73 and a one in four chance of not covering the trip at all. Waiting is the weaker choice unless next year's trip matters much more than this one.

The point of the arithmetic is not the probabilities, which are guesses. It is that writing them down shows when waiting is a bet on a specific outcome. If you cannot name the trip, you cannot even write the bet.

When a partner leaves the list

A partner disappearing from a bank list is the harshest change a saved balance can meet, because no price increase is involved: the route simply closes. Amex's terms plan for it: if a partner program stops participating, Amex may limit transfers to it before the exit, a rule covered above. Bilt's partner article, last updated October 20, 2026, shows the same gap from the holder's side without describing any exit: it lists Atmos Rewards, Southwest and Japan Airlines, does not list American AAdvantage, and does not say whether AAdvantage was ever a partner. A Bilt holder planning an American-operated trip has to reach it through another program that sells American flights, at that program's price, or not at all.

The defence is the same benchmark habit. A named trip priced in two or more programs survives one of them leaving, because the second price is already known. A trip priced in only one program is a single point of failure. When you write down a benchmark, write down the second-best program for it too, and the bank ratio into each.

It also explains why the issuers phrase their lists carefully. Capital One says participating partners are subject to change, and Bilt says the same of its loyalty partners. The list on the day you book is the one that counts, which is one more reason the transfer happens at the booking and not before.

Expiry is a separate question

Saving is mostly about price changes, but expiry rules can force the choice. Chase's Sapphire Preferred announcement says Ultimate Rewards points do not expire as long as accounts are open. Other currencies and programs have their own clocks, often tied to account activity or account closure, and the points expiry guide by issuer sets them out.

The practical rule is that expiry should never be the reason a balance gets spent. If a balance is close to an expiry or closure event, the cheapest fix is usually a small qualifying activity under that program's own rules, not a rushed booking. A rushed booking usually clears no target, which is the cost the keep-or-spend rule exists to avoid.

A closure is the case where expiry and saving collide, so it is worth one worked case. Arithmetic illustration with stated inputs: a holder has 42,000 Ultimate Rewards points on a card they intend to close at its next annual fee, and no trip named. Chase says points do not expire while the account is open, so the deadline is the closure, not a calendar date. Two calm exits exist before that day. One is moving the points to another eligible Chase account under the household rules covered in the pooling and sharing guide. The other is cash back, which Chase prices at 1 cent per point, so 42,000 points become $420 with no booking at all.

That $420 is also the bar for any rushed booking. An award grabbed only to beat the closure has to return more than 1.00 cent per point on its own fare, after fees that survive, or it loses to the cash exit it was meant to avoid. On a $380 fare with $56 in fees and a 30,000-point award, the booking returns ($380 minus $56) divided by 30,000, or 1.08 cents per point: barely above cash, and only if the trip was wanted. The closing a card guide sets out what each issuer says happens to points left on a closed account.

Transfer bonuses and the temptation to wait

Transfer bonuses are the most common reason people give for waiting, and they deserve the same arithmetic as any other change. A bonus is a dated promotion that lowers the bank points an award needs. It does not change the award's price or its availability, and it is never standing policy.

Waiting for a bonus makes sense only when the trip is already named, the seat is not at risk of disappearing, and the bonus partner is one you would use anyway. Otherwise it is a reason to delay a booking that already clears your target. The transfer bonus guide shows how to compute what a bonus does to one booking's cents per point.

And a bonus can never rescue a transfer made too early. It applies to points moved during the promotion; points already moved are already the program's.

What to do with a large balance and no named trip

A large balance with no trip in mind is the case where saving most often goes wrong, because nothing ever triggers the decision. The fix is to create a trigger, not a booking. Choose one or two trips you would genuinely take in the next eighteen months, price them now in every program your bank reaches, and write the prices down with the date. That turns a vague balance into a set of named benchmarks.

Then treat everything beyond those benchmarks as working balance. Any booking that clears your target can use it. When a program change raises a benchmark trip's price, rebook the benchmark sooner if a seat exists; when a better use appears, take it. The balance stops being a store of value and becomes what it is: a set of prices you can buy at today.

The transfers and balance hub collects the related guides on moving, pooling and protecting a balance. If the named trip is a hotel stay at Hyatt or another chart program, the program chart table shows which programs let you price a trip before searching, which is what makes a benchmark possible.

Keep-or-spend check for one balance

• Price the booking in front of you on the bank points it would use.
• Book it if it clears your target and you would take the trip.
• Name any future trip the remaining balance is for, with a season and destination.
• Price that named trip now in every program your bank reaches, and date it.
• Leave the rest at the bank; never transfer to wait.
• Recheck named trips when a program or bank announces a change.
• Handle expiry with a qualifying activity, not a rushed booking.