How to Value Amex Points: CPP Data

The Points Guy values an American Express Membership Rewards point at 2.0 cents in its April 2026 valuation. American Express will hand you 0.6 cents for the same point as a statement credit. That 3.3x spread is the entire game — and it is why “how much are Amex points worth” has no single honest answer.

Every valuation you have seen quoted is a behavioral assumption disguised as a number. TPG’s 2.0 cents per point (CPP) assumes you transfer to airline partners and book premium-cabin awards. NerdWallet’s figure assumes something more modest. Amex’s own portal assumes you never leave its website. The valuation is not a property of the point; it is a property of how you redeem it. A household earning $150k+ that defaults to “pay with points” at checkout is quietly accepting 0.7 cents on an asset the optimizers price at 2.0.

The numbers, before the argument

Sources: The Points Guy Membership Rewards valuation (April 2026); NerdWallet Amex points analysis (2026); American Express published redemption terms and Platinum Card terms (April 2026). CPP = cents per point.

Read that table as a ladder, not a list. The same point is worth anywhere from 0.6 to 2.0 cents depending solely on your exit. Most coverage reports the top rung and stops. The useful question is which rung you actually stand on.

Why three sources disagree by 3x

TPG, NerdWallet, and Amex are not measuring the same thing. They are measuring three different redemption behaviors and reporting each as “the” value.

TPG arrives at 2 cents per point through transfer partners by assuming you move points to airline programs and book aspirational awards — business class to Europe, Virgin Atlantic Saver fares, Hilton luxury nights. TPG values Membership Rewards points at 2.0 cents each in its 2025 valuations and frames that as the floor of a “good” redemption rather than the average. High-value redemptions should exceed 2 cents per point, while lower-value awards return less than 1 cent. The 2.0 figure is a target, not a forecast.

NerdWallet is more conservative. Its estimate puts the overall transfer value at 1.6 cents per point, with sharply lower numbers everywhere else: roughly 1 cent per point redeemed for flights directly through Amex, and about 0.6 cents per point for statement credits — a rate it explicitly advises avoiding. The gap between NerdWallet’s 1.6 and TPG’s 2.0 is not a contradiction. It is two firms drawing the “typical optimizer” line in slightly different places.

Then there is Amex’s own math, which is the most honest about what most people actually do. Flights booked through American Express Travel redeem at 1 cent per point; prepaid hotels and other portal travel at 0.7 cents; covering charges on your statement at 0.6 cents per point. Pay With Points at linked merchants like Amazon runs about 0.7 cents per point. No transfer skill required, and no pretense that you will book ANA first class.

The valuation you should use is the one matching your laziness

This is the part most guides bury. Pick the row that describes your real behavior over the last twelve months, not the behavior you aspire to. If you have never once transferred points to an airline program and booked an award seat, your Amex point is worth 0.6 to 1.0 cents — full stop. The 2.0 figure belongs to someone else. The comparison between Chase and Amex points turns on the same self-honesty: flexibility is only worth something if you use it.

What a point is worth, by redemption path

Sources: American Express published redemption schedule (April 2026); NerdWallet Amex Membership Rewards review (2026); The Points Guy valuation (April 2026). Transfer-partner figures are estimates and assume award availability; statement-credit and portal figures are fixed by Amex.

The vertical distance in that table is $1,400 on a 100,000-point balance. Same points, same account, same day — a four-figure swing decided entirely by which button you press. For a $150k+ household running six-figure annual card spend, balances of 100,000+ points accumulate fast, and the redemption discipline compounds into real money.

One caution on the top two rows. Transfers are irreversible, so the cash fare should always be compared against the award price before moving points, and fuel surcharges on some partners run close to $400 per person round-trip in business class. The 2.0 CPP headline quietly assumes you dodge the bad redemptions and the surcharge traps. Net of those, realized value drifts down.

The Finluxy Card Net Annual Value test

Valuation in the abstract is a parlor game. The number that decides whether to keep a card is net of its fee. The Finluxy Card Net Annual Value formula: total annual rewards earned (points at a stated CPP) plus annual credits actually used, minus the annual fee.

The hinge for the Amex Platinum is the fee, and the fee moved. The Amex Platinum annual fee rose from $695 to $895, the first increase since 2021. The $895 fee was implemented in September 2025; existing cardholders see it at their next renewal on or after January 2, 2026. Any net-value model still running on $695 is wrong by $200 before it starts.

Here is the metric calculated across the realistic CPP range, modeled at a representative $150k household card spend earning roughly 150,000 Membership Rewards points over a year (5x on flights and prepaid hotels through Amex Travel, 1x on everything else), with credits set at a defensible mid-case. Because model-specific spend distribution varies by household, the points-earned figure is a representative estimate; the methodology below lets you substitute your own.

Finluxy Card Net Annual Value = (points earned × CPP) + credits used − annual fee. Annual fee $895 (American Express, eff. Sept 2025). CPP figures: American Express (0.6, 1.0), NerdWallet (1.6), TPG (2.0), all 2026. Points-earned and credits-used are representative estimates for a $150k+ household; substitute your own figures using the methodology below. Credits-used set at $1,000 as a mid-case against the card’s $3,500+ face value, reflecting realistic partial utilization.

Two things fall out of this. First, even at the worst redemption rate, the Platinum clears positive net value for a high-spend household — but only because the credits do the heavy lifting, not the points. Strip the credits and the 0.6 CPP row collapses to +$5. The card is a credit-stacking vehicle that happens to earn points, not a points engine that happens to have credits. Whether you keep it should hinge on the $895 fee justification math, and whether you actually burn the statement credits the card bundles in.

Second, the credits assumption is doing more than the CPP assumption. The realistic spread on credits used — from near-zero for an occasional traveler to the full $3,500+ face value — dwarfs the points spread. Amex Platinum credits add up to more than $1,500 in face value if fully used, but realistic outcomes for most cardholders fall between zero and that ceiling. That is the variable to stress-test before the CPP.

What the data shows that most coverage misses

Here is the finding buried in plain sight: the most-quoted Amex valuation and the most-used redemption method are nowhere near each other, and the gap is structural, not behavioral.

TPG’s 2.0 CPP gets cited everywhere because it is the highest defensible number and it sells the dream. But Amex’s own data on what redemptions actually pay tells you the realized average sits far lower. NerdWallet flatly advises against statement-credit redemptions because they return just 0.6 cents — which is precisely the redemption Amex makes easiest and most visible in the app. The architecture nudges you toward the worst rate. The 2.0 valuation requires you to actively resist the path of least resistance, learn transfer partners, hunt award availability, and accept irreversibility. Most people don’t, which means the “value” of a Membership Rewards point, measured across the actual population of redemptions, is materially below every headline figure. The number isn’t lying. The number describes a discipline most cardholders never exercise.

For the $150k+ household: where the threshold actually sits

At six-figure income and six-figure card spend, the points math is rarely the binding constraint — the fee-versus-credits math is. A household clearing $150k+ likely generates enough flights, hotels, and dining to absorb the Platinum’s credit stack, which is what keeps the Finluxy Card Net Annual Value positive even at 0.6 CPP. The decision is not “are Amex points worth it.” It is whether your spend pattern triggers the credits you are pre-paying for in the $895 fee.

Two thresholds matter. If you reliably use $1,000+ in credits and earn 100,000+ points a year, the card clears its fee comfortably regardless of redemption skill — the question becomes whether a lower-fee competitor would clear it by more. If your credit utilization runs under roughly $600 and you redeem at statement-credit rates, you are paying $895 to earn points worth less than a no-annual-fee 2% cash-back card would return on the same spend, and the rational move is to evaluate the downgrade threshold. The transfer-partner premium only justifies the fee if you genuinely redeem there; for households that don’t, the honest comparison is against the Venture X net-value math or the full Sapphire Reserve versus Platinum breakdown. The broader build is covered in the card strategy for $150k+ household spending.