Skip to content
All posts
guideUpdated 5 min read

Understanding Spending Caps and Tiered Earn Rates

Many bonus rates have spending limits. How caps, tiers, and choose-your-category cards work — and how to model around them.

Not All Bonus Rates Are Unlimited

When a card advertises "4x on groceries," there’s often an asterisk. Many of the best earn rates come with spending caps — once you hit the limit, the rate drops to 1x. Understanding these caps is critical to building a card setup that doesn’t leave money on the table.

How Spending Caps Work

A spending cap limits the amount of spend that earns the bonus rate in a given period. After you hit the cap, additional spend earns the base rate (usually 1x or 1%).

Common examples:

  • Amex Gold: 4x on groceries up to $25,000/year. After $25K, drops to 1x. That’s ~$2,083/month before the cap kicks in — generous enough for most households.
  • Amex Blue Cash Preferred: 6% on groceries up to $6,000/year. After $6K, drops to 1%. That’s only $500/month — easy to hit if you have a family.
  • Discover it: 5% on rotating categories up to $1,500/quarter. After $1.5K, drops to 1%. That’s $500/month per quarter — tight for heavy spenders in that category.
  • Chase Freedom Flex: 5% on rotating categories up to $1,500/quarter. Same cap structure as Discover.
  • Capital One Savor: 3% on dining, groceries, streaming, entertainment — no cap. This is increasingly rare and valuable.

The Impact of Caps on Annual Earnings

Let’s see how caps affect a heavy grocery spender ($1,000/month):

Amex Blue Cash Preferred (6% up to $6K/year, then 1%):

  • First $6,000: $6,000 x 6% = $360
  • Remaining $6,000: $6,000 x 1% = $60
  • Total: $420/year (effective rate: 3.5%)

Amex Gold (4x up to $25K/year, at 1.5 CPP):

  • Full $12,000: $12,000 x 4 x $0.015 = $720
  • Total: $720/year (effective rate: 6.0% — no cap hit)

Capital One Savor (3%, no cap):

  • Full $12,000: $12,000 x 3% = $360
  • Total: $360/year (effective rate: 3%)

For a heavy grocery spender, the BCP’s 6% rate looks strong on paper, but the $6K cap cuts the effective rate to 3.5%. The Amex Gold’s uncapped 4x at transfer value is higher in this comparison. Even the Savor’s uncapped 3% nearly matches the BCP’s capped 6%.

The lesson: a lower uncapped rate often beats a higher capped rate for heavy spenders.

How Tiered Rates Differ

Some cards don’t just cap — they offer different rates at different spending levels. These tiered structures are more complex.

Bank of America Customized Cash Rewards:

  • Pick one category (gas, online shopping, dining, travel, drug stores, or home improvement)
  • That category earns 3% up to $2,500/quarter ($833/month)
  • Everything else earns 1%
  • With the BofA Rewards bonus (a percentage added on top of base rewards): 3% becomes 3.3% for Members (no minimum balance), 3.75% at Preferred Plus ($30K+), 4.5% at Preferred Honors ($100K+) and 5.25% at Premier ($1M+), while the 1% base rises to 1.1%, 1.25%, 1.5% and 1.75%

With a $100K+ balance in BofA and Merrill accounts (the Preferred Honors tier), this card becomes exceptional — 4.5% in a chosen category with a reasonable $2,500/quarter cap, and 5.25% at the $1M+ Premier tier.

US Bank Cash+:

  • Choose two categories at 5% (up to $2,000/quarter combined)
  • Choose one category at 2% (uncapped)
  • Everything else earns 1%

The Cash+ lets you pick categories each quarter. Popular choices: utilities (5% on electric/gas/water/internet), fast food (5%), select clothing stores (5%). The $2,000/quarter cap is shared across both chosen categories.

How the Optimizer Handles Caps

This is where manual optimization gets genuinely difficult — and where the solver shines.

The CP-SAT solver models each card’s earn structure as segments:

  1. Bonus segment: Earns the bonus rate, limited by the spending cap
  2. Base segment: Earns the base rate, unlimited

When building your modeled setup, the solver:

  • Allocates spend up to the cap at the bonus rate
  • Overflows remaining spend to the next-best card for that category
  • Balances across cards to maximize total earnings minus fees

Example: You spend $800/month on groceries. The solver might:

  • Route the first $500/month to the BCP (6% rate, $6K annual cap)
  • Route the remaining $300/month to the Savor (3%, no cap)
  • After the BCP cap is hit mid-year, shift all $800 to the Savor

This automatic overflow is nearly impossible to calculate by hand across dozens of categories and 5 cards — but the solver handles it in seconds.

Updated October 2, 2026: the optimizer now models dozens of spending categories (an earlier version of this post said 24), quarterly caps are divided by 3 to get the monthly limit (not 4), and the stacking example's comparison figures now use that example's $1,200 a month.

Updated October 11, 2026: Bank of America replaced Preferred Rewards with BofA Rewards, and its bonus tiers changed: Member 10% (no minimum balance), Preferred Plus 25% ($30,000), Preferred Honors 50% ($100,000) and Premier 75% ($1,000,000), added on top of base rewards (checked on Bank of America's BofA Rewards page). The Customized Cash bonus example above now reflects them: 4.5% in a chosen category at $100,000, and 5.25% only at $1,000,000.

Why This Matters for Heavy Spenders

If you spend above average in any category, caps directly affect which cards fit your inputs:

  • $1,000+/month groceries: BCP cap hits in 6 months. You may need a second grocery card (Savor, Amex Gold, or Citi Strata Premier) to maintain bonus earnings for the rest of the year.
  • $800+/month dining: Most dining cards have no cap or a very high one (Amex Gold 4x up to $50,000 a year, Savor 3% with no cap), but BofA Customized Cash caps at $2,500/quarter if you pick dining.
  • $500+/month rotating categories: Both Discover and Freedom Flex cap at $1,500/quarter. You could hold both and effectively double the cap by using each in different quarters.

The Category Stacking Strategy

Category stacking means holding multiple cards that earn bonuses in the same category, using them sequentially as caps are reached.

Grocery stacking example ($1,200/month):

  1. Amex Blue Cash Preferred: Use for the first $500/month (6% rate, fills the $6K annual cap)
  2. Amex Gold: Use for the remaining $700/month (4x MR, cap is $25K/year — plenty of headroom)
  3. After BCP cap hits (month 12): Shift all $1,200 to the Amex Gold

Annual grocery earnings with this stack:

  • BCP: $6,000 x 6% = $360
  • Gold: $8,400 x 4 x $0.015 = $504
  • Total: $864 — vs $432 from a single Savor or $444 from BCP alone ($6,000 x 6% plus $8,400 x 1%)

Stacking requires more cards in your wallet but can be worth hundreds per year for heavy category spenders.

Quarterly vs Annual Caps

Pay attention to the reset period:

  • Quarterly caps ($1,500/quarter for Discover, Freedom Flex): Reset every 3 months. You get a fresh cap each quarter, but the per-quarter limit is lower.
  • Annual caps ($6,000/year for BCP, $25,000/year for Amex Gold): Larger total cap, but no mid-year reset. If you hit it in September, you’re stuck at base rate for the rest of the year.

The optimizer accounts for both by turning every cap into an average monthly limit: annual caps are divided by 12 and quarterly caps by 3.

Find Your Modeled Setup

Caps make manual card comparison unreliable — a card that looks strong at $300/month spending might be wrong at $800/month because of a cap you didn’t notice. Run the optimizer with your actual spending amounts and let the solver model the cap, tier, and overflow rules automatically. The result status tells you whether the highest modeled net value was proven optimal or is the strongest valid setup found within the time limit; a timeout means optimality was not proven.

Not financial advice. OptimalCardSetup provides mathematical optimization tools for educational and informational purposes only. This does not constitute financial, investment, or credit advice. Card rates, fees, and benefits shown were reviewed as of Oct 11, 2026; later corrections are listed on the Corrections page. Terms may change — always verify current details with the card issuer before applying.

Solver examples on this site are bounded by the stated catalog, objective, spending assumptions, redemption values, constraints, and data snapshot. If a run timed out, optimality was not proven; issuer eligibility, approval, terms, and application-channel offers remain controlling.

Ready to model your card setup?

Run the optimizer — free