← All field notes

A YEAR IN VIEW · 4 min read

What an annual cash-back estimate leaves out

Our annual estimate repeats the month you enter twelve times and starts with unused annual caps. It compares a full year of eligible spending; it does not predict your remaining rewards this year or the cash arriving on your next statement.

Give one-off spending its real weight.

Suppose you normally spend $200 per month in a category, then make a planned $1,200 purchase once during the year. The annual category total is $200 × 12 + $1,200 = $3,600. Dividing by 12 gives a representative monthly input of $300.

Entering the unusually expensive $1,400 month would instead model $16,800 per year. At a hypothetical flat 2% rate, that shows $336 rather than $72 from the actual $3,600 plan. The problem is repeating a one-time purchase twelve times.

Build a planned full-year total for each category, then divide each by 12. Past statements can help you estimate, but account for changes you already expect. Round the monthly input to cents; dividing some annual totals by 12 introduces a small rounding difference. Enter each purchase in one category only.

Seasonal timing alone does not change this model’s full-year total.

If the same eligible purchases fall within one calendar year, the rates stay the same, and the category has one annual cap, spending them unevenly through that year does not change the modeled annual rewards. The category total matters. Timing matters when purchases move between cap years, terms change, or you are estimating only the remaining part of a year.

For a hypothetical category earning 3% on the first $6,000 each calendar year and 1% afterward, $7,200 in one year earns $180 + $12 = $192. Split the same total into $3,600 in each of two fresh calendar years and it earns $108 + $108 = $216. That is a different time window with two caps. Our single-year estimate does not model it.

A fresh cap is not your remaining cap.

Using that same hypothetical rule, suppose you have already spent $5,500 in the category this calendar year and expect another $1,500 before year-end. Only $500 of bonus capacity remains:

  1. Remaining bonus capacity: $6,000 − $5,500 = $500
  2. Next $500 × 3% = $15
  3. Other $1,000 × 1% = $10
  4. Rewards on the remaining purchases = $25

Treating all $1,500 as fresh-cap spending would show $45. That overstates the remaining rewards by $20. These amounts exclude rewards on the $5,500 already spent. Our calculator has no year-to-date spending input, so use your issuer’s cap tracker and this separate calculation for a remainder-of-year question. See how spending caps work for the full-year calculation.

Rewards earned and cash received are different clocks.

Citi Double Cash earns 1% when you purchase and another 1% as you make qualifying payments. Our full-payment assumption includes both portions; it does not mean both arrive when you buy. A purchase near year-end and its later payment can fall in different periods. Citi earning rules ↗

Redemption takes time too. PNC’s terms allow up to seven business days to apply a valid statement-credit or deposit redemption request, and returns reduce earned rewards. Its rules also round purchase rewards to cents; our model rounds each annual category total. PNC rewards and redemption terms ↗

Use the estimate for the question it can answer.

For “How would these cards compare over a representative full year?”, enter average monthly categories and inspect the annual breakdown. For “How much will I receive before December 31?”, check posted purchases, remaining caps, qualifying payments, and redemption timing with your issuer. Do not divide the annual estimate by twelve and treat it as a statement forecast.

All examples assume eligible purchases, full payment, and no interest or other fees. Keep spending within your existing plans. The methodology explains the model’s boundaries; the current-card comparison note shows how to turn a full-year estimate into a potential gain.

Explore the cards behind the examples

These selected cards illustrate the comparison; their detail pages show modeled rates, exclusions, and issuer sources.

Put it into practice