If your total is off by 0.01, or shows a tiny stray decimal such as 0.00000000001, floating point is one possible cause. Excel stores most decimals as binary approximations. Wrap your amounts and totals in ROUND(...,2) so rupees and paise are exact, then use a difference cell to prove your sum matches the statement.
Why do sums of decimals drift at all?
Excel keeps numbers in binary, and a figure like 0.1 cannot be written exactly in binary, just as one third cannot be written exactly in decimal. Add many such figures and the tiny leftovers can surface. You might see a total that looks right but fails an equality test such as =E2=F2, or a difference cell that shows 0.0000000001 instead of 0. Nothing is wrong with your data; the arithmetic is slightly untidy.
How is a real paisa error different?
Look at the size of the difference.
| Difference you see | Likely cause | What to do |
|---|---|---|
| Around 0.0000000001 or smaller | Floating point noise | Round the amounts and totals to two places |
| Exactly 0.01 or 0.02 | Rounding at source, such as a bank rounding each line | Compare line by line with the statement |
| A whole transaction amount | Missing or duplicated row | Compare row counts and sort by amount |
| A figure that looks random | Text amounts ignored by SUM | Test with COUNT versus COUNTA |
The fix: ROUND in the amount column
Do your rounding where the amount enters the sheet, then everything built on top of it is clean.
- Amount column: =ROUND(VALUE(SUBSTITUTE(C2,",","")),2). If the amounts are already numbers, =ROUND(C2,2) is enough.
- Totals: =ROUND(SUM(E2:E500),2).
- SUMIFS totals: =ROUND(SUMIFS($E$2:$E$500,$B$2:$B$500,"Food"),2).
Why two places? Rupees have paise, so two decimals is the smallest unit you actually spend. Rounding to fewer would change real money; rounding to more would keep the noise.
How do you check the sum against the statement?
Type the statement's printed total, for example total debits, in a cell, say H1. Then in H2 enter =ROUND(SUM(E2:E500)-H1,2). Read the result:
- 0 means your sheet agrees with the statement.
- 0.01 or another multiple of 0.01 means a real amount is slightly different somewhere.
- A large figure means a missing row, a duplicate or a text amount.
A visible flag helps: =IF(H2=0,"Ties to statement","Check rows"). Note that without ROUND, H2 can fail the equality test even though the figures look identical, which is the issue we are fixing.
Worked example with hypothetical numbers
Say three amounts are 1,250.10, 2,349.90 and 400.00 and the statement total is 4,000.00. Without rounding, Excel's internal addition might yield something like 4000.0000000000005; the cell displays 4,000.00, yet an equality test against the statement figure can return FALSE. With =ROUND(SUM(E2:E4),2) the answer is held at exactly 4000, and the comparison behaves.
Is it ever the statement that is off?
Occasionally, yes. Bank statements can round per line, show amounts after conversion from another currency, or list a charge on a different day from the one in your notes. If your amounts are clean numbers, row counts match and the gap is a tidy multiple of 0.01, ask the bank or compare a few lines to the printed originals instead of forcing a match.
Can a paisa difference be genuine?
Yes. Say a ₹1,000 bill is shared equally among three people. Each share is 333.33 when rounded to paise, and three shares add to 999.99. The missing paisa is real, not an Excel fault. Decide in advance where the leftover goes, for example to the largest share, and note it beside the figures so you remember next time.
Before you finish
- Add ROUND at the point of entry for each amount.
- Keep a Difference cell on every monthly sheet.
- Investigate any gap above zero rather than hiding it.
If you want the check built in, SpendixAI shows imported rows for review before they are saved and flags possible duplicate imports; your own sheet remains a good second look.
Useful ideas start with a clearer picture of your own spending.
Explore your reports

