Unable to obtain a true addition with- odd excel fuction.

imo, it might be the problem of binary conversion to decimal
especially for a decimal point number

Regards.

Typical issue with 64-bit binary floating-point arithmetic (*).

Change the formula in A2 to =ROUND(SUM(A4:A19), 2), if you expect it to be accurate to 2 decimal place.

FYI, not every number in A18 results in "a spurious decimal" in A2. For example, enter 456 in A18, and Excel displays 355.290000000000 in A2.

OTOH, sometimes we cannot see the "spurious decimal", because Excel formats only up to 15 significant digits, rounded.

For example, even when there is nothing in A18 and A2 displays -100.710000000000, there is an infinitesimal difference of about -7.11E-14. We can see that with a formula of the form =SUM(A2,-(A2&"")).

(We cannot write simply =A2-(A2&""), because Excel sometimes substitutes exactly zero for the actual arithmetic difference in that specific context. OTOH, =A2-(A2&"")=0 returns FALSE.)


-----
(*) Explanation....

The problem is: Excel uses a binary form to store numbers and perform arithmetic. And most decimal fractions cannot be represented in that binary form (64-bit binary floating-point). Moreover, the binary approximation of the same decimal fraction might vary, depending on the magnitude of the number. That is why, for example, 10.01 - 10 = 0.01 returns FALSE(!).

In general, the work-around is: when we expect a calculation to be accurate to some number of decimal places, explicitly round to that number of decimal places -- and not to an arbitrary number like 10, as some people suggest.

The operative words are "expect" and "calculation".

There are times when we do not want to round the calculation or constant, but we want to present (format) it with potentially less precision. Only you can make that decision. That is one of many reasons why I discourage setting the "Precision as displayed" option as an alternative to explicit rounding.