Percentage Change Calculator

What does this percentage change calculator compute?

Percentage change measures how much a value moved from an initial (old) amount to a final (new) amount, relative to |V₁|. Positive means increase; negative means decrease. Two of initial, final, and % change are inputs; all three can check consistency; Solve for selects the output. Solving for V₁ can have no, unique, two, or infinitely many solutions because the denominator uses |V₁|.

The published formula or method is % change = (V₂−V₁)/|V₁|×100%. V₂ = V₁(1+P/100) if V₁>0; V₁(1−P/100) if V₁<0. V₁ = V₂/(1+P/100) on the V₁>0 branch; V₂/(1−P/100) on the V₁<0 branch.. The following references support the stated mathematics or method:

• NIST — SI Unit Rules and Style Conventions: The symbol % represents the number 0.01. Thus p% = p/100 and a part of p% of base b equals b×p/100; inverse percentage equations are algebraic rearrangements. CalculatorX’s negative-baseline signed-change convention is separately specified, not attributed to NIST. Mathematical reference only; not implementation validation or an endorsement of CalculatorX.

Which inputs and limits should you check?

Percentage change uses the starting value as its reference. The signed-change and zero-reference cases follow this calculator’s documented convention.

The worked example uses 60 → 72 with 12/|60|×100%. Compare it with the Example calculations and the complete result for the same inputs.

How can you verify and cite this result?

CalculatorX is responsible for this page. The mathematical references above explain the model; implementation checks and their scope are documented separately in the math.percent_change v1.2.0 Evidence and capability contract. Read the displayed verification and production state together with the specification's numerical limits.

Use the clean calculator URL to cite the tool. A shared input link identifies a particular calculation. Content updates, numerical verification, and domain reviews have separate dates and scopes.

What is percentage change?

Percentage change tells you how much a quantity moved from an initial (old) value V₁ to a final (new) value V₂, relative to the start.

• Positive result → increase
• Negative result → decrease (a −25% change is a 25% decrease)

Use it when order matters (“from … to …”). For a directionless comparison of two numbers, use percentage difference instead.

Change vs difference vs error

CalculatorX Percent Error defaults to the absolute-value form, so the result is ≥ 0. A signed form is opt-in.

Signed-change convention

CalculatorX uses |V₁| as the denominator:

% change=(V₂−V₁)/(|V₁|)× 100% (V₁ ≠ 0)

so the sign follows the numerical direction of change when the initial value is negative. A move from −25 to 25 is a +200% increase; from −100 to −50 is a +50% increase (less negative).

This is an explicit convention, not a universal law. Some educational sources also use |Old|. Other statistical treatments divide by V₁ itself and may restrict the formula to a positive baseline.

Inverse formulas

Equivalently, V₂ = V₁ + (P/100) |V₁|. Expanding the absolute value gives the signed branches:

V₂ = V₁(1+P/100) when V₁>0

V₂ = V₁(1−P/100) when V₁<0

Solving for the initial value is the same relation run backward. Each branch is valid only when the solved V₁ has the assumed sign:

V₁ = V₂/(1+P/100) when that V₁>0 and 1+P/100≠ 0

V₁ = V₂/(1−P/100) when that V₁<0 and 1−P/100≠ 0

For V₂=25 and P=+200%, both V₁=25/3 and V₁=−25 satisfy the definition. The calculator returns two valid initial values instead of picking one.

Special cases: V₁=0 is undefined; P=−100% omits the positive-start reverse branch; P=+100% omits the negative-start reverse branch. If V₂=0 and P=−100%, every V₁>0 works; if V₂=0 and P=+100%, every V₁<0 works.

Numerical range and precision

Enter finite binary64 numbers or decimal strings (up to512 characters). Aliases must agree; unknown fields and invalid solve directions are rejected. Inputs round to binary64 first. Formulas are evaluated exactly for those normalized operands, then each published field rounds once, ties to even. The display uses shortest round-trip decimals; it does not track measurement significant figures.

Initial0 is undefined. Decimal-string underflow and nonzero result underflow or overflow, including secondary fields and any valid initial-value branch, return NUMERIC_RANGE: rescale. Raw JSON numeric literals that underflow during parsing return NUMERIC_UNDERFLOW; a partial inverse set is never returned as complete. Near-zero initial values and inverse denominators near P=±100 amplify input rounding. All three inputs check percentages within8ULP only when nonzero with the same sign; zero must match exactly. Solving final preserves the entered P even when the final value rounds unchanged (resolution_limited). An infinite outcome states a positive/negative initial constraint; no-solution is a recoverable error.

Interactive calculations stay local; shared URLs and REST/MCP transmit inputs to the server, and local history may retain results on this device.