Rate of change: the formula, the multiplier and successive changes
A rate of change is a subtraction followed by a division. Everything else (the multiplier, chained changes, the annual average) follows from that one formula.

TL;DR
A rate of change is the difference divided by the starting value: (final − initial) ÷ initial × 100. It is equivalent to a multiplier: 1.08 for an 8% rise. Changes that follow one another multiply their multipliers; they do not add.
The formula, and what it actually measures
A rate of change answers one question: how far has a value moved, relative to where it started?
rate of change = (final value − initial value) ÷ initial value × 100
A rent going from €850 to €879.75 changed by (879.75 − 850) ÷ 850 × 100 = 3.5%. The percentage increase calculator does the sum and shows the working. A turnover going from €1,200,000 to €1,080,000 changed by −120,000 ÷ 1,200,000 × 100 = −10%. The sign carries the whole direction: positive is a rise, negative a fall.
The important word in the formula is initial. The denominator is the starting value, never the ending value and never the average of the two. That is where the behaviour that surprises people most comes from, and it is not a quirk but the definition itself1.
In France this quantity has its own standing name, taux d'évolution, and it is taught as a topic in its own right rather than as a special case of percentages. The vocabulary below is that French vocabulary, because it is what you will meet on a French payslip, rent notice or INSEE release.
Why a rise and the fall that answers it are not the same rate
Going from 250 to 300 is a 20% rise. Coming back from 300 to 250 is not a 20% fall: it is a fall of 16.67%.
- 250 to 300: 50 ÷ 250 = +20%
- 300 to 250: −50 ÷ 300 = −16.67%
The difference in euros is identical, 50. The rate differs because the reference moved. That is how a headline can say, without lying, that a price "rose 20% then fell 17%" and describe an almost exact return to where it began.
The practical consequence: a rate of change only means something together with its direction of reading. "A 20% difference between the two" is an ambiguous sentence; "20% more than last year" is not.
The multiplier: the same thing, written differently
Applying an 8% rise means multiplying by 1.08. Applying a 20% fall means multiplying by 0.80. In France that number has a name of its own, the coefficient multiplicateur, and it is the real working tool1.
coefficient = 1 + rate ÷ 100 → rate = (coefficient − 1) × 100
| Rate of change | Multiplier |
|---|---|
| + 8% | × 1.08 |
| + 100% | × 2 |
| 0% | × 1 |
| − 20% | × 0.80 |
| − 100% | × 0 |
Once the change is written as a multiplier, the rest of this guide becomes obvious arithmetic: to chain changes you multiply the coefficients, to go backwards you divide by the coefficient, and for an average rate over n periods you take the nth root of the overall coefficient. Three operations, one object.
Going back to the earlier value: a division
You know the amount after the change and the rate that was applied, and you want the amount before. You divide by the multiplier.
A rent of €879.75 after a 3.5% rise was 879.75 ÷ 1.035 = €850.
The temptation is to take 3.5% off the new amount. That gives €848.96, which was never anybody's rent. The error is 79 cents here, because the rate is small; on a 20% rise it becomes crude:
- €100 + 20% = €120
- €120 − 20% = €96 ← wrong
- €120 ÷ 1.20 = €100 ← correct
The 20% had been computed on 100, not on 120, so taking it off the total takes off too much. It is exactly the sum that converts a TTC (VAT-inclusive) price back to the HT one, which the VAT calculator does, and it holds for any rate and in both directions.
Successive changes multiply, they do not add
This is the point where reasonable intuition always gets it wrong.
| Two changes | Working | Overall rate |
|---|---|---|
| + 10% then + 10% | 1.10 × 1.10 = 1.21 | + 21% |
| + 20% then − 20% | 1.20 × 0.80 = 0.96 | − 4% |
| − 30% then − 10% | 0.70 × 0.90 = 0.63 | − 37% |
| + 5% then − 10% | 1.05 × 0.90 = 0.945 | − 5.5% |
The second row is the one worth memorising: going up and then down by the same percentage does not return you to the start, but always below it2. And the third explains why "30% off in the window, another 10% at the till" is not a 40% discount: €100 less 30% is €70, and less another 10% is €63, which is precisely what the discount calculator works out.
Two properties are worth knowing. The order does not change the overall rate (multiplication is commutative), but it does change the intermediate amounts, which matters as soon as a threshold, a cap or a rounding step intervenes along the way. And a −100% step absorbs everything after it: the overall multiplier becomes zero and no later rise lifts it.
An average annual rate is a geometric mean
A price has doubled in five years. By how much did it rise per year?
The intuitive answer (100% divided by 5, so 20% a year) is wrong, and one line checks it: five rises of 20% give 1.20⁵ = 2.49, not 2. The right answer is the fifth root of 2:
average rate = (overall multiplier)^(1/n) − 1 → 2^(1/5) − 1 = 14.87% a year
And 1.1487⁵ does give 2. The average rate is a geometric mean, because each period applies to the result of the one before. The arithmetic mean overstates systematically: that is a mathematical inequality, not a tendency, and the wider the swings and the more numerous the periods, the more wrong it is.
Two reading cautions. An average rate is a reconstruction: it describes the smoothed path from one end to the other, not the path actually travelled, and two very different series can share the same average rate. And it requires a strictly positive starting value: from a start of zero there is no multiplier, so there is no mean.
When there is no rate of change
Two cases have no answer, and it is better to know it than to read an invented number.
The start is zero. Going from 0 to 300 is not a rise of 100% or of 300%: the formula divides by zero and the result does not exist. It is an appearance, not a change, and the honest way to put it is in absolute terms: "300 more, against nothing last year".
The fall reaches or passes 100%. If you look for the value before a 100% fall, there is not one: every starting value ends at zero, so the information is gone. Beyond a 100% fall the multiplier turns negative and the reverse calculation returns a value with the wrong sign.
One last trap of vocabulary, which is an ambiguity rather than an impossibility: when the quantity that changes is itself a percentage, a rate moving from 2% to 3% gains one percentage point and its rate of change is +50%. Both statements are correct; the second sounds far more dramatic, which explains its popularity. The four other questions that all go under the name "percentage", of which this is only one, are gathered in how to calculate a percentage.
Common mistakes
Reading a rise and the fall that answers it as the same percentage
250 to 300 is +20%, but 300 back to 250 is −16.67%: the reference has changed.
Adding successive rates
+10% then +10% is +21%. You multiply 1.10 × 1.10; you do not add the rates.
Subtracting the percentage to find the earlier value
After a 20% rise you divide by 1.20. Taking 20% off the new amount takes off too much.
Dividing the total change by the number of years to get an average rate
The average rate is the nth root of the overall multiplier; the arithmetic mean always overstates.
Frequently asked questions
How do you calculate a rate of change?
What is the coefficient multiplicateur?
How do you combine successive percentage changes?
How do you calculate an average annual rate of change?
Can a rate of change exceed 100%?
What is the difference between a rate of change and a percentage point?
Related reading & calculators
Sources
Author / Reviewed by
Author
Thorben Rasmus Idel
Co-founder & writer
Co-founder of Calculadora Capital and the writer behind the methodology on every calculator and article. An entrepreneur and active investor, Thorben founded Idel Versandhandel GmbH, an international trading company operating across 16 countries, and invests across stocks, ETFs and cryptocurrency. He writes the methodology and verifies the math behind each page, drawing on hands-on business and investing experience to keep the tools and explanations grounded in how money, markets and taxes actually work for everyday people in France.
Reviewed by
Nahar Geva
Co-founder & reviewer
Co-founder of Calculadora Capital and the independent reviewer behind every calculator and article. An entrepreneur and active investor, Nahar brings a data- and product-driven mindset together with hands-on experience in the markets, investing across stocks and ETFs as well as cryptocurrency and other digital assets, alongside broader personal finance and real estate. On each page Nahar reviews the methodology and double-checks the math and figures, pressure-testing how the tools and explanations hold up against the way money, markets and taxes actually work for everyday investors.
Published: Updated: Reviewed: