- The derivative of \( e^x \) is \( e^x \), and the derivative of \( \ln x \) is \( \frac{1}{x} \). Almost everything else follows from these two facts and the chain rule.
- With the chain rule: \( \frac{d}{dx}e^{f(x)} = f'(x)e^{f(x)} \) and \( \frac{d}{dx}\ln f(x) = \frac{f'(x)}{f(x)} \).
- Other bases bring in a natural log: \( \frac{d}{dx}a^x = a^x \ln a \) and \( \frac{d}{dx}\log_a x = \frac{1}{x \ln a} \).
- In economics, \( \frac{d}{dt}\ln Y \) is the growth rate, and \( \frac{d \ln y}{d \ln x} \) is an elasticity.
- Exam point: use log laws to simplify before you differentiate a logarithm. It saves time and errors.
The case: how fast is the economy growing?
An economy’s output, in billions of dollars, is modelled by \( Y(t) = 1500e^{0.03t} \), where \( t \) is years from now. (The figures are illustrative.) A finance ministry wants three answers. How fast is output rising today, in dollars per year? How fast will it be rising in ten years? And what is the growth rate, in per cent per year?
Today output is $1,500 billion. In ten years it will be \( 1500e^{0.3} \approx 2{,}024.79 \) billion. The rate of change in dollars per year keeps rising as the economy gets bigger. But the growth rate stays fixed at 3% a year. To see why, and to get the numbers, you need the derivative of the exponential function. Then, to turn “dollars per year” into “per cent per year”, you need the derivative of the natural logarithm.
This post covers both, with the chain rule, other bases, and three applications that exams and economics courses love: continuous growth, elasticity and optimisation.
The derivative of eˣ
The first rule says the height of the curve and its slope are the same number. At \( x = 0 \), \( e^0 = 1 \), so the tangent there has gradient 1: it is the line \( y = x + 1 \) in Figure 1. The second and third rules come from the chain rule: differentiate the outside (\( e^{u} \) stays as \( e^{u} \)), then multiply by the derivative of the inside [1].
Back to the case. \( Y(t) = 1500e^{0.03t} \), so \( Y'(t) = 1500 \times 0.03\,e^{0.03t} = 45e^{0.03t} \). Today, \( Y'(0) = 45 \): output is rising at $45 billion a year. In ten years, \( Y'(10) = 45e^{0.3} \approx 60.74 \) billion a year. Divide the rate by the level: \( \frac{Y'(t)}{Y(t)} = \frac{45e^{0.03t}}{1500e^{0.03t}} = 0.03 \). The growth rate is 3% every year.
Other bases: aˣ
What about \( 2^x \) or \( 5^x \)? Rewrite the base using \( e \): since \( a = e^{\ln a} \), we have \( a^x = e^{x \ln a} \). Now use the \( e^{kx} \) rule with \( k = \ln a \):
When \( a = e \), \( \ln e = 1 \) and the rule collapses back to \( e^x \). That is exactly why \( e \) is the “natural” base: it is the one where the extra factor disappears.
The derivative of ln x
Let \( y = \ln x \). Then \( x = e^{y} \), so \( \frac{dx}{dy} = e^{y} = x \). Turn it upside down: \( \frac{dy}{dx} = \frac{1}{x} \). The same argument, with the chain rule, gives the general form:
In Figure 1, the tangent to \( \ln x \) at \( (1, 0) \) has gradient \( \frac11 = 1 \): it is the line \( y = x – 1 \), the mirror image of the tangent to \( e^x \) at \( (0, 1) \). For other bases, change base first: \( \log_a x = \frac{\ln x}{\ln a} \), and \( \ln a \) is just a constant, so
Before differentiating any logarithm, simplify with the log laws: \( \ln(ab) = \ln a + \ln b \), \( \ln\frac{a}{b} = \ln a – \ln b \), \( \ln a^n = n\ln a \). Our free guide on exponentials and logarithms revises these laws.
Summary table of rules
| Function | Derivative | Example |
|---|---|---|
| \( e^{x} \) | \( e^{x} \) | gradient at \( x = 0 \) is 1 |
| \( e^{kx} \) | \( ke^{kx} \) | \( \frac{d}{dx}e^{3x} = 3e^{3x} \) |
| \( e^{f(x)} \) | \( f'(x)e^{f(x)} \) | \( \frac{d}{dx}e^{x^2} = 2xe^{x^2} \) |
| \( a^{x} \) | \( a^{x}\ln a \) | \( \frac{d}{dx}5^{x} = 5^{x}\ln 5 \) |
| \( \ln x \) | \( \frac{1}{x} \) | gradient at \( x = 1 \) is 1 |
| \( \ln f(x) \) | \( \frac{f'(x)}{f(x)} \) | \( \frac{d}{dx}\ln(3x – 2) = \frac{3}{3x – 2} \) |
| \( \log_a x \) | \( \frac{1}{x\ln a} \) | \( \frac{d}{dx}\log_{10}x = \frac{1}{x\ln 10} \) |
Applications: growth rates, elasticity and optimisation
Growth rates. For any positive quantity \( Y(t) \), \( \frac{d}{dt}\ln Y = \frac{Y'(t)}{Y(t)} \), the proportional growth rate. That is why economists plot GDP on a log scale: a straight line means a constant growth rate. With \( Y = 1500e^{0.03t} \), \( \ln Y = \ln 1500 + 0.03t \), whose derivative is 0.03. The doubling time is \( \frac{\ln 2}{0.03} \approx 23.1 \) years.
Elasticity. Price elasticity of demand is \( \frac{dQ}{dP}\cdot\frac{P}{Q} \), which is the same as \( \frac{d\ln Q}{d\ln P} \). For a constant-elasticity demand curve \( Q = 500P^{-1.5} \), taking logs gives \( \ln Q = \ln 500 – 1.5\ln P \), so the elasticity is \( -1.5 \) at every price. For \( Q = 200e^{-0.05P} \), \( \ln Q = \ln 200 – 0.05P \), so \( \frac{d\ln Q}{dP} = -0.05 \) and the elasticity is \( -0.05P \). Demand is unit elastic at \( P = 20 \), which is exactly where revenue \( R = 200Pe^{-0.05P} \) is highest, at about $1,471.52 [4]. See our price elasticity of demand post for the economics.
Optimisation. Functions like \( xe^{-x} \) rise then fall, and they model things such as a drug’s concentration in the blood or the sales of a product after launch. You find the peak with the product rule and the exponential rule together, as in the third worked example.
Worked examples
Differentiate (a) \( e^{3x} \), (b) \( 5^{x} \), evaluating the gradient at \( x = 2 \), and (c) \( 2^{3x} \).
Every non-\( e \) base produces a \( \ln a \) factor; every inside function produces a chain-rule factor.
Differentiate (a) \( \ln(x^2 + 1) \) and find the gradient at \( x = 1 \); (b) \( \ln\sqrt{x^2 + 4} \); (c) \( \log_{10}x \) at \( x = 100 \).
Using \( \ln a^n = n\ln a \) in part (b) removed the square root before any differentiation.
Find the maximum value of \( f(x) = xe^{-x} \) for \( x \ge 0 \), and confirm it is a maximum.
The function peaks at \( x = 1 \) with value \( \frac{1}{e} \). Every answer above was checked with a computer algebra system.
For the full path from indices and functions through limits and derivatives to their applications, with 282 worked examples, see the Mathematics Made Simple textbook. To build speed, the Mathematics Made Simple Practice Questions book has exam-style questions in AP Calculus, Cambridge 9709 and IB formats, every one with a fully worked answer. Revise the chain rule first in the chain rule step of our free guide.
Practice questions
EasyQ1. Differentiate \( y = 4e^{2x} \). Show worked answer
EasyQ2. Differentiate \( y = \ln(3x – 2) \), for \( x > \tfrac23 \). Show worked answer
\( \frac{dy}{dx} = \frac{3}{3x – 2} \).
MediumQ3. Find the gradient of \( y = 3e^{4x} – \ln(5x) \) at \( x = \tfrac14 \). Show worked answer
At \( x = \tfrac14 \): \( 12e – 4 \approx 32.62 – 4 = 28.62 \).
MediumQ4. Find the stationary point of \( y = x^2\ln x \) for \( x > 0 \) and determine its nature. Show worked answer
Set to zero: \( \ln x = -\tfrac12 \), so \( x = e^{-1/2} \approx 0.607 \).
\( y = e^{-1} \times (-\tfrac12) = -\frac{1}{2e} \approx -0.184 \).
\( \frac{d^2y}{dx^2} = 2\ln x + 3 = -1 + 3 = 2 > 0 \): a minimum.
ExamQ5. Monthly sales of a new app, in thousands, are modelled by \( S(t) = 20te^{-0.5t} \), where \( t \) is months after launch. (a) Show that \( S'(t) = 10(2 – t)e^{-0.5t} \). (b) Find when sales peak and the peak value. (c) Show that this is a maximum. (d) Find the rate of change of sales at \( t = 4 \) and interpret it. [8] Show worked answer
(b) \( e^{-0.5t} \ne 0 \), so \( t = 2 \) [1]; \( S(2) = 40e^{-1} \approx 14.72 \) thousand [1].
(c) \( S”(t) = 5(t – 4)e^{-0.5t} \) [1]; \( S”(2) = -10e^{-1} \approx -3.68 < 0 \), so a maximum [1].
(d) \( S'(4) = -20e^{-2} \approx -2.71 \) [1]: sales are falling by about 2,710 per month at four months [1].
Frequently asked questions
Why is the derivative of eˣ equal to eˣ?
Because \( e \) is defined as the base for which the exponential’s gradient at \( x = 0 \) is exactly 1. For any exponential, the gradient at each point is the height times the gradient at zero, so for base \( e \) the gradient equals the height everywhere.
What is the derivative of ln(kx)?
It is \( \frac{1}{x} \), for any positive constant \( k \). Since \( \ln kx = \ln k + \ln x \) and \( \ln k \) is a constant, only \( \ln x \) contributes.
Do I need log base 10 in calculus exams?
Rarely, but you should know \( \frac{d}{dx}\log_{10}x = \frac{1}{x\ln 10} \). Calculus works most naturally in base \( e \), so convert other bases using \( \log_a x = \frac{\ln x}{\ln a} \).
What is logarithmic differentiation?
Taking logs of both sides before differentiating, useful for functions like \( x^x \) or long products. For \( y = x^x \): \( \ln y = x\ln x \), so \( \frac{1}{y}\frac{dy}{dx} = \ln x + 1 \) and \( \frac{dy}{dx} = x^x(\ln x + 1) \).
References
- Strang, G. and Herman, E. (2016) Calculus Volume 1. Houston, TX: OpenStax. Section 3.9, Derivatives of exponential and logarithmic functions. Available at: https://openstax.org/books/calculus-volume-1/pages/3-9-derivatives-of-exponential-and-logarithmic-functions.
- College Board (n.d.) AP Calculus AB and AP Calculus BC Course and Exam Description. New York: College Board. Available at: https://apcentral.collegeboard.org/media/pdf/ap-calculus-ab-and-bc-course-and-exam-description.pdf.
- Cambridge International (n.d.) Cambridge International AS & A Level Mathematics 9709: Syllabus for examination in 2026 and 2027. Cambridge: Cambridge University Press & Assessment. Available at: https://www.cambridgeinternational.org/Images/697427-2026-2027-syllabus.pdf.
- Sydsæter, K., Hammond, P., Strøm, A. and Carvajal, A. (2016) Essential Mathematics for Economic Analysis. 5th edn. Harlow: Pearson.

