Deriving e Using the Limit Definition
e is a constant, usually shrouded in mystery, used mainly to terrorize students in finance and calculus classes. It's irrational (meaning it's a non-terminating, non-repeating decimal figure). It is not only similar to π and i; it's inextricably related to them (as proven by Euler).
Every respectable scientific calculator has some approximation of e, and the nicer ones can represent it symbolically (e.g. 3 * e = 3e, rather than 3 * e = 8.15484549). (I have a Casio fx-300ES, and the only gripe I have about it is that it doesn't have a button for π except if you shift first. It represents e and π symbolically.)
So for the person who wants to dig deeper and figure out what this e is all about, let's look at how it's derived using the limit definition:
This looks unpleasant, so let's see where this is going by trying a simpler approach at first.
Consider the case of compound interest: If you invested $1 into an interest-bearing account at 100% APR, in one year your account would be worth $2. What if you were compounding twice a year? Well, then you'd be compounding twice as often at half the APR (once on June 30 at 50%, once on December 31 at 50%). In this case, you'd end up with $2.25 at the end of the year. What about different compounding schedules? A picture is worth a thousand words:
| |
Investment |
APR |
Periodic |
Value as of... |
| Once Yearly |
$1 |
100% |
100% |
December 31 |
$2.00 |
EOY Total |
| |
|
|
|
|
|
|
| Twice Yearly |
$1 |
100% |
50% |
June 30 |
$ 1.50 |
|
| |
|
|
50% |
December 31 |
$ 2.25 |
EOY Total |
| |
|
|
|
|
|
|
| Quarterly |
$1 |
100% |
25% |
March 31 |
$ 1.25 |
|
| |
|
|
25% |
June 30 |
$ 1.5625 |
|
| |
|
|
25% |
September 30 |
$ 1.953125 |
|
| |
|
|
25% |
December 31 |
$ 2.44140625 |
EOY Total |
| |
|
|
|
|
|
|
| Twice Quarterly |
$1 |
100% |
12.5% |
February 15 |
$ 1.125 |
|
| |
|
|
12.5% |
April 30 |
$ 1.265625 |
|
| |
|
|
12.5% |
June 15 |
$ 1.423828125 |
|
| |
|
|
12.5% |
July 31 |
$ 1.601806640625 |
|
| |
|
|
12.5% |
September 15 |
$ 1.80203247070312 |
|
| |
|
|
12.5% |
October 31 |
$ 2.02728652954101 |
|
| |
|
|
12.5% |
November 15 |
$ 2.28069734573364 |
|
| |
|
|
12.5% |
December 31 |
$ 2.56578451395034 |
EOY Total |
| |
|
|
|
|
|
|
| 512 Times Yearly |
$1 |
100% |
100% / 512 |
December 31 |
$ 2.71563200016899 |
EOY Total |
| 1024 Times Yearly |
$1 |
100% |
100% / 1024 |
December 31 |
$ 2.71695572946643 |
EOY Total |
| 8192 Times Yearly |
$1 |
100% |
100% / 8192 |
December 31 |
$ 2.71811593626605 |
EOY Total |
| Ten Million Times |
$1 |
100% |
100% / 107 |
December 31 |
$ 2.71828169413208 |
EOY Total |
| Ten Billion Times |
$1 |
100% |
100% / 1010 |
December 31 |
$ 2.71828205323479 |
EOY Total |
So what's really happening here? Refer back to the limit definition above, and try to think of this as just (1 + 1/n)n. We're infinitely increasing the number of times we compound, and infinitely decreasing the amount we're compounding. In a sense, this is very similar to calculating area using an integral: we're making smaller and smaller slices, but zillions of them.
Before you get too bent out of shape that e is more precise in your calculator than it is calculating it using the limit as n approaches 10 billion, know that people much smarter than I have used computers to calculate it well past 500 billion, and it never terminates and never repeats. (I promise.)
What might really warp the mind is that e is so named for Euler ("Euler's Constant"), though actually a guy named Napier was more directly responsible for it. Euler found an interesting property of it, though (Euler's formula):
eix = cos x + i sin x
...and when you substitute x = π, you get
eiπ = cos π + i sin π
eiπ = -1 + 0
eiπ + 1 = 0
And this elegant little identity shows the fundamental relationship of e, i, and π.
|