Skip to content

Money lending

Let's suppose I lend you some money, £M0, and because I'm so generous I say you can pay me back double at the end of the year. So, after one year, you owe me

M0×(1+1)

Ok, that's a weird way of writing it, but it's useful because I'm going to make you an offer (because I'm so generous, you see): after each month, I'll increase your debt by a mere 112 and that's how much you owe me. Smaller increments, more often - fair is fair. You're welcome.

Read the small print

The wording of our agreement is very important. The amount you owe is

M0(1+112) after 1 month M0(1+112)2 after 2 months M0(1+112)3 after 3 months 

What if it still takes you a year to raise the money to pay me back? Then you owe me

M0(1+112)12

Without a calculator, you might think sure, probably fair. But this is

M0×2.613035290

Are you still happy with this arrangement?

Towards e

What about if I offer to increase your debt by a tiny 1365 each day? Surely you're happier with that? Then again, at the end of the year you owe

M0(1+1365)365=M0×2.714567482

That's still better for the lender, but they might be a little disappointed that the coefficient leaps from 2 to 2.6 (ish) the first time but only 2.6 to 2.7 the second time.

But still, the tendency is to push this further: what about every hour?

M0(1+1365×24)365×24=M0×2.718126692

The lender is a little happier! Every minute?

M0(1+1365×24×60)365×24×60=M0×2.718279243

Every second?

M0(1+1365×24×602)365×24×602=M0×2.718281785

We're making the coefficient grow less and less each time. It's approaching a limit, and that limit is called the number e

The definition of e

We define e as

limn(1+1n)n

If you choose a large number for n, say ten million, your calculator gives you

e2.718281693

whereas the true value of e would be

e=2.718281828

The digits would keep going and would have no pattern to them. In fact, e is an irrational number like π or 2.

The most money the lender could possibly make by increasing the frequency of the compounding is

M0×e

Differentiating ex

We're going to use differentiation from first principles to see why

ddxex=ex

Let f(x)=ex. Then

f(x)=limh0f(x+h)f(x)h=limh0ex+hexh=limh0ex(eh1)h=exlimh0eh1h

Now let k=eh1, so h=ln(1+k) and

k0h0

So

f(x)=exlimk0kln(1+k)

Finally, let n=1k so

k0n

Then we get

f(x)=exlimn1nln(1+1n)=exlimn1ln[(1+1n)n]=ex1ln(e)=ex

That was a lot of work! But we managed to use nothing but the definition of e and some laws of logarithms to prove what we wanted, which is a nice itch to scratch.

The uniqueness of e

Is ex the only type of function that differentiates to itself? Suppose g(x) is another function with the property that g(x)=g(x) and consider, by the quotient rule,

ddxg(x)ex=exg(x)exg(x)e2x=exg(x)exg(x)e2x=0

which implies

g(x)ex=k

for some kR. So g(x) must be a multiple of ex. This means e really is special! (Note: if we include the property that g(0)=1 then k=1 so g(x)=ex.)

Conclusions

The financial origins of the number e are interesting, and can be adapted into a fun task where students are led to the value of e by greedily trying to swindle more and more money from their customers.

The fact that the fundamental property of ex follows from this definition and first principles is pretty astonishing, but it is to be hoped for because these properties should depend only on the definitions. The working is probably a bit too technical to be enjoyable for most A Level students, but it could be demonstrated at break time if anyone ever asks.

Finally, ex is somehow unique amongst all functions for having this property, which should convince anyone of its importance. π and ϕ steal a lot of the limelight, but e is by far my favourite number.