THE POWER OF COMPOUNDING INTEREST
A Problem: 21 yr old vs 31 yr old (simple interest solution)
Yearly vs Monthly Contributions
Continuous Model: 21 yr old vs 31 yr old
Continuous Model vs Simple and Continuous Interest
Cyclic Interest Rates: Annual and Monthly Contributions vs Modified ODE Continuous Model
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Yearly vs Monthly Contributions
Now,
let’s take a closer look at Adam’s first 10 years with the account.Let’s
compare his annual contributions to monthly, and then continuously (with
continuous compounding interest).
Here
we have:
·One
lump sum of $2400 per year (at the end) for 10 years at 8% interest:
So, let P=2400 and
r =.08.
Our calculation
is of the form P[(1+r)^9 + (1+r)^8 + … + (1+r) +1]
So, 2400[(1+.08)^9
+ (1+.08)^8 + … + (1+.08) +1]
= 2400[(1-(1+.08)^10)/(1-(1+.08))]
=$34,767.75
But
what if he did this:
·$200
every month for 120 months (at the end) at 8% interest:
So, let P= 200 and
r = .08
Our calculation
is of the form P[(1+r/12)^119 + (1+r/12)^118 + … + (1+r/12) +1]
So, 200[(1+.08/12)^119
+ (1+.08/12)^118 + … + (1+.08/12) +1]
= 200[ (1-(1+.08/12)^120)
=$36,589.21
Now
suppose Adam continuously put money into his account:
·$2400
continuously for 10 years at 8% interest:
Here we will need
the ODE dS/dt –rS = k
So, let r =.08 and
k= 2400
Our differential
equation, dS/dt =rS + k, has the solution:
S(t)=
(k/r)(e^rt-1)
To derive this solution,
here, r=p(t)
integrating factor
e^ -(antiderivative of p(t))dt
=
e^(-rt)
So, e^(-rt) dS/dt
– re^(-rt)S = ke^(-rt)
Then, (Se^(-rt))’
= ke^(-rt)
So, Se^(-rt)=ke^(-rt)
dt
Thus, S= e^rt{(k/-r)(e^-rt)
+C}
S(0)=0iff
C=k/r
S=-k/r + (k/r)e^(rt)
That is, S= (k/r)(e^(rt)-1)
So, getting back
to our continuous model, we have:
S(t) = (k/r)(e^(rt)
– 1)
So, S(10)=(2400/.08){e^(10*.08)-1}
=$36,766.22
As we can see, the
continuous model is greater than the monthly model, which is greater than
the annual model.
Now, applying
this model to Adam and Bob, we have:
Adam (at age
21):
Contributes $2400
continuously/year for 10 years
Amount in account
at age 60 at 8% interest:
S(t)= (k/r)*{e^(rt)-1}
S(10)=(2400/.08){e^(10*.08)-1}
=$36,766.22
So, amount at age
60 is,
S(10)*{e^(.08*30)}
=$405,280.61
Bob (at age 31):
Contributes $2400
continuously/year for 30 years
Amount in account
at age 60 at 8% interest:
S(t)= (k/r)*{e^(rt)-1}
S(30)= (2400/.08){e^(30*.08)-1}
=$300,695.29
Comparing this
to their simple interest model, we see that both have a better account
with the continuous interest and contributions, and Adam still has more
than Bob. If we compare the continuous interest and contributions model
to the one below which just has continuous interest, we see that, although
continuous interest is better than simple interest, continuous contributions
with continuous interest is even better.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Cyclic Interest Rates: Annual and Monthly Contributions vs Modified ODE Continuous Model
Now, these
models are unrealistic since the interest rate won't stay constant. So,
let's vary the interest rate. I chose the cycle below by picking rates
between 2 and 8. So, let's go back and compare Adam's three models using
the cycle below, where the interest rate moves through the cycle once per
year. Below that, we will see that monthly contributions is still better
than yearly. Now, for the continuous interest and contributions model,
we must modify the ODE since r is no longer constant, but a function of
time. So, using sine regression I found the sine function which best fit
the interest rate cycle. This is shown below the yearly and monthly comparison
chart. Here we let the
interest rate be cyclic as follows: |
|||
|
|
{8,
4, 2.1, 3.4, 4.9, 6.1, 7.4, 8,…}
|
||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
dS/dt=r(t)S+k
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
By using a program, Runga Kutta,
I found Adam's balance at the end of the 10th year to be $31, 987.84. Once
again, this shows that continuous interest and contributions is better
than yearly or monthly contributions.