Math 165: Revenue Streams and Mortgages
Math 165: Revenue Streams and Mortgages
Best for printing: 165mortgage.pdf
I.
Income and Investment Streams
Simple Model
If a present value P is invested at time t=0 with continuous compounding (CC) at rate r, the future value at time t=T is
B = B(T) = P erT = P(0) erT.
If we wish to have future value B at time t=T, we should invest a present value P given by
P = P(0) = B e−rT = B(T) e−rT.
If we wish to withdraw amounts B1, B2, …, BN, at times T1, T2, …, TN, we must have present value
P = P1 + P2 + …+ PN = B1 e−rT1 + B2 e−r T2 + …+ BN e−rTN =
Bi e−rTi.
Continuous Model - Income Stream
We wish to withdraw a continuous income stream - to withdraw continuously at a rate R [dollars/ year] for T [years].
At a typical time t, over a period ∆t, we will withdraw ∆B ≈ R ∆t, a future value at time t. Thus we need a present value (investment) ∆P ≈ e−rt R∆t. The total present value needed is
P =
∆P ≈

t  from  0  to  T 
e−rt R∆t ≈
T

0 
R e−rt dt.
Figure 1.

strpic1.gif

Similarly, if we invest continuously at rate R [dollars/year] for T [years], the future value B at time T will be given by
B =
T

0 
R er(T−t) dt.




Figure 2.

strpic2.gif

Inflation Adjustments
Similar arguments can be made if the rate, R = R(t), depends on t. For example, the rate of contribution (investment) or income (revenue) might be continuously adjusted for inflation. In this case the formulas become:
If we wish to withdraw a continuous income stream - to withdraw continuously at a rate R(t) [dollars/ year] for T [years], we need a present value
P =
T

0 
R(t) e−r t dt.
In particular if we make a cost of living adjustment (COLA), of r1% annually,

R(t)
= R0 er1 t,
P
=
T

0 
R0 e(r1 − r)t dt.
If we invest continuously at a rate R(t) [dollars/year] for T [years], the future value B at time T will be given by
B =
T

0 
R(t) er(T−t) dt.
Pricing Annuities
For more on cost of living adjustments (COLA), see an exercise from Math 165 at UIC
II.
Mortgage Payments
A mortgage of P0 for T years with an annual interest rate r is traditionally paid off at a fixed [annual] rate R.
If P(t) is the principal remaining at time t, in a [small] time period ∆t, the payment is R  ∆t, the interest accrued is (or ≈ ) is r P(t) ∆t so that the change in principal is
∆P ≈ − R  ∆t+ r P ∆t.
Using the language of differentials,

dP = − R  dt + r P dt.
Using the magic of differentials, we divide by dt to obtain the differential equation
dP

dt
= −R + r P.
There are many ways to find all solutions. Let P(t) = u(t) ert. Then

P − r P
= u er t,
u er t
= −R,
u
= − R e−rt,
u(t)
= R

r
e−rt + C,
P(t)
= R

r
+ C ert.
In addition there are two boundary conditions satisfied by P(t):
P(0) = P0, P(T) = 0.
Using the initial condition, P(0) = P0,

C
= P0 R

r
,
P(t)
= R

r
+
P0 R

r

ert
If T is given, solve the equation P(T) = 0 for R:
R
= r P0 erT

erT −1
Since R is the annual rate of payment, the monthly payment is M = R/ 12.
If R is given, solve the equation P(T) = 0 for T:

0
= R

r
+
P0 R

r

erT
erT
= R

R − r P0
T
= 1

r
ln
R

R − rP0

N.B. In a practical application, R − rP0 > 0. The case R − rP0 = 0 corresponds to paying interest only.
Investigations
1.
Variable Rate Mortgages. At a time t = T1, the rate changes from r to r1. Investigate
If the monthly payment M is held constant, how does the term (expiration date) of the loan change?
If the term (expiration date) of the loan is unchanged, how does the monthly payment M change?
2.
Accelerated Payments. At time t = T1, make an extra payment R1 and continue making the monthly payment M. How is the expiration date of the loan changed?
3.
Rules of Thumb? Determine the veracity of the statements:
Doubling the monthly payment M halves the period T of the mortgage.
Doubling the period T of the mortgage halves the monthly payment M.
4.
Compare. Several mortgage calculators are available on the web.
Mortgage Calculator - Mortgage-calc.com
http://www.mortgage-calc.com/mortgage/simple.html
Take several common loans (for example $100,000, 30 years at 5.75%, ). Compare the monthly payment M you have calculated using continuous compounding (CC) and the monthly payment calculated by your favorite web mortgage calculator. Are the results different?
5.
Monthly Compounding. In practice, interest is not compounded continuously(CC). A discrete calculation using monthly compounding is used. If an annual rate r is compounded M equally spaced times in a year, the annual percentage rate (APR) is
APR=
1 + r

M

M

 
− 1.
The actual annual interest paid is the same as if the APR rate were simply compounded.
On most consumer loans, the APR is recorded.
Discrete Payments
Mortgages and other loans are usually paid on a monthly basis. We will assume that there is a nominal annual interest rate r, and that payments of $R are made M times per year so that in a time period of 1/M years, for a loan of $P , $(1 +[r/M])P is due.
Start with a loan of P0, and let Pn denote the principle remaining after the nth payment is made. The nth payment of R [dollars] covers the accrued interest ([r/M] Pn−1) and reduces the principal by R − [r/M] Pn−1.
There is a formula for Pn in terms of P0, r, and M. It is convenient to let
α
= 1 + r

M
.
Then

Pn
= αPn−1 −R
= α(αPn−2 −R)−R
= α2 Pn−2 − (α+ 1 )R
= α2 (αPn−3 −R) − (α+ 1 )R
= α3 Pn−3 − (α2 + α+ 1 )R
= …
= αn P0 − (αn−1 + …+ α+ 1 )R.
= αn P0 αn − 1

α− 1
R.
If the loan is paid off in T years with N = MT payments, we have the equation for R.

αN P0
= (αN−1 + …+α+ 1 )R
= αN − 1

α− 1
R,
R
= (α− 1) αN

αN − 1
P0.
Note that

α− 1
= r

M
,
αN
=
1+ r

M

MT

 
=
1+ r

M

[M/r] rT

 
=

1+ r

M

[M/r]

 

rT

 
.
Let

E
=
1+ r

M

[M/r]

 
.
Note that [r/M] is small and limϵ→ 0(1+ ϵ)[1/(ϵ)] = e, so that E ≈ e.
We obtain

R
= r

M
ErT

ErT − 1
P0,
E
=
1+ r

M

[M/r]

 
≈ e.
N.B. Since we have there are M payments in a year, the total yearly payment , Y, is
Y
= r ErT

ErT − 1
P0.
(†)
If Y and P0 are given, solve the equation (†) for T:

Y

rP0
= ErT

ErT− 1
,
(ErT− 1) Y

rP0
= ErT,
ErT
= Y

Y − rP0
,
T
= 1

ln(E)
  1

r
ln
Y

Y−rP0

.

Footnotes:

In our case, ≈ means that the error, which depends on t and ∆t, satisfies

lim
∆t→ 0 
error(t,∆t)

∆t
= 0.
divide by dt means: replace dt by ∆t ≠ 0, divide by ∆t, and let ∆t → 0.


File translated from TEX by TTH, version 4.05.
On 27 Aug 2014, 09:43.