Second-order Difference Equations

Basic Concepts

Definition 1: A second-order difference equation takes the form

xn = f(n, xn-1, xn-2)

for all positive integers n. A solution is a function g(x, y) such that

xn = g(x0, x1)

By induction on n, such a solution is unique.

Definition 2: A linear second-order difference equation with constant coefficients takes the form

xn = f(n, xn-1, xn-2) = uxn-1 + vxn-2 + w

Alternatively, this can be expressed as

xn+2 + bxn+1 + cxn = a

The difference equation is homogeneous if a = 0.

We start by looking at solutions in the homogeneous case.

Definition 3: The characteristic equation of a second-order difference equation is

y2 + by + c = 0

Solution for homogeneous equations

Property 1: A homogeneous linear second-order difference equation with constant coefficients has a solution based on its characteristic equation, as follows:

If b2 > 4c, then the characteristic equation has two distinct real roots r1 and r2. Any solution of the difference equation takes the form Ar1n + Br2n.

Using the binomial formula, these roots are

Binomial formula

If b2 = 4c, then the characteristic equation has one distinct real root r1 = –b/2. Any solution of the difference equation takes the form (A + Bn)r1n.

If b2 < 4c, then the characteristic equation has two conjugate complex roots u ± iv where

real and imaginary components

Any solution to the homogeneous differential equation takes the form

Complex number solution

Note that

complex roots

where

amplitude and theta

Alternatively

cos and sin theta

Thus

Complex solution 1

Complex solutions 2

Also note

Formula for r

Setting A = C+D and B = (C – D)i, we see that any solution to the homogeneous differential equation takes the form

xn = rn(A cos + B sin )

Using initial values

Property 2: If x0 and x1 are known, we can solve for A and B as follows:

If b2 > 4c, then any solution of the difference equation takes the form Ar1n + Br2n.

x0 = A + B          x1 = Ar1 + Br2

Here, x0, x1, r1, and r2 are known. We solve for A and B.

If b2 = 4c, then any solution of the difference equation takes the form (A + Bn)r1n.

x0 = A          x1 = (A + Bn)r1

Since x0, x1, and r1 are known, we can solve for A and B.

If b2 < 4c, then any solution of the difference equation takes the form xn = rn(A cos + B sin ).

x0 = r0(A cos 0 + B sin 0) = A          x1 = r(A cos θ + B sin θ)

Since x0, x1, r, and θ are known, we can solve for A and B.

Solution for non-homogeneous equations

Property 3: If a is not zero, then the solution to the difference equation needs to be modified by adding d to the corresponding  homogeneous difference equation solution, where

d = a / (b + c + 1)

If, however, b + c + 1 = 0, then we set

d = an / (b + 2)

provided that b ≠ -2. Otherwise, we set

d = an2 / 2

Let d0 be the value of d when n = 0, and d1  be the value of d when n = 1. When there are two distinct real roots, clearly d0 = d1 = d. For the other two cases, d0 = 0.

We also need to modify the approach for finding the A and B coefficients as follows:

If b2 > 4c, then any solution of the difference equation takes the form Ar1n + Br2n + d.

x0 = A + B + d0          x1 = Ar1 + Br2 + d1

If b2 = 4c, then any solution of the difference equation takes the form (A + Bn)r1n + d.

x0 = A + d0          x1 = (A + Bn)r1 + d1 

If b2 < 4c, then any solution of the difference equation takes the form xn = rn(A cos + B sin ) + d.

x0 = A+ d0          x1 = r(A cos θ + B sin θ) + d1

Convergence

Definition 4: A steady-state is achieved if the sequence xn converges, i.e. the limit of xn exists as n → ∞.

Property 4: The sequence converges provided

If two distinct real roots: |r1| < 1, |r2| < 1

If one real root: |r1| < 1

If two complex roots: r = √c < 1

Alternatively, the sequence converges if all the following conditions hold:                                        

1 + b + c > 0          1 – b + c > 0          c < 1

Examples and worksheet functions

Click here for examples and Real Statistics worksheet functions.

Links

↑ First-order difference equations

References

Osborne, M. J. (2025) Second-order difference equations
https://mjo.osborne.economics.utoronto.ca/index.php/tutorial/index/1/sod/t

Dowling, E. T. (1980) Introduction to mathematical economics. 3rd ed. Schaum’s Outline
https://ugess3.wordpress.com/wp-content/uploads/2015/08/schaum_introduction_to_mathematical_economics.pdf

Kumar, A. (2008) Linear, second-order difference equations
https://web.uvic.ca/~kumara/econ251/schap20.pdf

Leave a Comment