What Is A Nilpotent Matrix?
A nilpotent matrix is a square matrix $A$ for which some positive integer power equals the zero matrix, written $A^k = 0$. The smallest power $k$ that makes this true is called the index (or degree) of nilpotency.
The simplest non-trivial case is a $2 \times 2$ matrix with a single entry above the diagonal:
$$A = \begin{bmatrix} 0 & 1 \ 0 & 0 \end{bmatrix}, \qquad A^{2} = \begin{bmatrix} 0 & 1 \ 0 & 0 \end{bmatrix}\begin{bmatrix} 0 & 1 \ 0 & 0 \end{bmatrix} = \begin{bmatrix} 0 & 0 \ 0 & 0 \end{bmatrix}$$
Here $A$ itself is not the zero matrix, but $A^2$ is, so $A$ is nilpotent with index $2$. Note the two requirements baked into the definition: the matrix must be a square matrix, and the power that returns zero must be a whole number. The zero matrix counts too, as the trivial case with index $1$.
What Are Some Examples Of A Nilpotent Matrix?
Nilpotent matrices are not always as obvious as a single stray entry above the diagonal. This $2 \times 2$ matrix looks ordinary, yet it is nilpotent:
$$B = \begin{bmatrix} 2 & -4 \ 1 & -2 \end{bmatrix}$$
Square it, entry by entry using row-by-column multiplication:
$$B^{2} = \begin{bmatrix} (2)(2)+(-4)(1) & (2)(-4)+(-4)(-2) \ (1)(2)+(-2)(1) & (1)(-4)+(-2)(-2) \end{bmatrix} = \begin{bmatrix} 0 & 0 \ 0 & 0 \end{bmatrix}$$
So $B^2 = 0$ and $B$ has index $2$. Its determinant is $(2)(-2)-(-4)(1) = 0$ and its trace is $2 + (-2) = 0$, the two fingerprints every nilpotent matrix carries. The lesson: you cannot judge nilpotency by appearance, you have to test the powers.
How Do You Find The Index Of A Nilpotent Matrix?
To find the index, raise the matrix to successive powers until you first hit the zero matrix, and record that exponent. Take this $3 \times 3$ matrix with entries only above the diagonal:
$$C = \begin{bmatrix} 0 & 1 & 2 \ 0 & 0 & 3 \ 0 & 0 & 0 \end{bmatrix}$$
Square it first:
$$C^{2} = \begin{bmatrix} 0 & 0 & 3 \ 0 & 0 & 0 \ 0 & 0 & 0 \end{bmatrix}$$
$C^2$ is not zero, so the index is not $2$. Multiply once more:
$$C^{3} = C^{2},C = \begin{bmatrix} 0 & 0 & 0 \ 0 & 0 & 0 \ 0 & 0 & 0 \end{bmatrix}$$
Now $C^3 = 0$, so the index of $C$ is $3$. There is a ceiling on how large the index can get: for any $n \times n$ nilpotent matrix (a matrix of order $n$), the index is at most $n$. That bound comes from the Cayley-Hamilton theorem, which forces $A^{n} = 0$ once every eigenvalue is zero.
How Do You Check Whether A Matrix Is Nilpotent?
The decisive test is short: for an $n \times n$ matrix, compute $A^{n}$, and if it equals the zero matrix the matrix is nilpotent. This works because the index can never exceed $n$, so if raising the matrix to the $n$-th power has not reached zero, no higher power will either.
Before grinding through the powers, three quick screens can rule a matrix out, since each is a necessary condition:
Trace check: the trace (sum of the diagonal) must be $0$.
Determinant check: the determinant must be $0$, so the matrix must be singular.
Eigenvalue check: every eigenvalue must be $0$.
If any screen fails, stop, the matrix is not nilpotent. If all three pass, confirm with the power test, because passing the screens is necessary but not sufficient. Consider $\begin{bmatrix} 1 & 0 & 0 \ 0 & -1 & 0 \ 0 & 0 & 0 \end{bmatrix}$: its trace and determinant are both $0$, yet its eigenvalues are $1$, $-1$, and $0$, not all zero, so it is not nilpotent. Computing each power means repeated multiplication of matrices, which is why the necessary screens are worth running first.
Why Is Every Triangular Matrix With A Zero Diagonal Nilpotent?
Any triangular matrix whose main diagonal is entirely zero is nilpotent. A strictly upper triangular matrix (zeros on and below the diagonal) pushes its nonzero band one step further toward the corner every time you multiply it by itself, exactly as the $3 \times 3$ example above did, until the band falls off the edge and the matrix becomes zero.
The eigenvalues of any triangular matrix are just its diagonal entries. Set that diagonal to all zeros and every eigenvalue is zero, which is the spectral signature of nilpotency. A diagonal matrix with a zero diagonal is the extreme case, since it is already the zero matrix, with index $1$.
What Are The Properties Of A Nilpotent Matrix?
Every property below follows from one root fact: all eigenvalues of a nilpotent matrix equal zero.
Table: The defining properties of a nilpotent matrix and why each holds.
Property | What it says | Why |
|---|---|---|
Must be square | Only a square matrix can be nilpotent | $A^k$ needs $A$ times itself, which requires equal rows and columns |
Trace is zero | The diagonal entries sum to $0$ | Trace equals the sum of eigenvalues, all of which are $0$ |
Determinant is zero | $\det(A) = 0$ | Determinant equals the product of eigenvalues, all of which are $0$ |
Not invertible | No inverse exists; it is singular | A zero determinant rules out an invertible matrix |
Eigenvalues all zero | The only eigenvalue is $0$ | $Av = \lambda v$ with $A^k = 0$ forces $\lambda = 0$ |
Index at most $n$ | For order $n$, the index $k \le n$ | Cayley-Hamilton gives $A^n = 0$ |
Rank below $n$ | The rank is less than $n$ | A singular matrix has dependent rows |
Because its determinant is always zero, a nilpotent matrix is never a non-singular matrix. That is the property students lean on most in exams, and the one the next section protects against misreading.
How Is A Nilpotent Matrix Different From Idempotent And Singular Matrices?
These three labels describe different behaviours, and they overlap in only one place. A nilpotent matrix satisfies $A^k = 0$; an idempotent matrix satisfies $A^2 = A$ (it reproduces itself); a singular matrix simply has determinant zero.
Table: Nilpotent, idempotent, and singular matrices compared.
Feature | Nilpotent | Idempotent | Singular |
|---|---|---|---|
Defining rule | $A^{k} = 0$ | $A^{2} = A$ | $\det(A) = 0$ |
Eigenvalues | all $0$ | only $0$ or $1$ | includes at least one $0$ |
Determinant | always $0$ | $0$ or $1$ | always $0$ |
Invertible? | never | only the identity matrix | never |
Zero matrix qualifies? | yes, index $1$ | yes, $0^2 = 0$ | yes, $\det = 0$ |
Every nilpotent matrix is singular, but the reverse fails: most singular matrices are not nilpotent. The only matrix that is both nilpotent and idempotent is the zero matrix, where $A^2 = A = 0$.
Why Do Nilpotent Matrices Matter?
Nilpotent matrices are not a curiosity invented to trap students. They are the algebraic model of a process that moves in one direction only and cannot cycle back, so repeating it eventually exhausts everything to nothing.
The pure "shift" building block. In the theory of the Jordan form, every matrix splits into a scaling part (the eigenvalues) and a shifting part that carries no scaling of its own. That shifting part is exactly a nilpotent matrix, which is why understanding nilpotency unlocks the structure of every linear map.
A finite, terminating chain. Differentiate a polynomial of degree $n$ enough times and you reach zero. That "runs out after finitely many steps" behaviour is nilpotency, and it appears anywhere a quantity strictly decreases at each step.
No hidden loops. A transformation that is nilpotent can have no cycle, because a cycle would repeat forever instead of dying out. That links nilpotent matrices to processes guaranteed to finish.
Strip away scaling and rotation, and what remains of a linear map is a nilpotent core. Studying it is how mathematicians describe the part of a transformation that only pushes forward and never returns.
Who Discovered Nilpotent Matrices?
The behaviour came first, and the name came later, from an American mathematician classifying every algebra he could imagine.
Two more mathematicians shaped how we understand these matrices today:
Arthur Cayley (1821–1897, England) proved the Cayley-Hamilton theorem, which explains why an $n \times n$ nilpotent matrix always satisfies $A^n = 0$.
Camille Jordan (1838–1922, France) developed the Jordan canonical form, in which every nilpotent matrix appears as a tidy block with $1$s just above the diagonal and $0$s everywhere else.
Where Are Nilpotent Matrices Used In The Real World?
The "runs out after a fixed number of steps" behaviour turns up in more fields than its abstract definition suggests.
Calculus and differential operators: the operation of differentiating polynomials of degree at most $n$ is nilpotent, because enough repeated derivatives send any such polynomial to zero.
Computer science and scheduling: the adjacency matrix of a directed acyclic graph is nilpotent, which is the algebra behind task schedulers and build systems that must contain no circular dependency.
Control engineering: a discrete system whose state matrix is nilpotent settles to exactly zero in a finite number of steps, the idea behind deadbeat controllers that reach their target and stop.
Quantum physics: raising and lowering operators for finite spin systems are nilpotent, since applying the raising operator past the top state annihilates it.
One quiet idea, a process that cannot loop and must terminate, connects calculus, scheduling, control systems, and quantum mechanics. That reach across fields is what makes the concept worth naming.
What Are The Most Common Nilpotent Matrix Mistakes?
These three errors account for most lost marks on nilpotency questions, and each traces back to reading one property as if it were the whole definition.
Assuming any singular matrix is nilpotent.
Where it slips in:
A student sees that a matrix has determinant zero, remembers that nilpotent matrices are singular, and concludes the matrix must be nilpotent.
Don't do this:
Do not run the implication backwards. Singular is necessary for nilpotency, not sufficient.
The correct way:
Check that every eigenvalue is zero, or compute $A^n$ directly. The matrix $\begin{bmatrix} 1 & 0 \ 0 & 0 \end{bmatrix}$ is singular, yet $A^2 = A \neq 0$, so it is not nilpotent.
Confusing a nilpotent matrix with an idempotent matrix.
Where it slips in:
A student blends the two rules and writes $A^2 = A$ for a nilpotent matrix, or expects a nilpotent matrix to reproduce itself.
Don't do this:
Do not mix the defining equations. Nilpotent means $A^k = 0$; idempotent means $A^2 = A$.
The correct way:
Keep the rules separate. The only matrix satisfying both at once is the zero matrix, where $A^2 = A = 0$.
Stopping the power test too early.
Where it slips in:
Working with a $3 \times 3$ matrix, a student computes $A^2$, finds it nonzero, and declares the matrix not nilpotent.
Don't do this:
Do not stop before the $n$-th power. A $3 \times 3$ nilpotent matrix can have index $3$, so $A^2$ may be nonzero while $A^3 = 0$.
The correct way:
For an $n \times n$ matrix, carry the powers up to $A^n$ before deciding. Only when $A^n$ is still nonzero can you conclude the matrix is not nilpotent.
Practice Problems On Nilpotent Matrix
Test each matrix using the power method, and state the index where it applies. Answers follow each problem.
Is $\begin{bmatrix} 0 & 5 \ 0 & 0 \end{bmatrix}$ nilpotent?
(Answer: Yes. Its square is the zero matrix, so the index is $2$.)Show that $\begin{bmatrix} 1 & 1 \ -1 & -1 \end{bmatrix}$ is nilpotent and give its index.
(Answer: The square is $\begin{bmatrix} 0 & 0 \ 0 & 0 \end{bmatrix}$, so it is nilpotent with index $2$; trace and determinant are both $0$.)Find the index of $\begin{bmatrix} 0 & 2 & 1 \ 0 & 0 & 3 \ 0 & 0 & 0 \end{bmatrix}$.
(Answer: The square is nonzero but the cube is the zero matrix, so the index is $3$.)Is $\begin{bmatrix} 1 & 0 \ 0 & 0 \end{bmatrix}$ nilpotent? Explain.
(Answer: No. It is singular but equals its own square, so it is idempotent; its eigenvalues are $1$ and $0$, not all zero.)A $4 \times 4$ nilpotent matrix has index $k$. What is the largest possible value of $k$?
(Answer: $4$, because the index of an $n \times n$ nilpotent matrix is at most $n$.)
Where Should You Go Next After Nilpotent Matrix?
Nilpotency sits at a crossroads of ideas about eigenvalues, structure, and matrix families, and several natural doors open from here.
Eigenvalues. The fact that every eigenvalue is zero is what makes a matrix nilpotent, so eigenvalues are the deeper concept sitting underneath this whole topic.
Triangular matrix. The strictly triangular matrices are the readymade source of nilpotent examples, and they show how the diagonal controls the eigenvalues.
Matrices and determinants. Strengthen the determinant and singularity ideas that decide when a matrix can be nilpotent in the first place.
For a live walkthrough of these matrix families, a Bhanzu trainer teaches nilpotency starting from what the powers actually do, in the Bhanzu algebra program.
Was this article helpful?
Your feedback helps us write better content