Infer BA When AB Is the Identity

Quick Overview

Analyze what AB = I implies about BA for both same-size square matrices and compatible rectangular matrices, including the projection case.

Infer BA When AB Is the Identity

Company: Goldman Sachs

Role: Data Scientist

Category: Statistics & Math

Difficulty: easy

Interview Round: Online Assessment

# Infer BA When AB Is the Identity Matrices `A` and `B` satisfy `AB = I`. Explain what can be concluded about `BA`. The dimensions are not specified, so analyze both the case where `A` and `B` are square matrices of the same size and the compatible rectangular case. ### Constraints & Assumptions - Matrices are over the real numbers. - `I` is the identity matrix of the row dimension of `A`. ### Clarifying Questions to Ask - Are both matrices square, and if not, what are their dimensions? - Is the question asking whether `BA` is an identity, a projection, or merely constrained in rank? ```hint Interpret one-sided inverses For rectangular matrices, a right inverse need not also be a left inverse; examine what multiplying `BA` by itself does. ``` ### What a Strong Answer Covers - Invertibility and the determinant or rank argument in the square case. - The distinction between right and left inverses for rectangular matrices. - The idempotence and rank of `BA` when `AB = I` is rectangular. - A counterexample to the claim that `BA` must always be identity. ### Follow-up Questions - What are the eigenvalues of `BA` in the rectangular case? - Under what dimension condition can `AB = I` hold?

Quick Answer: Analyze what AB = I implies about BA for both same-size square matrices and compatible rectangular matrices, including the projection case.

|Home/Statistics & Math/Goldman Sachs
Goldman Sachs logo
Goldman Sachs
Sep 2, 2026
easyData ScientistOnline AssessmentStatistics & Math
1
0

Infer BA When AB Is the Identity

Matrices A and B satisfy AB = I. Explain what can be concluded about BA. The dimensions are not specified, so analyze both the case where A and B are square matrices of the same size and the compatible rectangular case.

Constraints & Assumptions

  • Matrices are over the real numbers.
  • I is the identity matrix of the row dimension of A .

Clarifying Questions to Ask Guidance

  • Are both matrices square, and if not, what are their dimensions?
  • Is the question asking whether BA is an identity, a projection, or merely constrained in rank?

What a Strong Answer Covers Guidance

  • Invertibility and the determinant or rank argument in the square case.
  • The distinction between right and left inverses for rectangular matrices.
  • The idempotence and rank of BA when AB = I is rectangular.
  • A counterexample to the claim that BA must always be identity.

Follow-up Questions Guidance

  • What are the eigenvalues of BA in the rectangular case?
  • Under what dimension condition can AB = I hold?
Loading comments...