SQL in the Assessment

Most engineers prep the Morgan Stanley lateral assessment as an algorithm test and never open a SQL editor, which quietly leaves points on the table. The experienced-hire online assessment runs on HackerRank at roughly seventy to ninety minutes: two coding questions, a heavy multiple-choice block on Core Java and the systems around it, and a SQL query. That SQL query is graded like everything else, so a strong coder who skipped query practice can walk out having answered only part of the assessment well. This lesson closes that gap. The query shapes a bank reaches for are stable, and rehearsing them until they are automatic frees your attention for the harder half of the test.

Grade yourself the way the assessment does: a query that runs, returns the right rows, and reads cleanly on the first pass. The bar is Easy to Medium, never exotic. Nobody is asking you to tune a query planner or unwind a recursive CTE. You will be asked to join two tables, aggregate a column, filter groups with the right clause, and occasionally rank or accumulate with a window function. Every query below is written for SQLite, the engine HackerRank runs, and each one has been run against real rows so the answer is not a guess.

The Morgan Stanley angle

A SQL query is a first-class item on the Morgan Stanley lateral OA. The experienced-hire HackerRank blocks come phrased as "two coding questions and one SQL," or a wall of MCQs with a single SQL query attached. The domain wears the firm's vocabulary: trades, accounts, desks, and positions instead of the textbook orders-and-products. The SQL underneath is identical, so do not let the framing throw you. Read the shape hiding in the words. A phrase like "for each account" is a GROUP BY or a PARTITION BY. "Keep only the ones where the total exceeds" is a HAVING. "The position as the blotter runs down the day" is a window function ordered by trade date. Learn to hear the pattern and the domain stops mattering.