Just recently had it done. J.P. Morgan 2026 New Grad OAoverall experience is very smooth. The questions are not tricky, but the logical rigor and code specification requirements are quite high. The entire test has two questions, both classic question types, as long as well-prepared, clear thinking, it is not difficult to pass.
Test Process Overview
- Platform: HackerRank
- quantity of questions: 2 questions
- duration: 90 minutes
- language restriction: Unlimited (Python, Java, C++ are all acceptable)
- additional link: No personality assessment, no behavioral questions
- direction of an exam point: sliding windows, greedy algorithms, heaps (priority queues), basic data structures
The whole test experience is very smooth, the topic is biased towards basic algorithmic logic, the amount of code is not large, but pay attention to boundary processing, time complexity, and output format consistency.
Question 1: Stock Price Intervals
Question.
Given an array of stock prices and an integer k, determine how many contiguous subarrays of length k have strictly increasing prices.
Thought Analysis:
This question is actually a typical sliding window problem. The question asks to find all the objects of length k Whether the stock prices are "strictly increasing" in consecutive subarrays of the
The thought process is:
- Iterate through all the possible lengths of
kThe window. - Checks whether neighboring elements in each window satisfy
prices[i] < prices[i+1]. - If the entire subarray satisfies strict incrementation, add one to the counter.
Pay attention to the boundaries in detail: when k is greater than the length of the array, and in the case of consecutive equality (e.g. [3,3,4]) is not considered incremental. The overall logic is straightforward, but write it in such a way as to control the loop boundaries.
Question 2: Minimum Cost to Combine Elements
Question.
You are given an array of integers. Each time, you can remove two elements from the array, add their sum back to the array, and pay the cost equal to their sum.
Return the minimum total cost to reduce the array to one element.
Thought Analysis:
This is a classic greedy problem, along the lines of the "optimal merger problem" or "Huffman coding". To minimize the total cost, the two smallest current numbers should be merged each time.
Implementations can use a min-heap to maintain the minimum value in the current array:
- Put all the numbers in the smallest pile.
- Remove the two smallest values at the top of the heap one at a time, compute their sum and add it back to the heap.
- Accumulate the cost of all merges until there is only one element left in the heap.
It can be done with sorting + linear processing, but the heap writeup is more efficient (complexity O(n log n)).
Lessons Learned & Preparation Advice
The biggest impression I got from this OA is that Morgan puts a lot of emphasis on "engineering sense" and "logical stability".
The topic is not a flowery endeavor, but requires concise, unambiguous writing with clear boundaries.
Suggested directions for preparation are as follows:
bonus points
If the language supports it (e.g. Python or Java), properly demonstrating standard library usage such as STL, heap, deque, etc. will make the code more professional.
Algorithm review focus
Sliding window (e.g. successive incremental subarrays, maxima and subintervals)
Greedy algorithms (optimal merging, interval coverage, task scheduling, etc.)
Basic data structures: heaps, queues, hash tables
Code Conventions Requirements
Note the variable naming and indentation style;
The output format is strictly on-topic (HackerRank will strictly determine this);
Add a few comments to facilitate logical self-testing.
Time management recommendations
An average of 40-45 minutes per question is sufficient;
When you encounter a choke point, write the base correct solution first, then consider optimization.
From OA to Offer, a one-stop, no-holds-barred helper for steady entry into big firms like JPMorgan!
If you are also preparing for JPMorgan OA or other large firms' written tests, you can learn about us! Programhelp The one-stop remote assistance program.
Our team consists of North American CS engineers and former Amazon interviewers, who have been accompanying students to complete all kinds of HackerRank / Codesignal / Codility / company's self-developed platform OA.
Support Real-time voice reminders, no trace remote assistance, guide ideas while writing, to help you get points accurately within the time limit.
In addition to OA, we also offer:
- VO real-time assists: Real-time voice prompts + thought leadership to help you answer questions in a logical and natural way in interviews;
- Mock Interview Training: Built on 500+ real questions from companies like Amazon, Meta, J.P. Morgan, and more;
- Personalized strategy guide: Customized preparation programs for each company's question trends.
Whether you're looking to take down a fintech (J.P. Morgan, Goldman Sachs) or a quantitative powerhouse (Citadel, IMC, Optiver), the
Programhelp can provide you with safe, stable and hidden technical support, so that you can pass every hurdle smoothly and reach your dream offer easily.
More J.P. Morgan Interviews to Share
J.P.Morgan Interview - J.P.Morgan Interview - J.P.Morgan Generation Interview - VO Aid
JPMorgan pre-interview HireVue Interview Experience Sharing
JP Morgan VO Interview Guide | Coding, System Design, Behavioral Full Analysis
