Optiver OA on 1point3acres 2026: High-Frequency Questions + VO Coaching Guide

Every September to December, Optiver threads on 1point3acres (the largest Chinese-speaking tech-jobs forum) surge. Unlike traditional tech companies, Optiver is an Amsterdam-based global market maker, and its OA centers on trading scenarios: order book matching, market-making quotes, trading-sequence profit maximization, plus mental math and probability for Traders. This article aggregates the 2026 1point3acres reports by question type and adds a realistic VO coaching / mock-interview plan.

Optiver OA at a Glance (2026)

Dimension	SWE Track	Trader Track
Platform	HackerRank	Optiver in-house + HackerRank
Duration	90 minutes	60-80 minutes
Format	2-3 coding problems	Mental math + probability + 1 coding
Focus	Simulation, greedy, heaps	Expectation, Sharpe, market-making intuition
Difficulty	LC Medium-Hard	Very fast arithmetic + strong probability

Recurring feedback on 1point3acres: 90 minutes for 3 problems is tight. Solving 2 with full AC is already strong enough to clear the OA bar.

Question Type 1: Trading Sequences (Max Weighted Subarray)

Problem

Given price series prices[] and matching volumes[], pick a contiguous subarray [l, r] (length ≥ 2) maximizing the weighted profit.

Idea

Compute step profit: profit[i] = (prices[i+1] - prices[i]) * volumes[i]
Apply a Kadane variant
Return 0 if every step is negative (no trade)

Python

def max_trading_window(prices, volumes):
    n = len(prices)
    if n < 2:
        return 0

    profits = [(prices[i + 1] - prices[i]) * volumes[i] for i in range(n - 1)]

    best = cur = 0
    for p in profits:
        cur = max(p, cur + p)
        best = max(best, cur)
    return best

Time O(n)　Space O(n)

Question Type 2: Order Book Matching

Problem

Implement ADD / CANCEL / MATCH:

ADD side price quantity — limit order
CANCEL order_id — withdraw
MATCH — when best_bid ≥ best_ask, trade at the resting (counter) price

Idea

Bids: max-heap by price
Asks: min-heap by price
Cancels via lazy deletion — skip removed ids when peeking

Python

import heapq

class OrderBook:
    def __init__(self):
        self.bids = []        # (-price, id, qty)
        self.asks = []        # (price, id, qty)
        self.cancelled = set()
        self.next_id = 0

    def add(self, side, price, qty):
        self.next_id += 1
        oid = self.next_id
        if side == "BUY":
            heapq.heappush(self.bids, (-price, oid, qty))
        else:
            heapq.heappush(self.asks, (price, oid, qty))
        return oid

    def cancel(self, oid):
        self.cancelled.add(oid)

    def _clean(self, heap):
        while heap and heap[0][1] in self.cancelled:
            heapq.heappop(heap)

    def match(self):
        trades = []
        while True:
            self._clean(self.bids)
            self._clean(self.asks)
            if not self.bids or not self.asks:
                break
            best_bid = -self.bids[0][0]
            best_ask = self.asks[0][0]
            if best_bid < best_ask:
                break
            bp, bid, bq = heapq.heappop(self.bids)
            ap, aid, aq = heapq.heappop(self.asks)
            qty = min(bq, aq)
            trades.append((best_ask, qty))
            if bq > qty:
                heapq.heappush(self.bids, (bp, bid, bq - qty))
            if aq > qty:
                heapq.heappush(self.asks, (ap, aid, aq - qty))
        return trades

Time ADD/CANCEL O(log n), MATCH O(k log n)

Question Type 3: Allocation under Capital / Risk Constraints

Problem

n traders (capital, risk tolerance) and m opportunities (required capital, risk level, expected profit). Assign each opportunity to at most one qualifying trader, maximizing total profit.

Idea

Sort opportunities by expected profit descending
For each, pick the trader who qualifies and has the least slack (capital − required) — saves high capital for big tickets later
Classic greedy

Python

def allocate(traders, opps):
    # traders: [(cap, risk_tol)]
    # opps:    [(need_cap, risk_lv, profit)]
    opps.sort(key=lambda x: -x[2])
    used = [False] * len(traders)
    total = 0

    for need, risk, profit in opps:
        pick, slack = -1, float("inf")
        for i, (cap, tol) in enumerate(traders):
            if used[i] or cap < need or tol < risk:
                continue
            if cap - need < slack:
                slack = cap - need
                pick = i
        if pick >= 0:
            used[pick] = True
            total += profit
    return total

1point3acres Frequency Table

Category	Frequency	Core prep
Order book matching	★★★★★	Heaps + lazy delete
Max weighted subarray	★★★★	Kadane variant
Capital/risk allocation	★★★★	Greedy + sort
Simplified market-making	★★★	bid/ask spread
Mental math (Trader only)	★★★★★	4-digit math in 8s

VO Coaching / Mock Interview Roadmap

Optiver onsites on 1point3acres are described as long and high-pressure: technical + Trading Game + mental math + behavioral + hiring manager. The real value of VO coaching is not losing what you already know when stress kicks in.

1) VO loop breakdown

Round 1: Algorithms / system design (45 min)
Round 2: Order book / matching engine discussion
Round 3: Behavioral (collaboration, pressure)
Round 4: Trading game / probability (Trader only)
Round 5: Hiring manager / team fit

2) How VO coaching is typically used

First pass: bucket the last 30 days of 1point3acres reports by topic; list 5 variants each for Order Book and Trading Sequences
Shadow interviews: Zoom + screen share; mentor plays interviewer with a 45-minute timer
Replay debrief: record the session; focus on "started coding without clarifying" and "missed edge case" patterns

3) oavoservice's combined VO Coaching + VO Proxy package

oavoservice offers both VO Proxy (live interview support) and VO Coaching (mocks + question bucketing + recorded debriefs) for the full Optiver SWE / Trader loop:

Pre-interview: VO Coaching mocks at realistic 90-minute pacing; minute-by-minute debriefs
During the interview: VO Proxy live support so stress doesn't wipe out what you already know
Between rounds: targeted follow-up patches before the next round

Reach out on WeChat Coding0201 for the full plan and pricing.

7-Day Sprint

Day	Task
D1	Read last 90 days of Optiver 1point3acres reports, bucket by type
D2	Implement Order Book (with lazy delete) once; 5 test cases
D3	Trading Sequences + Allocation: 3 variants each
D4	Probability/expectation: 20 problems (Trader only)
D5	One full 90-minute HackerRank-style mock
D6	System design: matching engine + low-latency queues
D7	Behavioral: STAR debrief, 2 team-conflict stories

FAQ

Is Optiver OA basically LeetCode Hard?

The surface reads Medium, but 90 minutes × 3 problems + long English prompts + many edge cases puts the overall pressure at LC Hard. On 1point3acres, candidates who solve 2 with full AC have the highest passthrough.

Can I just memorize 1point3acres reports?

No. Optiver rotates ≥ 30% of problems yearly, but themes (matching, greedy allocation, profit subarray) are stable. Memorize patterns and templates, not statements.

SWE or Trader track?

If you can do two-digit multiplication in 8 seconds and have probability intuition, try Trader. Otherwise SWE — better suited to ACM-style backgrounds with strong Python.

Failed the OA — can I reapply?

Usually a 12-month cooldown. A few 1point3acres reports describe successful re-applications after 9 months with a referral and a stronger resume.

Preparing for Optiver via 1point3acres reports?

oavoservice tracks Optiver SWE / Trader OA and VO updates and offers question bucketing, variant walkthroughs, timed mocks and behavioral debriefs. Our mentors come from Optiver, Citadel, Jane Street and Five Rings, and can build a 1-2 week sprint around your weak spots.

👉 Add WeChat: Coding0201 — get Optiver high-frequency questions + VO coaching.

Contact

Email: [email protected]
Telegram: @OAVOProxy