Wolverine Trading Interview: Roles, Process and Preparation

Know which Wolverine business is interviewing you

Wolverine is a group of related but distinct financial businesses. Its trading-businesses page separates:

• Wolverine Trading, the proprietary market-making and valuation-arbitrage group established in 1994;
• Wolverine Execution Services, an independent broker-dealer serving institutional clients; and
• Wolverine Asset Management, an alternative asset manager investing across equity, credit, commodity and volatility strategies.

That distinction changes an interview. A proprietary trader may discuss pricing and inventory; an execution-services engineer may care about client orders and controls; an asset-management researcher may work at a different horizon entirely. “Wolverine interview questions” is too broad until you know the legal entity, desk and product.

The group is headquartered at 175 West Jackson Boulevard in Chicago. Current vacancies describe offices in Chicago, New York and London. Wolverine does not publish a firm-wide headcount, and its own site does not name individual founders; industry references such as MarketsWiki identify Robert Bellick and Christopher Gust as its co-managing partners.

What the public hiring pages reveal

Wolverine's vacancy platform can display a six-stage route: application review, technical assessment, first interview, second virtual interview, on-site final interview and offer. For example, that sequence appears on the Business Intelligence Engineer hiring-process page.

It is evidence for that vacancy and perhaps a common template, not proof that every trader, researcher and engineer receives the same loop. The official open-positions page is the right starting point. Candidate reports of timed arithmetic tests, market-making games and coding rounds can guide practice, but their timing, platform and scoring are not official unless they appear in your invitation.

There is no reliable public support for “80 questions in eight minutes”, a fixed four-to-six-interview superday, one-week decisions or a universal 12-month reapplication rule.

Match your evidence to the business

Start by translating the vacancy into three fields: business, decision and consequence. In proprietary trading, the decision might be a quote, hedge or allocation of risk, and the consequence belongs to Wolverine's capital. In execution services, the system may handle a client's order, making routing quality, instructions and auditability central. In asset management, the research horizon, mandate and portfolio constraints can differ again.

This prevents superficially impressive but misplaced preparation. A low latency benchmark may be relevant to an automated trading platform but insufficient for a client workflow where correctness and recoverability dominate. An options surface project may fit a valuation role but should not be forced into a business intelligence interview. Follow the actual job description and ask which users, products and production decisions the role supports.

Choose evidence with inspectable detail. Traders should bring one decision in which they updated a price or risk view as information arrived. Researchers should be able to reconstruct data chronology, validation and trading assumptions. Engineers should show a system where they measured the constraint, protected correctness and handled failure. In each case, state your contribution precisely. Wolverine's public language about ownership is better served by an honest account of one consequential decision than by taking credit for an entire team's result.

Trader preparation: derivatives with consequences

Wolverine officially describes its proprietary business as market making and valuation arbitrage across global markets. Current trading vacancies include quantitative trading analysis, high-frequency futures and systematic options. The common thread is not a particular puzzle; it is turning a valuation into a controlled position.

Price the risk, not just the option

Know the mechanics of Black-Scholes well enough to challenge its assumptions. Explain how spot, volatility, time, rates and dividends alter value and Greeks. Then move beyond a lone option: skew, term structure, discrete dividends, borrow, early exercise, liquidity and hedging costs all matter to a quoted market.

The options market-making guide, Black-Scholes explainer and volatility smile guide cover that progression.

Think in inventory

After a trade, restate the position and its risks before moving the quote. Candidates often rush to say “widen the market” without identifying delta, gamma, vega, concentration or the information contained in the customer's action. There may be several defensible hedges; compare their cost, basis risk and liquidity.

Explain both sides of a market

A market-making answer needs a bid, an offer and a reason for each. Begin with fair value, then account for estimation uncertainty, spread costs, adverse selection, inventory and limits. Size matters as much as price. If you would quote differently for ten contracts and a thousand, explain the capacity, information and hedge consequences rather than merely widening by instinct.

When an interviewer changes spot, volatility or order flow, update the state before giving another market. Which Greeks moved? Did the trade reveal information, or did inventory alone change your reservation price? Is the best hedge the underlying, another option or reduced quoting? There can be more than one defensible response, but it should preserve a coherent account of position and risk.

Also recognise when not to quote. A stale underlying, broken hedge route, uncertain corporate action or breached limit can make withdrawal the controlled choice. Market making rewards participation, but continuous presence is not sensible when the inputs needed to value and manage the position are unavailable.

Keep mental arithmetic proportionate

Fast percentages, expected values and rough Greek changes are useful because they preserve attention for the market. They are not the job itself. Train for clean estimation and error checking rather than importing a speed target from another firm's branded test.

Technology preparation: use Wolverine's own stack description

The technology page says Wolverine builds low-latency systems for automated trading and market making, streaming market data, risk evaluation and automated hedging. It names C++, C#, Python, XML and SQL, while stressing that teams choose technology according to the system.

Current openings reinforce that range: some explicitly seek C++ for trading platforms or latency-critical systems; another seeks C#. Preparation should follow the advert.

For a latency-sensitive C++ role, revise object lifetime, memory layout, concurrency, measurement and network failure. For C# or data-facing work, runtime behaviour, asynchronous design, SQL and reliable integration may be more relevant. Across both, Wolverine's published priorities are reliability, speed, efficiency and risk. An answer that optimises median latency while weakening controls has missed the business problem.

Research and risk roles

Wolverine's public materials do not define one “quant researcher” curriculum. A researcher supporting options valuation needs numerical methods, volatility modelling and empirical validation. A high-frequency futures role needs microstructure and reliable event data. Asset-management research can involve equity, credit, commodity or volatility strategies.

Risk questions should connect a metric to an action. Greeks, scenarios and VaR are descriptions until you explain data freshness, aggregation, limits, escalation and hedge feasibility under stress.

Turn preparation into candidate evidence

Build a compact evidence pack rather than a list of claimed strengths. One item might be a simple options notebook that compares model value with hedge behaviour under changed volatility and dividends. Another might be a trading system component with latency and correctness tests. A third might be a written review of a forecast that failed outside its development sample.

The artefact is a prompt for discussion, not proof that Wolverine requires a portfolio. Keep it small enough that you can explain every assumption. Be ready to identify the weakest part, the next test and the decision that the work could responsibly support. If the data are simulated or public, say so. If a team built it, separate your contribution.

For behavioural preparation, attach each example to a business consequence. “Improved performance” is vague. Explain whether you reduced stale prices, prevented duplicate orders, shortened a research cycle or exposed model instability before deployment. Then describe the control or measurement that supports the claim. This makes ambition, collaboration and ownership observable without relying on private information about Wolverine's internal environment.

Original Wolverine-style exercises

These are new practice prompts, not leaked or verified Wolverine questions.

Quote after two buys

Your fair value for an option is $2.50. A customer buys twice at your offer. What do you need to know before showing the next market?

Rebuild inventory, Greeks and any hedge already executed. Consider trade size, observed underlying and volatility moves, related markets, customer information and limits. Explain whether you change fair value, spread, size or hedge first. A quote with no position arithmetic is not an answer.

Dividend uncertainty

Two otherwise identical American calls differ only because one stock may pay a special dividend before expiry. How does that affect valuation and hedging?

Discuss the expected price drop, timing uncertainty, early-exercise incentives and scenario treatment. A single adjusted spot input can conceal meaningful event and model risk.

Tail latency versus average latency

A market-data service is faster on average after a rewrite, but its 99.99th-percentile latency is worse. Would you deploy it?

Tie the decision to stale-book risk, event bursts and the downstream strategy. Inspect the distribution and cause, establish a representative benchmark, define a rollback threshold and include correctness. “Low latency” without a percentile and workload is advertising.

Hedge service outage

Automated quoting remains live but the hedge gateway becomes unavailable. What should the system do?

Use exposure, product liquidity and outage duration to define controlled degradation: reduce size, widen, cap inventory or stop quoting. Cover independent health signals, kill controls, recovery and reconciliation of uncertain orders.

Research result with a beautiful surface

An implied-volatility model fits yesterday's surface almost perfectly but produces unstable hedge ratios. Which objective is wrong?

Separate interpolation fit from stable dynamics. Examine regularisation, arbitrage constraints, out-of-sample hedging error and parameter sensitivity. The best visual fit need not be the safest live model.

Behavioural evidence that fits the firm

Wolverine says it values ambitious, entrepreneurial people, collaboration and ownership. Choose examples where those qualities coexist with control:

• you found and fixed a risk before it became an incident;
• you challenged a model with evidence and preserved the working relationship;
• you improved a system after measuring the actual bottleneck; or
• you took an idea from rough prototype to something another person could trust.

Avoid unsupported culture comparisons with SIG, Optiver or other Chicago firms. A corporate careers page is also not independent proof of retention, collegiality or compensation rank.

Frequently Asked Questions

What is Wolverine Trading's interview process?

At least one current Wolverine careers page shows application review, technical assessment, first interview, second virtual interview, on-site final and offer. The exact route is vacancy-specific, so confirm it with the recruiter.

Is every Wolverine interview options-heavy?

No. Options and volatility are central to parts of the business, but current roles also cover high-frequency futures, software, business intelligence, execution services and asset management. Prepare for the named team and instruments.

Which programming languages does Wolverine use?

Wolverine's technology page names C++, C#, Python, XML and SQL and says technology is chosen for the system. Current vacancies include both C++ and C# roles. Use the job description to prioritise.

Where is Wolverine based?

Its headquarters is in Chicago. Current official job descriptions also identify New York and London offices.

Are the questions above genuine Wolverine interview questions?

No. They are original exercises grounded in Wolverine's published businesses and technology. Candidate reports can suggest themes, but no question should be described as genuine without an authoritative source.