This article was guest-written by Zeke Lee, a Stanford graduate, former management consultant, and former derivatives trader on Wall Street. He founded the GMATPill.com, a video-based GMAT prep course for ambitious MBA candidates, and recently launched the Practice Pill Platform for interactive questions on the web, as well as an iPad app for GMAT practice (iPhone and Android versions coming soon!).
Imagine that a hedge fund approaches you with an enticing proposition: they’ve earned average annual returns in excess of 35% over a 30-year period that saw multiple recessions, stock market crashes and other calamities… and they want you to invest in their new fund.
To put that number in perspective, a 35% average annual return means that $1,000 invested 30 years ago would be worth $8.1 million today.
So how much would you pay in management fees and carry for the “privilege” of investing in such a great fund?
The answer: 5 and 44.
Yes, Renaissance Technologies is well-known for charging a management fee of 5% and carry of 44%, way above the industry average of 2% and 20%… because they’re just that good.
And this scenario is not just in your imagination: their premier fund has actually delivered 35% average annual returns after fees over 30 years (good luck getting an invitation to invest, though).
It’s no coincidence that Renaissance is also a quant hedge fund – the type that employs a small army of mathematicians, physicists, statisticians, and PhD’s all with a singular purpose: make as much money as humanly possible.
Our interviewee today also works at a quant hedge fund, and while he can’t detail his specific trading strategies due to the secrecy of the industry, he will share with you a bunch of insights on the space, including:
- How to make sense of the quant fund space and what types of positions may be available
- What an average day in your life will be like at a quant fund
- How much money you’ll make at the entry-level, and as you move up the ladder
- What prerequisites you need to win interviews at a quant fund, and what kinds of questions you might get in a quant interview
Let’s get started with “quantifying” all of this…
Quantifying Quant Funds & Predicting Positions
Q: I know you don’t want to share too many details of your own story breaking into the industry, so let’s start with an industry overview instead. What is a “quant hedge fund” and what are the different categories of quant funds?
A: Essentially, it’s any fund that relies on statistical techniques and math modeling rather than fundamental analysis to make investments.
A value-oriented hedge fund might try to find companies that are undervalued based on qualitative and quantitative criteria, such as industry positioning, growth potential, and valuation multiples vs. peers in the industry.
A quant hedge fund, on the other hand, would say, “Here are 200 possible conditions that might result in the stock price of Company X rising. Let’s test the statistical significance of all those conditions and use them to predict the direction that the stock is headed in.”
You can divide the industry according to a few different criteria:
- Mathematically Complex Assets: Mortgage-backed securities and many other fixed income securities and variations.
- Less Mathematically Complex Assets: Equities and anything else where the effective yield can’t be stated with a formula.
The quants spend a lot of time calculating metrics like cash flow and arbitrage-free valuation with the first one; the second asset class is less complex, but funds will still use statistical techniques to predict price movements.
Then you can divide the strategies into high-frequency and low-frequency; the former requires strong IT infrastructure, while the latter demands strong math modeling.
Q: OK, so what do you mean by “math modeling” there? I’m assuming this is completely different from the traditional 3-statement modeling and valuation that bankers do?
A: Good quant hedge funds typically treat trading questions as “scientific questions”: for example, what is the probability that MSFT will increase tomorrow?
To answer this question, a fund would study the conditions that made MSFT rise vs. fall historically. They would quantify these conditions and test the statistical significance of their conclusions. Examples:
- Over the past 10 years, whenever Company X stock rose by 10%, MSFT declined by 5%; there’s a 65% correlation between the two.
- Over the past 5 years, whenever trading volume of MSFT exceeded Y, it increased by X% on average; there’s a 50% correlation between the two.
- In the past year, whenever oil exceeded $Z per barrel, MSFT declined by X% on average; there’s a 70% correlation between the two.
So you think about conditions like these, determine the significance and correlations between all of them, and then come up with an overall model that tells you whether an asset such as a stock will increase or decrease in value.
Fundamental shops may go through similar procedures, but some of the conditions may not be quantifiable and they might not test for statistical significance.
Q: So it’s definitely a different kind of modeling than what bankers do – almost like a series of predictive statistical experiments.
What exactly do high frequency funds trade on? And how are they structured?
A: Their strategy is typically a variant of the following: “I know that somebody else will buy X, so let’s buy X first and sell it to them at a higher price.”
Some high frequency funds build algorithms to detect large orders that are about to be executed and front-run them accordingly. So in contrast to low-frequency funds, the high-frequency ones focus more on predicting these types of orders in advance rather than testing for the statistical significance of conditions before trading.
In terms of structure, there are traditional high-frequency funds that work the same way as any other hedge fund or quant fund, but there are also firms that are like “hotels for hedge funds.”
In other words, the company has a number of small teams that operate independently of each other. These teams may or may not be trading the same assets or strategies, and the company and the team members split the revenue. Examples of such firms are Tudor, Millennium, SAC, and Tower.
Each smaller fund has its own PM, team, and track record, but they will follow certain guidelines put together by the parent fund.
Q: I see – so what about the internal structure at those places? Do people specialize or is it more of an ad hoc structure?
A: It depends on the firm – I don’t think you can generalize. At some places, each employee performs a specialized task and “plugs in” his output into a large, automated system. For example, some employees may forecast returns while others will model transaction costs.
Renaissance works a bit like that where the roles and responsibilities are very structured; Citadel is an example of a fund where it’s more ad hoc and people do whatever is most useful at the moment.
Q: Speaking of these specific funds, any thoughts on the big names in the industry and how they all differ from each other?
A: Honestly, most of these places guard information so closely that no one outside really knows what goes on; my comments above are all I know about how some of these funds work.
Big names include Citadel, Renaissance, SAC, Tudor, and Millennium; there’s a good thread on Quora that lists other top names here.
There are some notable spin-off firms such as Two Sigma as well.
Your Mission, If You Choose to Accept It…
Q: Can you tell us about the different roles at a quant hedge fund? For example, at a normal hedge fund you see the Portfolio Manager, Investment Analyst, and Trader roles. How does that structure differ at a quant fund?
A: Those roles still exist, but you see a few additional positions as well. At most quant funds, the 3 main categories of junior-level employees are:
- Traders: Similar to what traders anywhere else do: they execute trades by finding willing buyers and sellers and sometimes also come up with ideas on their own; relationships with brokers are critical here. You see mostly former traders from bulge bracket banks and occasionally from other hedge funds here.
- Quants: They build tools to help traders assess potential rewards and risks of trades, or they try to predict asset returns. This is where you see the small army of statisticians, physicists, and mathematicians.
- Programmers: They develop data access and analytical tools; they have to understand not only the markets and trading strategies, but also how all the software and hardware works, and they must write clean, extensible, and robust code because the software changes all the time.
So, essentially you have 5 roles at quant hedge funds: Portfolio Manager, Investment Analyst, Trader, Quant, and Programmer.
Q: Awesome, thanks for explaining that. With the trend toward quant / “black box” systems, is the role of the trader becoming less important? Do you think they’ll still exist in the long-term?
A: The importance of a human trader increases as assets that people trade become more complex. It decreases as more information becomes quantifiable and people better understand how to price and predict these assets.
So far, the second trend is winning. But there is probably some fundamental lower bound on the importance of traders, because there will always be “un-quantifiable information.”
I think the outlook is probably slightly better for you if you’re interested in joining a quant fund in more of a quant or programmer role, but there will always be demand for top traders no matter how much certain funds switch over to “black box” systems.
A Day in the Life of a Quant
Q: Thanks for weighing in there. I know readers have asked a lot of questions on whether or not certain roles will still exist going forward.
So what’s it like to work at a quant fund? How are the hours? What’s a typical day like?
A: The hours vary; it’s probably around 45-65 hours per week on average depending on what’s going on in the market.
The total entry-level compensation is between $200K and $300K USD.
Compensation thereafter depends on the individual’s and firm’s performances. It could stay the same or potentially go up to the millions.
Q: I just want to pause for a minute so everyone can appreciate the sheer amount of cash you just mentioned: $200K – $300K USD for entry-level positions.
But going back to my actual question, what’s a typical day like for you in terms of schedule and work tasks?
A: Someone in a quant role at a hedge fund might complete the following tasks each day:
- Idea Generation – They’ll read academic research, attend conferences and talk to each other to generate ideas; then they’ll gather data and write programs to test their ideas.
- Turn Ideas Into Trades – They’ll build models to turn their ideas into trades and will monitor the performance and risk of their trades, tweaking their models as necessary.
- Get Better Trades – They’ll talk to traders, dealers, brokers, and so on to get better trading terms, lower transaction fees, and other “concessions” that make it easier to trade profitably.
- Fundraising and Interviewing – These tasks probably take up less time than the rest, but occasionally quants will help with speaking to new investors to raise money and with interviewing new candidates for open quant positions.
You don’t have much of a “set schedule” because it really depends on what’s going on in the market – it’s not quite as crazy as the wildly unpredictable investment banking hours, but I’ve definitely seen people here stay late and work weekends.
So the notion that “the hours will be better” is a bit of a myth, and it depends heavily on the type of fund you’re at and what your role is.
Quantifying the Quant Recruiting Process
Q: For readers who want to get into a quant hedge fund, how would you recommend that they start? Are internships important? Are PhD’s important? What prerequisites are there to get into a top quant fund?
A: My #1 recommendation is to develop a strong understanding of statistics and some programming skills.
Q: Right, I think everyone knows that those skills are required, but what about getting the interview in the first place?
A: There are basically 3 ways to get interviews at quant funds: go through on-campus recruiting (not many funds participate, even at top schools), get friends at hedge funds refer you, or go through headhunters.
It’s just like the networking process for any other role: there’s a certain amount of randomness and luck involved, and you have to grind it out until you start landing interviews.
One really important point here is that the “soft skills” they look for in banking interviews such as leadership experience, charisma, and so on don’t matter at all for most quant interviews.
They’ll ask you questions about past quantitative projects, math/stats questions, and brainteasers, but they don’t care what your “greatest weakness” is or what your “greatest challenge” as a leader was.
Any major or program that helps you develop your knowledge of statistics and programming is helpful: math, statistics, science, engineering, etc.
Master’s degrees are helpful, and PhD’s are required for some firms and roles.
So if you’re a college freshman and you’re sure that you want to go into quant finance, you should pick a major that develops rigorous thinking and communication skills and one where you’ll work with data and do some programming.
Your major may not matter much for investment banking, but it matters a lot here: math, statistics, and computer science are the best choices.
Q: So if you’re a number-crunching machine who can answer brainteasers in your sleep, but you hate talking to people or interacting with carbon-based life-forms, quant roles might be for you.
A: Haha, well, that’s the non-PC way to put it I suppose. I would point out that you still have to interact with your team and communicate your ideas even if you’re in a quant role that’s more technically intense.
So it’s not like you can just sit and hibernate in a cave and never talk to people.
The difference is that your job isn’t dependent on calling people, developing relationships, thinking quickly on your feet, and cutting deals as it would be in the senior levels of banking and PE.
Of course, this is also true of many other roles at hedge funds and asset management firms, so I’m not sure if it’s unique to quant funds or not.
Q: What if you follow the quant route but end up not getting a quant job? What “Plan B” options are there in case you don’t get into your dream hedge fund?
A: Work for technology companies or anywhere where data analysis is important – for example, insurance companies, credit card companies, or social networks that have lots of user-generated data.
Your goals should be to learn statistics very well, gain market knowledge, and make money in the process.
Q: You mentioned before how they don’t care too much about “fuzzy factors” in quant interviews. What types of questions can you expect, beyond the general categories of “lots of math and brainteaser questions”?
A: Interview questions for traders within quant shops test your understanding of market events and intuition for asset pricing models. For example, they might ask you what typically happens to the US mortgage market when treasury bonds rally.
Other likely questions include those based on probability, statistics, and optimization.
One really common probability question in quant interviews is the “Birthday Problem”: what is the probability that two people in a room of 20 have the same birthday?
Another common quant interview question is the Monty Hall problem: you’re at a game show and there are three doors. Behind one door is a gift, and there’s nothing behind the other two doors.
You randomly pick a door, and the show’s host, who knows where the gift is, picks another door and opens it, revealing that there’s nothing behind it.
At this point, you’re given a choice to stick with your original pick or switch to the other door. Should you switch?
Q: So these questions have clear, correct answers, even though they may be counter-intuitive.
The Monty Hall one is interesting because it is in your benefit to switch, even though some people just can’t understand intuitively why that’s the case.
A: Yeah, the questions tend to be mathematically complex or counter-intuitive, or both.
So you need to think out loud in front of your interviewer so they can get a sense of what’s going through your mind. They can also help steer you in the right direction. Simply sitting there thinking in silence won’t help.
The Grand Exit?
So are you planning to stick around at your fund or move onto greener pastures?
A: As with most other buy-side roles, there aren’t real “exit opportunities” here: you stay at your firm and move up the ladder, or you go off to start your own fund instead.
So far I’ve done well here and the next few years seem promising, so I’m going to stick around. Eventually, it would be nice to start my own quant shop and run the show.
Q: Right, string theory + algorithms can be a powerful combination, but once you mix in entrepreneurship with all of that you can do a lot better than mere “binders full of money.”
Thanks for taking the time to speak with us and best of luck!
A: Thanks. Same to you.