lordkebab/How to Get a Quant Job in Finance (Joshi)

## How to Get a Quant Job in Finance (Joshi)
How to Get a Quant Job in Finance

By Mark Joshi, December 2014

First, let’s be clear there are many different sorts of finance jobs that require mathematics to varying levels of ability and with very different characteristics in terms of both day to day work and hiring.

The classic “quant” job involves learning high-level stochastic calculus to develop pricing models for complex derivatives securities using the Black-Scholes model and various extensions of it. These jobs in finance are much scarcer on the ground than they used to be. The modern focus is often more about how to take additional factors such as credit and liquidity into account than about pricing ever more complex securities. For example, CVA desks (credit-valuation adjustment) are tasked with pricing and managing the credit risk that arises from trading derivatives in other asset classes.

Be clear that even such classical quants do not spend their day sitting with a pencil and paper doing mathematics. Most of their work lies in implementing tweaks to existing pricing models and in reading other people’s papers in order to find out what the latest techniques are. In many ways, most of the challenges of their job lie in numerical methods. How to evaluate an integral quickly, how to compute sensitivities, how to eliminate a numerical instability, how to resolve the fact that two models are giving different prices when they shouldn’t? Hirers of entry level people are generally looking for mathematical quickness, strong coding skills, practicality and some knowledge of financial maths.

Such quant jobs may be in the front or middle office. To be in the front office means to be working close to the traders. As such the jobs are closer to where the money is made and so can lead to greater salaries. They also offer the greatest potential for moving into trading where the real money is. It is not unusual for a bank’s highest paid employee to be a trader. Middle office pay less but can be more interesting in that schedules are less tight and there tends to be more emphasis on understanding rather than on how to get a number as quickly as possible. The front office tends to be a bit snobby regarding the middle office, however. So whilst starting in the middle office can be a good way of getting your foot in the door, if your objective is to work in the front office, don’t spend more than a year or two there.

The interview process for such finance jobs tends to be done directly by the hiring team. As such it is generally focussed on technical questions which can range from derive the Black-Scholes equation to do a contour integral to how many piano tuners are there in New York City? Brainteasers are popular. Soft behavioural-type questions are uncommon for these sorts of roles but can occur. Fortunately (or perhaps not), interviewers tend to be unimaginative in their questions and reuse ones they have been asked. This means that careful study of books of interview questions (including my own!) definitely pays off.

First, you have to get the interview, however. Quant teams tend to hire directly via personal contacts and failing that adverts and recruiters. (i.e. headhunters.) So try to get in touch with anyone you know who already has such a job and get some advice and contacts. Think ahead and if you are contemplating a move in a couple of years, try to build connections with those further ahead in the process for when the time comes.

Classical quant jobs are sometimes called “Q” measure jobs because the pricing is done in the risk-neutral or other martingale measure rather than in the real-world “P” measure. In the last 10 years, there has been a lot of growth in P-measure jobs. These are sometimes called statistical-arbitrage jobs or “stat arb” jobs. Here the objective is to use mathematics/statistics/computers to predict short-term movements of asset prices and then to execute trades as quickly as possible. Short term can mean anything from milliseconds to hours. These sorts of jobs tend to be in hedge funds. Hirers tend to be looking for strong programming and data management skills as much as mathematical ability. Study of classical financial mathematics is not really relevant for this sort of job since it deals with the pricing and hedging of derivatives across longer time periods rather than the movement of fundamental asset prices. Interviews will therefore be focussed on coding questions such as how to manage complex data structures rather than how to derive the Black-Scholes equation. Expect the usual plethora of brainteasers, however. eg. “what is your confidence interval for the population of X?” followed by “how much are you willing to bet on that?”

As well as these very mathematically intense jobs, there is a whole range of jobs working in banks that require mathematical skills. Post the GFC, the issue of capital modelling has become huge and this will never go away. Here the problem is how to model the bank’s losses in a very bad scenario - eg a 1 in 3000 year event. Ideally, the bank should hold enough capital to weather such events. Of course, the GFC made a mockery of many bank’s capital models. The 1 in 3000 years event only took a couple of years to arrive and then the bank didn’t weather it. As with terrorism, the failure of the experts to see what was coming has resulted in a massive increase in the number of jobs and resources put into the area rather than a wholesale sacking. Whilst maths is important for these jobs, it is also about data gathering and understanding, and interaction with senior management. How can we model events that have never occurred?

In these days of big data, there are increasingly many jobs in banks that consist of trying to analyse retail client behaviour. This may mean assessing their credit risk. Computer algorithms long ago replaced manager judgment for credit cards and personal loans. Or it may mean trying to detect fraudulent behaviour.  The principal skills here are statistical.

Risk management in general is a semi-mathematical discipline. There are many different time horizons. Often it is ten-day market risk that the regulators demand. So the problem becomes how to build models for how assets co-move across ten days and work out their tail movements. Since the answers are never precise and concrete, judgement and people skills play their part here as well as mathematical skills. It is possible to get into these sorts of roles via graduate analyst programs. Such programs tend to focus on grades when choosing for interview. The interviews are often much more behavioural than technical. As well as roles in market risk, there is increasing need for the modelling of credit portfolios. This is often in tandem with derivatives pricing. For example, suppose a long-dated contract has been entered to exchange variable cash-flows with a company. Under certain market conditions, this may result in the company owing  the bank a lot of money.There is then the question of how to model the bank’s credit exposure even if initially the contract is balanced. Again the skills tend to be a mix of mathematics, judgement and management handling.

In all areas, it is a bad no-no if you are not aware of banking news. What do banks do? How do they make money? What actually happened in the GFC? Why has this particular bank been in the news lately? Why are bankers arguing about the bonus cap? What is the price of oil today? Why is that important? What was regulation Q? What is a Euro-bond and why are they important historically? How do you play “Liar’s poker”?

In conclusion, getting a quant job in banking requires careful preparation, leveraging of personal contacts, a genuine interest in the work and plenty of willingness to work hard!

Mark Joshi obtained a B.A. in mathematics (top of year) from the University of Oxford in 1990, and a Ph.D. in pure mathematics from the Massachusetts Institute of Technology in 1994. He was an assistant lecturer in the department of pure mathematics and mathematical statistics at Cambridge University from 1994 to 1999. Following which he worked for the Royal Bank of Scotland from 1999 to 2005 as a quantitative analyst at a variety of levels, finishing as the Head of Quantitative Research for Group Risk Management. He joined the Centre for Actuarial Studies at the University of Melbourne in November 2005 as an associate professor and is now a full professor.
	How to Get a Quant Job in Finance

	By Mark Joshi, December 2014

	First, let’s be clear there are many different sorts of finance jobs that require mathematics to varying levels of ability and with very different characteristics in terms of both day to day work and hiring.

	The classic “quant” job involves learning high-level stochastic calculus to develop pricing models for complex derivatives securities using the Black-Scholes model and various extensions of it. These jobs in finance are much scarcer on the ground than they used to be. The modern focus is often more about how to take additional factors such as credit and liquidity into account than about pricing ever more complex securities. For example, CVA desks (credit-valuation adjustment) are tasked with pricing and managing the credit risk that arises from trading derivatives in other asset classes.

	Be clear that even such classical quants do not spend their day sitting with a pencil and paper doing mathematics. Most of their work lies in implementing tweaks to existing pricing models and in reading other people’s papers in order to find out what the latest techniques are. In many ways, most of the challenges of their job lie in numerical methods. How to evaluate an integral quickly, how to compute sensitivities, how to eliminate a numerical instability, how to resolve the fact that two models are giving different prices when they shouldn’t? Hirers of entry level people are generally looking for mathematical quickness, strong coding skills, practicality and some knowledge of financial maths.

	Such quant jobs may be in the front or middle office. To be in the front office means to be working close to the traders. As such the jobs are closer to where the money is made and so can lead to greater salaries. They also offer the greatest potential for moving into trading where the real money is. It is not unusual for a bank’s highest paid employee to be a trader. Middle office pay less but can be more interesting in that schedules are less tight and there tends to be more emphasis on understanding rather than on how to get a number as quickly as possible. The front office tends to be a bit snobby regarding the middle office, however. So whilst starting in the middle office can be a good way of getting your foot in the door, if your objective is to work in the front office, don’t spend more than a year or two there.

	The interview process for such finance jobs tends to be done directly by the hiring team. As such it is generally focussed on technical questions which can range from derive the Black-Scholes equation to do a contour integral to how many piano tuners are there in New York City? Brainteasers are popular. Soft behavioural-type questions are uncommon for these sorts of roles but can occur. Fortunately (or perhaps not), interviewers tend to be unimaginative in their questions and reuse ones they have been asked. This means that careful study of books of interview questions (including my own!) definitely pays off.

	First, you have to get the interview, however. Quant teams tend to hire directly via personal contacts and failing that adverts and recruiters. (i.e. headhunters.) So try to get in touch with anyone you know who already has such a job and get some advice and contacts. Think ahead and if you are contemplating a move in a couple of years, try to build connections with those further ahead in the process for when the time comes.

	Classical quant jobs are sometimes called “Q” measure jobs because the pricing is done in the risk-neutral or other martingale measure rather than in the real-world “P” measure. In the last 10 years, there has been a lot of growth in P-measure jobs. These are sometimes called statistical-arbitrage jobs or “stat arb” jobs. Here the objective is to use mathematics/statistics/computers to predict short-term movements of asset prices and then to execute trades as quickly as possible. Short term can mean anything from milliseconds to hours. These sorts of jobs tend to be in hedge funds. Hirers tend to be looking for strong programming and data management skills as much as mathematical ability. Study of classical financial mathematics is not really relevant for this sort of job since it deals with the pricing and hedging of derivatives across longer time periods rather than the movement of fundamental asset prices. Interviews will therefore be focussed on coding questions such as how to manage complex data structures rather than how to derive the Black-Scholes equation. Expect the usual plethora of brainteasers, however. eg. “what is your confidence interval for the population of X?” followed by “how much are you willing to bet on that?”

	As well as these very mathematically intense jobs, there is a whole range of jobs working in banks that require mathematical skills. Post the GFC, the issue of capital modelling has become huge and this will never go away. Here the problem is how to model the bank’s losses in a very bad scenario - eg a 1 in 3000 year event. Ideally, the bank should hold enough capital to weather such events. Of course, the GFC made a mockery of many bank’s capital models. The 1 in 3000 years event only took a couple of years to arrive and then the bank didn’t weather it. As with terrorism, the failure of the experts to see what was coming has resulted in a massive increase in the number of jobs and resources put into the area rather than a wholesale sacking. Whilst maths is important for these jobs, it is also about data gathering and understanding, and interaction with senior management. How can we model events that have never occurred?

	In these days of big data, there are increasingly many jobs in banks that consist of trying to analyse retail client behaviour. This may mean assessing their credit risk. Computer algorithms long ago replaced manager judgment for credit cards and personal loans. Or it may mean trying to detect fraudulent behaviour. The principal skills here are statistical.

	Risk management in general is a semi-mathematical discipline. There are many different time horizons. Often it is ten-day market risk that the regulators demand. So the problem becomes how to build models for how assets co-move across ten days and work out their tail movements. Since the answers are never precise and concrete, judgement and people skills play their part here as well as mathematical skills. It is possible to get into these sorts of roles via graduate analyst programs. Such programs tend to focus on grades when choosing for interview. The interviews are often much more behavioural than technical. As well as roles in market risk, there is increasing need for the modelling of credit portfolios. This is often in tandem with derivatives pricing. For example, suppose a long-dated contract has been entered to exchange variable cash-flows with a company. Under certain market conditions, this may result in the company owing the bank a lot of money.There is then the question of how to model the bank’s credit exposure even if initially the contract is balanced. Again the skills tend to be a mix of mathematics, judgement and management handling.

	In all areas, it is a bad no-no if you are not aware of banking news. What do banks do? How do they make money? What actually happened in the GFC? Why has this particular bank been in the news lately? Why are bankers arguing about the bonus cap? What is the price of oil today? Why is that important? What was regulation Q? What is a Euro-bond and why are they important historically? How do you play “Liar’s poker”?

	In conclusion, getting a quant job in banking requires careful preparation, leveraging of personal contacts, a genuine interest in the work and plenty of willingness to work hard!

	Mark Joshi obtained a B.A. in mathematics (top of year) from the University of Oxford in 1990, and a Ph.D. in pure mathematics from the Massachusetts Institute of Technology in 1994. He was an assistant lecturer in the department of pure mathematics and mathematical statistics at Cambridge University from 1994 to 1999. Following which he worked for the Royal Bank of Scotland from 1999 to 2005 as a quantitative analyst at a variety of levels, finishing as the Head of Quantitative Research for Group Risk Management. He joined the Centre for Actuarial Studies at the University of Melbourne in November 2005 as an associate professor and is now a full professor.