What do Quants do?
SearchTeam asked Quantitative Analysis Expert Marek Bukacek what actually Quantitative Analysts (Quants) do?
1. What do you do as a Quantitative Analyst?
The role of quantitative analyst (quant) is very variable, it differs with the need of an employer, the set of assigned tasks and the personality of employee. Generally, quant is responsible for answering some complex question by a comprehensive answer based on sophisticated facts. Mostly they use mathematical and statistical tools to analyze any available data, quantify key variables and build the answer on it.
In my company (financial data provider), I am responsible for data flow and data quality. My team monitors incoming data, validates calculations and checks output that we are providing to our clients. In case of any detected issue, we have to patch the problem and improve the system to prevent it. That implies we must understand the whole concept, be able to communicate with the other team within the company (programmers, economical guys, …) and with clients as well.
2. How can somebody become a Quantitative Analyst?
You became quant when you focus on the task instead of a method you are using (as mathematicians do).
Anyone thinking of this job should get experiences with mathematical and statistical tools starting with linear algebra, touching differential calculus and get deep understanding of regression, clustering or principal component analysis. You don’t have to be the expert; you just need to have large portfolio of tools that you can use and correctly interpret.
Then, you should start applying these tools in various fields to understand needs of experts with completely different background, from doctors over engineers to heads of marketing or politics.
When you start understanding their problems and when you are able to solve them, you are in the good path.
3. What are the traits of an excellent Quantitative Analyst?
There is a large set of qualities that are beneficial in many positions, quant should be definitely smart, agile and dutiful. But there are several key fey features that distinguish quant from others.
Empathy. Quantitative analyst must understand given task. He is communicating with people who don’t understand the language of math or technology. It is his responsibility to convert a discussion to well structured project plan and it is his responsibility to deliver output in form that a client wants.
Project oriented approach. When a task is set, quant have to solve it the most efficient way. It might be tempting to apply new interesting method, but one should be aware of wasting of resources.
Know your limits. Any applied method has its limits. Data are often noisy, mathematical methods are builds on assumptions that are not completely fulfilled, and computer simulation may lack of details. Therefore good quantitative analyst should attach comprehensible explanation of any inaccuracy affecting the result. Clients must be aware of the risk they could carry.
4. Do I have to know some programming language to become a Quantitative Analyst?
Such skill is (fortunately) not required. As quants are using math and statistical tools to analyze data, they have to operate with some software dedicated for them. It depends of the preference whether you can do everything in MS Excel or you will use MATLAB, R-studio or any other sophisticated software.
In case a company wants to incorporate quantitative analytical tools to their production system, it will be probably written in Java, C++, Python or other language. But it will be the responsibility of programmers to do it.
On the other hand, it is beneficial to know the algoritmization principles at least to organize your work and to be able to explain to programmers how they should implement your tools.
5. In what companies are Quantitative Analyst positions?
Generally, all decent size companies dealing with data should have at least one guy responsible for quantitative analysis.
Larger companies would have a team with specific positions covering data mining, modeling, risk analysis or reporting. In this case, the real quantitative analyst would be the guy with the big picture knowledge, may be in the position of manager or team leader.
The most common branches consuming quants are finance, consultant services and a research generally. The most developing sector nowadays is marketing due to the data from social media.
6. What is the best way to get an entry level Quantitative Analyst job?
Probably you should start at some junior position in any department dealing with data. Most likely you will be responsible for some routine job. Think about it, learn why are you doing that and understand why you are applying predefined method. If you do so, next time you will be able to improve it or even come with your approach.
7. What are the main challenges of the Quantitative Analyst job?
From my perspective, the communication with people without technical background is the most demanding part. Quant has to understand their needs, he has to explain methods he will use and present the results at the end. This process does not have to be smooth; you have to balance client’s decisions and your opinions.
8. What do you like mostly about Quantitative Analyst job?
I like the opportunity to participate various project from beginning to the end, to see the impact of my work and to work with experts from different field.
9. What in your opinion is a future of quantitative analysis?
I think quants will work for wider band of companies than now. With increasing usage of technology in daily routine it is much easier to get relevant data from all aspect of life and with this data one can analyses and optimizes any human’s activity.
The Artificial Intelligence is just a word, there is no sign that a smart computer could replace skilled expert who is actually using computers. Modern fashionable tools as neural networks with deep learning are just another method that could be used by quant to solve a problem. I can see only synergy, not the risk.
10. Are there any good quantitative analysis courses? Are there any quantitative
I am not aware of any. The practice combined by guided training in mathematical and stochastic tools is the best way to move forward.