Quant Research vs Model Development 

You said you've been using logistic regression most of times. Does it work usually better or is that it makes the modelling smoother and easier to interpret/explain in your experience? As far as I've seen in my short credit risk experience, there's basically no difference between a simple logistic regression and a neural network/random forest performance for probability of default estimation

very interesting, thanks for giving some insight into quant world!

Hey Dimitri, I have been watching your channel a lot recently and I would like to thank you because you have really helped me find a path that I am excited to follow. I am currently in my undergrad and I really love mathematics but I was considering every major besides a math major due to the fact that I didn’t think there were any interesting jobs I could do with a math major. So I settled with data science because I was interested in the model development aspect and I saw a lot of people saying you are constantly learning which is something I realized I want to do for the rest of my life. But then I stumbled upon your videos and you had talked about how data science is currently very muddied and it seemed data science didn’t really have the amount of rigor or math that I wanted in a career. And then I saw you talk about quant finance which is something I knew of but was scared to do because I had thought that every job in quant finance was like quant trading where your job isn’t very stable and it’s extremely high pressure and stress. But you showed me what quant finance really is and I have been extremely interested ever since. I plan on taking the route of a PhD in applied mathematics because I really just want to learn high level and rigorous math and I am highly considering a career in quant research afterwards because I am interested in the rigor, modeling and the opportunity to be constantly learning in my career. However I have some questions, I am currently enrolled in 2 majors at my university, applied mathematics and Statistics, the statistics major on top of the applied math major would only require me to take 5 more stats classes. However, I decided to do both when I was planning on having a career in data science, but now that I want to get a applied math PhD I was wondering if you think it would be smarter and more beneficial to just do the applied math major and focus on my math courses and even do math related projects and research. I am interested to know how much statistics I would need to learn to be a quant and if a statistics major on top of an applied math major would be overkill. I still of course have to take statistics classes but I don’t want to take more than I need to and take away from the math classes that I really care about. Thank you so much and keep up the great content it’s truly helped me so much!

Thank you so much for giving such insights! It helps a lot gauging what actually happens in the Quant world and that not everything in Quant Research is purely Research.

Hey Dimitri, I’m a UK student who is about to be a first year at Cambridge (if exams go well lol) studying economics. I know this isn’t the ideal background to break into quant finance, but would a masters in statistics be enough to get me through the door, especially to entry level analyst roles/ risk modelling/management(essencially I think I would prefer to work on the buy side rather than the sell side, and risk management seemed quite appealing to me). The cambridge econ course is known to be mathematically rigorous for an economics course, although I’m sure its certainly not to the level of a physics computer science or math degree. The relevant maths that I will learn from undergrad is: Linear Algebra(vector spaces, eigenvectors/values) Univariate and Multivariable Calculus Optimisation techniques Limits and continuity(nowhere near as much as a real analysis course unfortunately) ODE’s and PDE’s, and difference equations Regression(OLS, maximum likelihood and GMM estimation) PDF’s, DRV’s etc Time series and ARMA models Lots and lots of econometrics There is a lot I haven’t mentioned simply because I can only list what I can identify in end of year exam papers and some guidance documents, but would this plus a masters in statistics allow me to break in? Would a different masters degree be better(e.g an MFE)? My initial reasoning for doing a masters in statistics is that from what I have seen most top statistics programs seem more accommodating to a wider variety of backgrounds, specifically economics, than top MFE programs. Is there anything that I should be doing alongside my degree(books to read for example) that will help prep me for this career? Thanks for reading, I know I often ramble slightly.

I've recently joined a bank that's doing transaction monitoring. The team is hiring like crazy, however I feel like they shouldn't have to get as many people if they changed their methods. They are currently adding a bunch of 'if statements' (rules) to detect money laundering in transactions. If a transaction triggers on 1 of the if statements, it is send to human review. They have no hierarchy in which rules are important. Not surprisingly they run into false positive issue's. The human review section gets overloaded with transactions that are 'obviously wrong'. I want to tell them they should get a sort of hierarchy in their rules to see which ones do perform well. I don't feel like they want to listen, and perhaps, their goal is just to adhere to policy and not even generate a good model. I'm sort of confused, what do ya think I should do? (it's strange cause I've joined the company for

Hey Dimitri, this is such a great explainer video. Vintage FQ always giving out great information!

can you make a video on the Financial Engineering Masters program in Imperial College London.

Good and detailed as always. But a little misleading no? Rather than quant rsh vs model dev, it was more academic rsh vs model dev. I was hoping for a clearer understanding of Quant Research vs Model Dev - my thoughts are quant researcher is typically at algo driven automated funds, creating automated stat-arb strats for example. while model dev is the work of quant analysts at banks and discretionary funds doing pricing models etc. Would you say Quant Analyst and Quant Researcher are two different jobs, or the same? Obvs both are quants, and the QA title is used interchangeably, but I've not seen a QR doing pricing model - always QA's. I've seen QA's doing QR work though.

Any good programs in Canada for quant finance , maths or econ ? which universities would u recommend because we only have few here ( good ones ), thats only my option and it’s cheaper here relative to usa

Nice jabs on the Data Science side as well 😂... But the community is learning many new things and have started doing more rigorous methods. At least discussions of distribution properties have come in but most projects just assume everything is normal every time every where. The community will use new distributions and Quant based techniques over time.

Does the Markov models use it in the quant field

Hey dimitri, i am from India and i am currently studying MA Economics and i have cleared FRM part 1. I want to get into model development jobs. what can i do to get a job in this field?

So helpful as always Dmitri. Btw I got my post-grad job in tech instead of finance but still a big win :) Just wanted to keep you updated on my journey.