Very glad to hear this, as someone who is majoring in Economics this video offers great insight into the applications of econometrics in quant finance!
Hey dimitri, given your background/applied econ degree would you consider making a video about career paths in econometrics and applied econ? I am an undergrad student studying econ and stats.
Econometrics, my love. I loved this course during my Master's degree. 'Scary' at first, maybe, but simply amazing. And yes, I got a top mark at my exam :-)
Good luck to all of your trying to learn econometrics, it needs efforts but it is definitely doable. If you already know stochastic calculus, calculus 2 and Linear algebra, you are on the very right path!!! Good luck guys!
I would love to hear an exposition on econometrics as a Quant. Dr. Marcos Lopez De Prado a renowned Quantitative Researcher and Fund Manager wrote several articles detailing the Mathematics used by Economists that result in inadequate models representing the arbitrage opportunities. Unfortunately, from what I hear econometrics uses dated tools and commits Statistical sins according to ASA (American Statistical Society) and Mathematical Societies in the US.
Instead, he recommends Economists apply Topology, Graph Theory, Discrete Mathematics, and deeper abstract mathematical concepts to further have an accurate model of representing markets. Given his extensive background and pedigree, evidently he even took a second PhD when he was already an Economics PhD he does make some sound statements on how mathematics in economics keeps getting the wrong assumptions where he suggests an non-linear econometric approach.
Hey Dimitri, in your opinion what are the programming languages / statistical packages most important to know for quantitative finance? I've seen most master programs teach C++, Java, Python, R or Matlab, but I'd like to know what do you think based on your experience
It depends on what section you work in. I can only talk about prop trading, primarily HFT/Low Latency, and here the primary language for low-level coding is C++ as it has very convenient low-level implementation in terms of interacting with the hardware (GPUs, custom chips etc). This is handled by the software and hardware engineers. As a "quant", be it a researcher or trader, you rely mostly on SQL and Python. Daily. A lot.
It depends on the area you are working in. For time-series (my area) I don't think you can beat SAS. It is setup to do time-series at an expert level. David Dickey worked with SAS which is why it is a lot better than free software. For stats programming that's free, I like R the best. They have more useful packages and I trust it more than Python. For data science/machine learning Python will be the best. A lot of banks have been adding Python for data science. All the above I consider quant work as it is more focused on quickly being able to test model theories and model development. If any of that work is used in an application from banking to HFT, it will be implemented in a low level language like C++, C#, Java, or maybe Python. I still feel C++ is king when it comes to low level programming. If you want to work in C++, check out the Boost library as a lot of quants use it.
Hey Dimitri! In one of your past videos you mentioned that right after graduation you wanted to land a trading position. My question is why did you want to be a trader and what were the reasons that made you pursue a career in risk management?
what about running simulations (not really econometrics, but you are running statistics on financial data so econometrics)? Curious to see if IV regressions are used at all in industry or focus on casual methods like regression discontinuity? t seems that realistically the econometrics course you need are basic regression (OLS, logistic), time series (ARIMA, Garch), and advance econometrics (SEM, SUR, Markov Chain, panel data).
Hey Dimitri, which of the following paths would think is better for getting into quant finance? 1: Double Major in Economics and Finance and a MSc in Econometris 2: Finance Degree and a MSc in Data Science Note that this BSc in Finance is pretty quantitative though I'd apreciate your answer. Thank you!
To me, econometrics is all about causality and carefully selecting parameters and instrument variables. How do they bridge the gap between something like OLS and neural networks, which are essentially black boxes? I know things like causal random forests exist but I’m just curious
This is the current struggle that quant finance is facing. For applications in banking, we need to understand what is going on for many applications and how regulators will react is another hurdle.
Hi Dimitri, I’ve watched all your videos on R programming and time series analysis in R. Do you know of any guides (books, TH-cam series, skill share) for learning the actual ins and out of programming in R? Thanks mate!
Below is an affiliate link to my favorite R book. The reason I like this book the best is because it covers how to actually program in R instead of how to write code for statistics or a specific problem. If you know the language well enough you can troubleshoot quicker and you can write your own packages fairly easily. The Art of R Programming amzn.to/2G1u3Ua
Do you think computer science or econometrics would be a better choice of major for someone interested in pursuing a career in quantitative finance? Specifally a trading role.
I'd focus on statistics or econometrics. Also, an lot of traders at quant firms have business degrees. The skills required to trade are different than being a quant. They do need a basic understanding of methods though.
It is a possible route however it is not the most common or likely. For example, I have a masters in applied economics where I focused on financial engineering and statistics.
Great video. Thanks for answering my question Dimitri
Thanks for asking! In the first 20 hours it is ranking #1 out of the last 10 published,
Very glad to hear this, as someone who is majoring in Economics this video offers great insight into the applications of econometrics in quant finance!
Hey dimitri, given your background/applied econ degree would you consider making a video about career paths in econometrics and applied econ? I am an undergrad student studying econ and stats.
Found this super useful! Thanks Dimitri!
MBS prepayment modeling is also econometrics
Oh that’s a fun one!
Econometrics, my love. I loved this course during my Master's degree. 'Scary' at first, maybe, but simply amazing. And yes, I got a top mark at my exam :-)
Good luck to all of your trying to learn econometrics, it needs efforts but it is definitely doable. If you already know stochastic calculus, calculus 2 and Linear algebra, you are on the very right path!!! Good luck guys!
curious to hear your thoughts on SAS and its future in banking and, frankly, its viability as a (profitable) going-concern as R and Python take over
I would love to hear an exposition on econometrics as a Quant. Dr. Marcos Lopez De Prado a renowned Quantitative Researcher and Fund Manager wrote several articles detailing the Mathematics used by Economists that result in inadequate models representing the arbitrage opportunities. Unfortunately, from what I hear econometrics uses dated tools and commits Statistical sins according to ASA (American Statistical Society) and Mathematical Societies in the US.
Instead, he recommends Economists apply Topology, Graph Theory, Discrete Mathematics, and deeper abstract mathematical concepts to further have an accurate model of representing markets. Given his extensive background and pedigree, evidently he even took a second PhD when he was already an Economics PhD he does make some sound statements on how mathematics in economics keeps getting the wrong assumptions where he suggests an non-linear econometric approach.
Thank you for the video, would be amazing to know which are the most common optimization methods quants or portfolio managers use.
Hey Dimitri, in your opinion what are the programming languages / statistical packages most important to know for quantitative finance? I've seen most master programs teach C++, Java, Python, R or Matlab, but I'd like to know what do you think based on your experience
It depends on what section you work in. I can only talk about prop trading, primarily HFT/Low Latency, and here the primary language for low-level coding is C++ as it has very convenient low-level implementation in terms of interacting with the hardware (GPUs, custom chips etc). This is handled by the software and hardware engineers. As a "quant", be it a researcher or trader, you rely mostly on SQL and Python. Daily. A lot.
It depends on the area you are working in. For time-series (my area) I don't think you can beat SAS. It is setup to do time-series at an expert level. David Dickey worked with SAS which is why it is a lot better than free software. For stats programming that's free, I like R the best. They have more useful packages and I trust it more than Python. For data science/machine learning Python will be the best. A lot of banks have been adding Python for data science. All the above I consider quant work as it is more focused on quickly being able to test model theories and model development. If any of that work is used in an application from banking to HFT, it will be implemented in a low level language like C++, C#, Java, or maybe Python. I still feel C++ is king when it comes to low level programming. If you want to work in C++, check out the Boost library as a lot of quants use it.
Thank you both for your reply!!
Hey Dimitri! In one of your past videos you mentioned that right after graduation you wanted to land a trading position. My question is why did you want to be a trader and what were the reasons that made you pursue a career in risk management?
I'll consider making a video on this.
what about running simulations (not really econometrics, but you are running statistics on financial data so econometrics)?
Curious to see if IV regressions are used at all in industry or focus on casual methods like regression discontinuity?
t seems that realistically the econometrics course you need are basic regression (OLS, logistic), time series (ARIMA, Garch), and advance econometrics (SEM, SUR, Markov Chain, panel data).
I haven't used IV regression for work but I'm sure it is used somewhere. Simulations are used a lot in the market risk side.
IV is really not useful in industry.
Hey Dimitri, which of the following paths would think is better for getting into quant finance?
1: Double Major in Economics and Finance and a MSc in Econometris
2: Finance Degree and a MSc in Data Science
Note that this BSc in Finance is pretty quantitative though
I'd apreciate your answer. Thank you!
To me, econometrics is all about causality and carefully selecting parameters and instrument variables. How do they bridge the gap between something like OLS and neural networks, which are essentially black boxes? I know things like causal random forests exist but I’m just curious
This is the current struggle that quant finance is facing. For applications in banking, we need to understand what is going on for many applications and how regulators will react is another hurdle.
Hi Dimitri, I’ve watched all your videos on R programming and time series analysis in R. Do you know of any guides (books, TH-cam series, skill share) for learning the actual ins and out of programming in R? Thanks mate!
Below is an affiliate link to my favorite R book. The reason I like this book the best is because it covers how to actually program in R instead of how to write code for statistics or a specific problem. If you know the language well enough you can troubleshoot quicker and you can write your own packages fairly easily.
The Art of R Programming
amzn.to/2G1u3Ua
@@DimitriBianco Perfect, that seems spot on to what I was looking for. Many thanks.
What about the other aspects of economics such as micro theory or macro theory? Are they at all touched upon in quant jobs?
They are used but from a theoretical perspective. For example, discussing why a variable should or shouldn't be in a model.
“CCAR” Comprehensive Capital Analysis Review*
Do you think computer science or econometrics would be a better choice of major for someone interested in pursuing a career in quantitative finance? Specifally a trading role.
Econometrics in my opinion gives you a way more rigorous understanding of the math while cs focusses more on application
I'd focus on statistics or econometrics. Also, an lot of traders at quant firms have business degrees. The skills required to trade are different than being a quant. They do need a basic understanding of methods though.
@@DimitriBianco Thanks I appreciate it
Hi from Brazil!
Is econometrics degree suitable to become quant?
It is a possible route however it is not the most common or likely. For example, I have a masters in applied economics where I focused on financial engineering and statistics.