So where are the results of education and "research"? All professors support the same policies that really originate from political establishment. All they want is free pass to print paper money backed by nothing to support their political agenda that is destroying the economy on purpose. They need professors on TV with economic title to justify their policies. All professors talk the same "economics" conveying slightly different fear level (good cop, bad cop). No one comes with independent opinion that would differ from others. By now we should have some sound economic advice coming from all the geniuses with Nobel prizes. Where are they when you need them since we are in difficult economic environment for years now? Oh, I know, it is about money. Those who asks for opinion pay their salary.
Ernie Chan claims that statistical arbitrage is better than momentum like trend
- Thread starter GloriaBrown
- Start date
So where are the results of education and "research"? All professors support the same policies that really originate from political establishment. All they want is free pass to print paper money backed by nothing to support their political agenda that is destroying the economy on purpose. They need professors on TV with economic title to justify their policies. All professors talk the same "economics" conveying slightly different fear level (good cop, bad cop). No one comes with independent opinion that would differ from others. By now we should have some sound economic advice coming from all the geniuses with Nobel prizes. Where are they when you need them since we are in difficult economic environment for years now? Oh, I know, it is about money. Those who asks for opinion pay their salary.
Two words: Ed Thorp.
Yeah, I have to say that I disagree with @vicirek here, although I usually respect his opinion on stuff. My point was never that economists are market wizards, although I also disagree with you about this idea that top econ professors are just trying to collect a check. I was saying that most professors are passionate about their fields and that their work isn't invalid just because they don't work on Wall Street.
Well, when was the last time you found some highly useful and insightful content in academic papers? And limit that to teaching academicians. I did indeed find a lot of good stuff but mostly written by PhD candidates or other researchers that later on proceeded straight into finance space.
Take Paul Krugman as prime example: This is the guy who boldly pronounced "Greece will leave the Euro for sure" and "Europe will be done in a matter of weeks" (during the height of the crisis). Or take Joseph Stiglitz: A PAID poster boy for Greece who was subsequently eaten alive by hedge fund trader Hugh Hendry in front of cameras
Both (Krugman and Stiglitz) are Nobel Price laureates. Neither have provided much of value that was/is implemented into today's economic landscape nor helped to explain economic relationships that nobody understood before despite their nominations and awards. Another example is the still stubborn belief among leading economic professors that printing of money by central banks always leads to inflation. This line of thought is so disconnected from reality that empirical evidence should serve to disprove such academic theories. Inflation has nothing to do directly with how much money is put into circulation. It has to do with demand for and supply of goods, one among many, money. If the money that is put into circulation is not used but parked then such money creates zero inflation. That is exactly what we are seeing today, however, this empirical evidence is vehemently rejected by "leading" academicians".
Most professors are not passionate about their field, sorry if I starkly disagree with you. If you had passionate professors then all the more power to you but it certainly is not the norm, even at top schools. But I like to get back to the main point, which is not how enthusiastic professors are but what value they add to our actual understanding of finance and economics.
Take Paul Krugman as prime example: This is the guy who boldly pronounced "Greece will leave the Euro for sure" and "Europe will be done in a matter of weeks" (during the height of the crisis). Or take Joseph Stiglitz: A PAID poster boy for Greece who was subsequently eaten alive by hedge fund trader Hugh Hendry in front of cameras
Both (Krugman and Stiglitz) are Nobel Price laureates. Neither have provided much of value that was/is implemented into today's economic landscape nor helped to explain economic relationships that nobody understood before despite their nominations and awards. Another example is the still stubborn belief among leading economic professors that printing of money by central banks always leads to inflation. This line of thought is so disconnected from reality that empirical evidence should serve to disprove such academic theories. Inflation has nothing to do directly with how much money is put into circulation. It has to do with demand for and supply of goods, one among many, money. If the money that is put into circulation is not used but parked then such money creates zero inflation. That is exactly what we are seeing today, however, this empirical evidence is vehemently rejected by "leading" academicians".
Most professors are not passionate about their field, sorry if I starkly disagree with you. If you had passionate professors then all the more power to you but it certainly is not the norm, even at top schools. But I like to get back to the main point, which is not how enthusiastic professors are but what value they add to our actual understanding of finance and economics.
Last edited:
I'm not really willing to debate the premise that economists are right or wrong or have their heads in the sands, or anything like that. I think econometrics is a field with numerous applications to trading, so maybe that counts as me backing economists, but I don't content that Krugman or any of these guys are worthwhile, I'm not even interested in economics--I haven't taken an econ course besides econometrics and its prereqs. I don't condemn or validate it as a field, it's just not of interest to me outside of econometrics.
What I was saying is that in my experience in the CS, math, stats, and financial engineering departments at my school I've had numerous professors--both former quants and career academics--who care both about their fields and their students. If you don't believe they've made valid contributions or that they care for my education, that's fine, but CS, math, stats, and financial engineering are all empirical fields and it takes capable and motivated individuals to become professors. They don't just teach because they can't make it on Wall Street. If you want to see particular research papers go to columbia.edu or ssrn, I'm not going to debate the validity of their research, but I am going to say that it's my opinion that they generally care about advancing their fields and their students. If you don't think they have what it takes to hack it on Wall Street you can find a lot of their resumes online too. I'm not really interested in defending the likes of Krugman, but I think some professors efforts both as researchers and as educators at least deserve acknowledgement
What I was saying is that in my experience in the CS, math, stats, and financial engineering departments at my school I've had numerous professors--both former quants and career academics--who care both about their fields and their students. If you don't believe they've made valid contributions or that they care for my education, that's fine, but CS, math, stats, and financial engineering are all empirical fields and it takes capable and motivated individuals to become professors. They don't just teach because they can't make it on Wall Street. If you want to see particular research papers go to columbia.edu or ssrn, I'm not going to debate the validity of their research, but I am going to say that it's my opinion that they generally care about advancing their fields and their students. If you don't think they have what it takes to hack it on Wall Street you can find a lot of their resumes online too. I'm not really interested in defending the likes of Krugman, but I think some professors efforts both as researchers and as educators at least deserve acknowledgement
Actually, maybe this will be a good learning experience. Pick any research paper from the CS, math, or OR:FE department here and we can both read it and debate it if you want.
Fair points that some deserve acknowledgement. Emphasis on some. Most do not. We also do not sit down and nitpick the few self-less, concerned, and respectful to society at large practitioners at "Wall Street". It is a minority.
Else, academia would pump out much better prepared students, students with high ethics and moral values, students that are very solid in the basics of their fields, students that learned to think outside the box. My estimate of students that possess such are broadly a maximum of 10 out of 100 (according to CVs that passed my desk). My point is that there is a causal relationship between the quality of teachers/professors and the quality of students they "educate" as well as there is a causal relationship between the quality/motivation/passion of professors and the quality of the academic literature they publish. Every professor has to publish papers...that fact is not debated, it is the lack of quality, the lack of applicability of said research that I find in most academic research that that imho directly leads to the quality and motivation of the originating author.
For example why is it that there is hardly any groundbreaking research performed in hft space? Why do most academicians limits themselves to the limitations of some R packages rather than turning it around and find packages that serve their needs and then write code themselves to perform those analysis certain R packages do not manage? Pretty much anything I read these days that deals with hft in equity or currency space is utter garbage and entirely useless. The more complex some of those papers are the less applicable often times. Lately the first thing I do is to read the conclusion and summary statistics. And that is where I most often toss the paper away because some elaborate statistical techniques lead to results that are so poor that it becomes apparent the author seems to lack an understanding of what most likely works or not in the real world. How some pretty well known academic authors can still today work on papers that research the efficiency of popular technical indicators only to conclude that technical analysis does not work anymore is entirely beyond my understanding. It should be obvious that this is the case for anyone who has even only paper traded. ...
Else, academia would pump out much better prepared students, students with high ethics and moral values, students that are very solid in the basics of their fields, students that learned to think outside the box. My estimate of students that possess such are broadly a maximum of 10 out of 100 (according to CVs that passed my desk). My point is that there is a causal relationship between the quality of teachers/professors and the quality of students they "educate" as well as there is a causal relationship between the quality/motivation/passion of professors and the quality of the academic literature they publish. Every professor has to publish papers...that fact is not debated, it is the lack of quality, the lack of applicability of said research that I find in most academic research that that imho directly leads to the quality and motivation of the originating author.
For example why is it that there is hardly any groundbreaking research performed in hft space? Why do most academicians limits themselves to the limitations of some R packages rather than turning it around and find packages that serve their needs and then write code themselves to perform those analysis certain R packages do not manage? Pretty much anything I read these days that deals with hft in equity or currency space is utter garbage and entirely useless. The more complex some of those papers are the less applicable often times. Lately the first thing I do is to read the conclusion and summary statistics. And that is where I most often toss the paper away because some elaborate statistical techniques lead to results that are so poor that it becomes apparent the author seems to lack an understanding of what most likely works or not in the real world. How some pretty well known academic authors can still today work on papers that research the efficiency of popular technical indicators only to conclude that technical analysis does not work anymore is entirely beyond my understanding. It should be obvious that this is the case for anyone who has even only paper traded. ...
Last edited:
There is definitely a deficit of research in HFT. Most of the HFT research I've seen is essentially the same paper time and time again reworded trying to prove or disprove the impacts on liquidity and volatility. That's obviously of no use to either of us, and really boring. There's not much research about HFT, but there definitely is research regarding things that are useful for HFT. My school has some:
http://cfe.columbia.edu/research-FE
Stanford also has a lot of really good stuff (particularly on vol arb) some of it even specifically HFT related.
I've found good stuff from NYU too. Not all of the stuff from these schools is good, but there's are places I've seen good stuff. I'm sure CMU/Cal/MIT etc have their own gems.
If you want to find terrible trading strategies or research, look no further than a search on SSRN. I can almost never reproduce anything similar to the results of most of the papers on there. I'm sure there's some good content on SSRN, but almost everything I've found on there is garbage.
I have no problem with them using R. Although I don't think backtesting an HFT strategy in R over a lot of data is really viable, but not all their research is HFT related, obviously. I personally do not really trust strategies in academic papers, we probably agree there. I don't really see creating strategies and writing papers about it as the mission of Financial Engineering departments either though.
http://cfe.columbia.edu/research-FE
Stanford also has a lot of really good stuff (particularly on vol arb) some of it even specifically HFT related.
I've found good stuff from NYU too. Not all of the stuff from these schools is good, but there's are places I've seen good stuff. I'm sure CMU/Cal/MIT etc have their own gems.
If you want to find terrible trading strategies or research, look no further than a search on SSRN. I can almost never reproduce anything similar to the results of most of the papers on there. I'm sure there's some good content on SSRN, but almost everything I've found on there is garbage.
I have no problem with them using R. Although I don't think backtesting an HFT strategy in R over a lot of data is really viable, but not all their research is HFT related, obviously. I personally do not really trust strategies in academic papers, we probably agree there. I don't really see creating strategies and writing papers about it as the mission of Financial Engineering departments either though.
I fully concur with you when it comes to research that does not pertain to trading or trading strategies directly. There is a lot of highly interesting content out there with regards to pricing mechanisms for derivatives and the like. Some papers introduce techniques that become market standard (Patrick Hagan on pricing caps, floors, swaptions via SABR, to just name one single example [though I think he actually developed the model as practitioner]). But this thread seems to be directly related to trading and trading strategies and hence my gripe with most academic research performed directly in this particular space.
My comment about R was meant in the sense that it often times has become the lazy man's research tool of choice. What I mean with that is that most in academic space have accepted all the limitations that it poses (extremely poor charting, even the JS script or Ruby stuff is incredibly poor in comparison to modern charting libraries, essentially not even real dynamic charting, which makes visualizing results pretty boring and limited; then the single threaded architecture; then the poor storing and reading of data structures; then the extremely limited speed of execution; the non optimization of loops internally, to just name a few out of the so many shortcomings. One of the largest shortcomings, however, is that most academicians actually learned to use packages that someone else wrote rather than being trained in writing their own code. That leads to limiting themselves to only gain access to the functionality provided in the package. So, you see a lot of time series research performed off the back of R that uses the very same packages to make a point but actually never gets to make a point for lack of out-of-the-box thinking simply because all the thinking is done by the author of the package.
I like to wrap up my point in this thread: Either good research is performed in the trading strategy space but never is published because it turns out to be most profitable to secure the IP. Or, someone in academia is performing such research to attempt to make himself/herself a name, and I find it to be exceptionally rare to come across very high quality work in this aspect. The rest of research is published by otherwise relatively mediocre people who cannot hack it out themselves and accepted the fact that they would fall short of entry into a lucrative position in the financial industry. They try to get tenure (and no, it does not take a brilliant mind to achieve tenure, even at top schools. It takes publishing, at least above average teaching and student feedback, it takes relationships and a good network, and it takes luck for an open position to emerge). Or others sell books. Books, NOT trading strategies that make money.
My comment about R was meant in the sense that it often times has become the lazy man's research tool of choice. What I mean with that is that most in academic space have accepted all the limitations that it poses (extremely poor charting, even the JS script or Ruby stuff is incredibly poor in comparison to modern charting libraries, essentially not even real dynamic charting, which makes visualizing results pretty boring and limited; then the single threaded architecture; then the poor storing and reading of data structures; then the extremely limited speed of execution; the non optimization of loops internally, to just name a few out of the so many shortcomings. One of the largest shortcomings, however, is that most academicians actually learned to use packages that someone else wrote rather than being trained in writing their own code. That leads to limiting themselves to only gain access to the functionality provided in the package. So, you see a lot of time series research performed off the back of R that uses the very same packages to make a point but actually never gets to make a point for lack of out-of-the-box thinking simply because all the thinking is done by the author of the package.
I like to wrap up my point in this thread: Either good research is performed in the trading strategy space but never is published because it turns out to be most profitable to secure the IP. Or, someone in academia is performing such research to attempt to make himself/herself a name, and I find it to be exceptionally rare to come across very high quality work in this aspect. The rest of research is published by otherwise relatively mediocre people who cannot hack it out themselves and accepted the fact that they would fall short of entry into a lucrative position in the financial industry. They try to get tenure (and no, it does not take a brilliant mind to achieve tenure, even at top schools. It takes publishing, at least above average teaching and student feedback, it takes relationships and a good network, and it takes luck for an open position to emerge). Or others sell books. Books, NOT trading strategies that make money.
I'm close enough to fully agreeing with you not to justify a long reply haha. I still think most of my professors are genius, but maybe I lack context there, and I don't want to argue something subjective when we are generally in agreement that trading strategies found in academic papers are either generally terrible or that the good ones are wrapped in NDAs. Also agree that most of the good research is done by people at least heavily involved in industry. Again, I also don't think it's the job of academics to hunt for holy grail strategies either. They should be working on things that are more generally applicable like pricing models or portfolio level methods, so I think it's no surprise that good departments don't publish papers entitled "A Strategy for Trading..."
Pivoting a bit back to the lack of research on HFT, this is a topic I find really frustrating also. HFT is definitely not a lazy man's game, just storing and maintains the data necessary to Backtest strategies that would hardly count as HFT is becoming nearly a full time hobby for me. Also, it's no surprise that the best current HFT practitioners and the OGs of HFT aren't just average academics, they're undeniable geniuses. David E Shaw, Simons, Blair Hull, Malyshev. These are largely top physics and CS people; they wouldn't even have considered writing theses on trading! My point is that HFT is just not something academics seem to care about. It might be too hard, too dependent on specific firms infrastructure to generalize, too far from their areas of expertise; whatever be the case. I've found some decent research on the impact of latency and some informative stuff on order book dynamics, but that's it really. I'd love to share papers later after school here or on PM if you like.
Pivoting a bit back to the lack of research on HFT, this is a topic I find really frustrating also. HFT is definitely not a lazy man's game, just storing and maintains the data necessary to Backtest strategies that would hardly count as HFT is becoming nearly a full time hobby for me. Also, it's no surprise that the best current HFT practitioners and the OGs of HFT aren't just average academics, they're undeniable geniuses. David E Shaw, Simons, Blair Hull, Malyshev. These are largely top physics and CS people; they wouldn't even have considered writing theses on trading! My point is that HFT is just not something academics seem to care about. It might be too hard, too dependent on specific firms infrastructure to generalize, too far from their areas of expertise; whatever be the case. I've found some decent research on the impact of latency and some informative stuff on order book dynamics, but that's it really. I'd love to share papers later after school here or on PM if you like.
Last edited: