Professional quant here. I have to say I strongly disagree with the conclusions of the OP.
> They were all found by using phrases like "predict stock market" or "predict forex" or "predict bitcoin" and terms related to those.
Yeah, searching for any finance papers with "predict" or "machine learning" is literally the lowest quality tier you can get. These papers are often written by grad students who can pump an easy paper out by "applying" some already known ML algorithm to financial markets. Of course it's not gonna work. It also kills me when I see ML models who need stationarity assumptions applied to non-stationary time series data. Yeah, good luck with that.
THAT being said, there is lots of high quality research which has been replicated over and over, showing that alpha does exist in the market (and which funds have made billions off of). I would like to see the OP try to replicate some of these instead. To give some simple examples:
1. Try searching for papers with the keywords "and the cross section of expected returns". For example, the momentum factor which can be tested and replicated with only linear regression.
> There is substantial evidence that indicates that stocks that perform the best (worst) over a three- to 12-month period tend to continue to perform well (poorly) over the subsequent three to 12 months.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=299107
2. Statistical arbitrage strategies which were known to work well until the mid 2000s. Also been replicated many times, furthermore, you can see the gradual decline in profitability pointing to the theory that "alpha decay" in this case is real. https://www.math.nyu.edu/faculty/avellane/AvellanedaLeeStatA...
3. High frequency strategies. No way OP or any retail trader can replicate this, but firms make billions of dollars per year consistently doing this.
In conclusion, to make a claim that there is no alpha in the market seems highly suspect, and perhaps just needs a more nuanced view of how trading firms make their profits.
I also find it highly unlikely anyone is able to implement 130+ papers in 7 months.
This would require insane productivity, implausible access to pricing and news data resources (which are often not freely available) and expertise in machine learning, natural language processing, finance, and data science.
OP had to implement financial, time-series and linguistic feature engineering pipelines, as well infer the architecture and hyper-parameters used AND train all these models.
He also claims he "web scraped" all the data which is highly unlikely as pricing datasets are often sold for a pretty penny and not publicly available in the detail described in several of these papers.
OP must be a genius to pull this off, all the while being a trader at "a Tier 1 US bank" (in itself that description is ridiculous).
All OP has to show for all this work is a hastily written Reddit post with dubious claims. There is no proof of the work done whatsoever, no code samples, not even result tables or graphs.
And at the end OP chills his cryptotrading bot.
What's worse HN seems to gobble it up naively.
Seemingly because OP is critical of something that is popular to criticize.
> OP must be a genius to pull this off, all the while being a trader at "a Tier 1 US bank" (in itself that description is ridiculous).
Mostly agree with you, but what's ridiculous about that description? His LinkedIn says he worked at Merrills and Citi. Those are normally considered top tier US banks?
Generally the banks are no longer the place to do prop trading though. You're better off at a hedge fund. They will not only pay better but have way better access to resources, tech, experienced traders, etc
A lot of papers use data from a few sources which are typically available to universities. Also a surprising (maybe not?) scrape data themselves. That being said as someone who was involved in paper replication and investigation there is not enough time to implement substantive papers in that time frame.
Perhaps he lead an R&D team at his bank, and most of this work was being done in parallel by his reports.
Or perhaps there are fewer ideas contained in these papers than there are papers themselves. I can “replicate 130 papers”, too, if they’re all substantially the same experiment on different data-sets. Just download all the data-sets and loop through them :)
Hey, at no point did I made a claim that there's no alpha in the market. I generated around 25% annually myself on a fairly large balance sheet and made reference to rentech's stellar performance several times. AND I'm working on a new commercial project which relies on finding alpha. but yes the rest of your comments are valid and fair.
Also, for the other comments I edited the post to add links because I was asked several times what I was up to now.
This is a really informative comment, but the OP explicitly excluded the three strategies you mentioned. (There’s a blurb about ignoring “alpha” strategies, and he only went back 8 years).
I read this article more as answering the question “Did anything useful come out of the last few batches of finance PhDs?”, than “Are investment strategies are totally futile?”
Most academic white papers on trading strategies are horrible. There are written by academics or grad students (as you point out) that have no firm understanding of the actual market mechanics. You can often tell this right away just by the language and terminology used... Even before you get to the actual math or strategy.
Nobody will make any money in the markets relying on others work. It's just how arbitrage works. I've you want to succeed you need to be creative beyond what's already been done.
Once you're there you should follow the first rule of trading:
Don't talk about your strategy.
The second rule of trading is the same as the first rule.
I cannot emphasize enough how much you need to keep things secret in trading.
Actually, I think the author is in agreement with you that there is alpha. He apparently wrote this article to attract attention to his scheme for selling access to his cryptocurrency trading "bots", which he claims are guaranteed to be highly profitable once he finishes developing them. Certainly, his survey is not credible and does not appear any more genuine than his business offering.
Awesome work, have been looking for something like this for a while. I would love to see tags for people and a search bar. If I'm trying to organize a casual chess match or pickup soccer game, it would be super convenient to search all my friends with said interest.
> Because the "points" have no intrinsic value except in comparison to your classmates
This is where your logic is wrong. The professor is not saying the class is graded on a relative scale (for example, top 10% of the class gets As). If everyone gets two points, and it bumps up everyone's GPAs, then of course it will benefit them all, especially in comparison to people who are not in the class.
This is not necessarily true. If you're already a good student, this devalues your GPA, mostly due to the generally awful GPA system. Assuming your school considers everything above a 92 to be a "4.0", if you're already in that range, you definitely don't want to help anyone else into that range.
Note that, of course, this would be fixed by just making GPAs decimal numbers out of 100, the average of all of your actual, numeric grades from each course (multiplied by credit hours?). You can divide it by 25 if you really want the weird 4.0 scale for some reason.
Yes, it's true that it sort of benefits everyone in the class vs people who did not take the class, but some classes are easy A's and some aren't, and everyone knows that. If the class becomes a slightly easier A because we all picked 2 points, then the effect will be relatively small. It would be a better experiment if he was handing out cash, even small amounts of cash.
I run https://rrandomize.com (it ships random items to you). It's currently earning about $150 a month or so. Most of the money comes in through one-time payments, since I pocket the change after purchasing an item for the user. The idea is simple, so it didn't take much work coding at all. I'm still pretty new to web development so this was a chance to practice execution and security.
That's awesome, I was about to ask if it was inspired by the xkcd before reading your about page. How many people currently have the "adventurer" or "conquistador" plan? Can you provide examples of items you've bought in the past? Actually, you should probably put that info on your homepage.
Around 50 people have those plans right now. Some recent items that have been shipped are: a solar powered cockroach, bacon candy canes, and a pedometer. I was planning on building a live stream of items being ordered, but I'm waiting on more users to sign up before implementing that feature.
This is an interesting concept. I suppose you could do similar things with USD as well. You could declare, "Anyone with a dollar bill with this specific serial number now has control of my company." I would argue that the dollar bill still has a value of $1 by itself, but you can assign arbitrary value to it by making it representative of another thing.
> They were all found by using phrases like "predict stock market" or "predict forex" or "predict bitcoin" and terms related to those.
Yeah, searching for any finance papers with "predict" or "machine learning" is literally the lowest quality tier you can get. These papers are often written by grad students who can pump an easy paper out by "applying" some already known ML algorithm to financial markets. Of course it's not gonna work. It also kills me when I see ML models who need stationarity assumptions applied to non-stationary time series data. Yeah, good luck with that.
THAT being said, there is lots of high quality research which has been replicated over and over, showing that alpha does exist in the market (and which funds have made billions off of). I would like to see the OP try to replicate some of these instead. To give some simple examples:
1. Try searching for papers with the keywords "and the cross section of expected returns". For example, the momentum factor which can be tested and replicated with only linear regression. > There is substantial evidence that indicates that stocks that perform the best (worst) over a three- to 12-month period tend to continue to perform well (poorly) over the subsequent three to 12 months. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=299107
2. Statistical arbitrage strategies which were known to work well until the mid 2000s. Also been replicated many times, furthermore, you can see the gradual decline in profitability pointing to the theory that "alpha decay" in this case is real. https://www.math.nyu.edu/faculty/avellane/AvellanedaLeeStatA...
3. High frequency strategies. No way OP or any retail trader can replicate this, but firms make billions of dollars per year consistently doing this.
In conclusion, to make a claim that there is no alpha in the market seems highly suspect, and perhaps just needs a more nuanced view of how trading firms make their profits.