Jump to content

Are Renaissance Technologies just trend followers?


RuleNumberOne

Recommended Posts

There are many different ways to make money in the market. But it’s really hard to scale with size. Ren is doing very high frequency trading, resulted in very high sharpe but size can’t be scaled up to more than a certain percentage of a stock ADV

 

Yes but the idea is you trade every single stock, every commodity, every currency pair and derivative contracts.  After all a neural network that can follow one trend has an easier time of generalizing to other assets.  With that you can manage a lot of assets.

 

Making money through stats arb doesn’t mean they use neural network or even any sophisticated AI. The neural network is actually a pretty new thing. Any alpha/ideas employed by hedge fund human traders can be automated by Stats arb. But it’s not easy to apply most ideas in equity to other assets.

 

AFAIK rentech is not stat arb typically (although at this point they probably do a bit of everything).  I think stat arb is not just programming hedge fund ideas.  It is but is grounded in mathematical or at least statistically valid arbitrage opportunities ie actual situations where the same asset has different prices.  Rentech is mainly a trend following quant shop, which means they use techniques to identify then follow trends.  While you don’t have to use cutting edge machine learning to identify those trends, most of the big quant shops have teams that apply machine learning and deep learning to these problems.  I know for sure that Rentech hires machine learning phds.  I’m not saying generalizing a statistical model is easy, but it’s much easier if you have one model the follows trends to extend that to other assets and even asset classes.  While they might pick up fundamentals, to some extent when you are trading based on trend a lot of that is probably picking up subtle behavioral reactions to price movement and that generalized across any asset class. 

Link to comment
Share on other sites

  • Replies 61
  • Created
  • Last Reply

Top Posters In This Topic

Good points cameronfen. Another item to add is conviction - machines have a lot more compared to humans. Six of Renaissance's top 10 positions right now are drug stocks.

 

A "fundamentals-based" hedge-fund builds its portfolio around the hedge-fund favorites - FAAMG, V, MA. The only stock in that list that appears in the Renaissance top 10 is FB at #10.

 

The Renaissance people are betting their own money on models that they themselves built. Who would think of betting their money on drug stocks in an election year with Liz and Bernie tweeting threats all the time.

 

Even when "fundamentals-based" investors jump into drug stocks, they come out with a Valeant as 10-30% of their portfolio. Hard to say "fundamentals-based" can in any way be superior to Renaissance's machines.

 

There are many different ways to make money in the market. But it’s really hard to scale with size. Ren is doing very high frequency trading, resulted in very high sharpe but size can’t be scaled up to more than a certain percentage of a stock ADV

 

Yes but the idea is you trade every single stock, every commodity, every currency pair and derivative contracts.  After all a neural network that can follow one trend has an easier time of generalizing to other assets.  With that you can manage a lot of assets.

 

Making money through stats arb doesn’t mean they use neural network or even any sophisticated AI. The neural network is actually a pretty new thing. Any alpha/ideas employed by hedge fund human traders can be automated by Stats arb. But it’s not easy to apply most ideas in equity to other assets.

 

AFAIK rentech is not stat arb typically (although at this point they probably do a bit of everything).  I think stat arb is not just programming hedge fund ideas.  It is but is grounded in mathematical or at least statistically valid arbitrage opportunities ie actual situations where the same asset has different prices.  Rentech is mainly a trend following quant shop, which means they use techniques to identify then follow trends.  While you don’t have to use cutting edge machine learning to identify those trends, most of the big quant shops have teams that apply machine learning and deep learning to these problems.  I know for sure that Rentech hires machine learning phds.  I’m not saying generalizing a statistical model is easy, but it’s much easier if you have one model the follows trends to extend that to other assets and even asset classes.  While they might pick up fundamentals, to some extent when you are trading based on trend a lot of that is probably picking up subtle behavioral reactions to price movement and that generalized across any asset class.

Link to comment
Share on other sites

Owning drug stocks with Liz and Bernie hovering is very painful. Bernie caused a lot of pain to drug stock investors in 2016.

 

Humans don't want that pain. Machines don't feel the pain and Renaissance is content to let the machine do its thing.

 

When it came to Valeant, it was not just Bill Ackman. The list of Valeant luminaries included Lou Simpson (12% of portfolio), Sequoia Fund (> 30% of portfolio), Glenn Greenberg (36% of portfolio), Wally Weitz. They probably derived conviction from each other.

 

 

Most fundamental investors (myself included) have zero industry specific knowledge in healthcare r&d. The ones that do hire specialists (baupost), and their portfolio looks much different than bill ackman and valeant.

Link to comment
Share on other sites

Owning drug stocks with Liz and Bernie hovering is very painful. Bernie caused a lot of pain to drug stock investors in 2016.

 

Humans don't want that pain. Machines don't feel the pain and Renaissance is content to let the machine do its thing.

 

When it came to Valeant, it was not just Bill Ackman. The list of Valeant luminaries included Lou Simpson (12% of portfolio), Sequoia Fund (> 30% of portfolio), Glenn Greenberg (36% of portfolio), Wally Weitz. They probably derived conviction from each other.

 

Valeant was easy to avoid - lots of red flags: high debt levels, rollup with exponential growth, unconventional business model, promotional “ Outsider” management, heavy promotion of non-GAAP accounting.

 

People got greedy and were neglecting the downside because so much money was made with this stock. If you just focus on downside, you never would have gotten into Valeant to begin with.

 

 

Most fundamental investors (myself included) have zero industry specific knowledge in healthcare r&d. The ones that do hire specialists (baupost), and their portfolio looks much different than bill ackman and valeant.

Link to comment
Share on other sites

The new book is great, and it's illuminating.  Sad to see how much errors there are in this thread about RenTech. 

 

1) Medallion hasn't been publicly available for a LONG TIME.  Their institutional funds are and have very public performance data.  However, the institutional fund performance #'s are nothing to write home about.

 

2) Medallion is limited to around $10B, I think.  They disburse all gains above and beyond to stakeholders.  Even at this size, their performance #'s are unreal.

 

3) Buffett doesn't even come close to Medallion #'s.  He wishes he could.  Even if he were trading with a small AUM, I HIGHLY doubt young Buffett would come close to Medallion Fund #'s

 

4) Medallion isn't really stat arb.  Not even close.  They did do some stat arb before, but not with equities.  They were pretty good in commodities and debt investments.  But, what really ramped up their #'s since late 90's was pattern recognition software.  They hired two guys from IBM from their speech recognition division who finally was able to get Medallion a significant edge in equities.  As a side note, one of the guys is a far right conservative who likely helped Trump win the election. 

 

5) They use leverage but not LTCM levels of leverage.  They can scale in and out of leverage fairly easily. 

 

6) Jim Simons gets way too much credit for the success of Medallion.  His main contribution was probably the idea of using quantitative models for trading and team gathering. 

 

7) They are definitely not fundamental investor types.  At least not since the early 90's.

 

Link to comment
Share on other sites

Edit: the links don't work. Need to Google for author-name followed by "dblp"

 

Robert Mercer's papers are here: https://dblp.uni-trier.de/pers/hd/m/Mercer:Robert_L=

 

David Magerman's papers are here: https://dblp.org/pers/hd/m/Magerman:David_M=

 

No doubt these ideas were used at Medallion

 

"Jelinek wrote, "The performance of the Renaissance fund is legendary, but I have no idea whether any methods we pioneered at IBM have ever been used. My former colleagues will not tell me: theirs is a very hush-hush operation!"

 

If they didn't use any of it, they would have told their lead co-author Jelinek.

 

Link to comment
Share on other sites

I know very little about neural networks, but the speech recognition research that happened in IBM in the 1990s was antiquated long ago. But suppose we start with what kind of queries can neural networks answer that would be relevant to stock trading?

 

Fill-in-the-blanks.

 

Given a sequence of winners and losers for training the model, the neural network could predict missing winners or losers. E.g. given the set of today's winners, which stocks are missing from the list... 5000 stock tickers per day over 10 years would be a total of 12.5 million words for training.

 

They would still need to add fundamental data as input to justify why a stock flips from reliable winner to loser and vice versa.

 

If multiple stocks are missing from today's stock market winners, there are likely no outliers, as opposed to a single stock missing from the list of winners...

Link to comment
Share on other sites

I think they also use some sentiment analysis of analyst reports or news articles for drug companies. Drug approvals in the case of NVO and VRTX, merger in the case of BMY and CELG.

 

The appearance of CMG could be explained by detecting an outlier in the sequence of winners.

 

I think the publication of this book will result in a lot of competitors. They have shown it works and there are a lot of smart people out there who can replicate these results - who never tried before because they had no idea such techniques would work.

Link to comment
Share on other sites

Edit: the links don't work. Need to Google for author-name followed by "dblp"

 

Robert Mercer's papers are here: https://dblp.uni-trier.de/pers/hd/m/Mercer:Robert_L=

 

David Magerman's papers are here: https://dblp.org/pers/hd/m/Magerman:David_M=

 

No doubt these ideas were used at Medallion

 

"Jelinek wrote, "The performance of the Renaissance fund is legendary, but I have no idea whether any methods we pioneered at IBM have ever been used. My former colleagues will not tell me: theirs is a very hush-hush operation!"

 

If they didn't use any of it, they would have told their lead co-author Jelinek.

 

His research at least as of 2019 is predominately neural network based.  Transformers, LSTMs with attention that like what is hot now and what was hot 3 years ago in language models with neural networks.  He has some other stuff but I am sure his work is mainly Neural Network based.  I bet by now there is no way Rentech are not using transformers to forecast these time series based on his expertise and what works.  It is likely he ran a giant transformer with like maybe 1billion+ parameters on the time series of every single financial asset and with maybe slight modification is running the same transformer to predict movement in those given assets.  If I had to guess you can copy the structure of the largest transformer, megaton-ln, from here: https://blog.exxactcorp.com/megatron-lm-unleashed-nvidias-transformer-megatraon-lm-is-the-nlp-model-ever-trained/ , steal the transformer base architecture from here: https://github.com/tensorflow/tensor2tensor and with little coding knowledge but 2 or 3 million dollars to spend on AWS you can likely replicate 50% of there returns just by training it on every possible asset class time series.  My guess is that’s the core of the model and all the ml smarts in the world gets only somewhat marginal improvements from there. 

Link to comment
Share on other sites

Robert L. Mercer (the Renaissance guy) last published a research paper in 1995.

 

Robert E. Mercer is a professor somewhere and he is still publishing research papers.

 

Yeah, it is very easy now to run the latest algorithms on any amount of data compared to the 1990s. Lot of machine learning packages available.

 

But the key is what inputs to feed - how do you structure the inputs.

 

Edit: the links don't work. Need to Google for author-name followed by "dblp"

 

Robert Mercer's papers are here: https://dblp.uni-trier.de/pers/hd/m/Mercer:Robert_L=

 

David Magerman's papers are here: https://dblp.org/pers/hd/m/Magerman:David_M=

 

No doubt these ideas were used at Medallion

 

"Jelinek wrote, "The performance of the Renaissance fund is legendary, but I have no idea whether any methods we pioneered at IBM have ever been used. My former colleagues will not tell me: theirs is a very hush-hush operation!"

 

If they didn't use any of it, they would have told their lead co-author Jelinek.

 

His research at least as of 2019 is predominately neural network based.  Transformers, LSTMs with attention that like what is hot now and what was hot 3 years ago in language models with neural networks.  He has some other stuff but I am sure his work is mainly Neural Network based.  I bet by now there is no way Rentech are not using transformers to forecast these time series based on his expertise and what works.  It is likely he ran a giant transformer with like maybe 1billion+ parameters on the time series of every single financial asset and with maybe slight modification is running the same transformer to predict movement in those given assets.  If I had to guess you can copy the structure of the largest transformer, megaton-ln, from here: https://blog.exxactcorp.com/megatron-lm-unleashed-nvidias-transformer-megatraon-lm-is-the-nlp-model-ever-trained/ , steal the transformer base architecture from here: https://github.com/tensorflow/tensor2tensor and with little coding knowledge but 2 or 3 million dollars to spend on AWS you can likely replicate 50% of there returns just by training it on every possible asset class time series.  My guess is that’s the core of the model and all the ml smarts in the world gets only somewhat marginal improvements from there.

Link to comment
Share on other sites

Robert L. Mercer (the Renaissance guy) last published a research paper in 1995.

 

Robert E. Mercer is a professor somewhere and he is still publishing research papers.

 

Yeah, it is very easy now to run the latest algorithms on any amount of data compared to the 1990s. Lot of machine learning packages available.

 

But the key is what inputs to feed - how do you structure the inputs.

 

Edit: the links don't work. Need to Google for author-name followed by "dblp"

 

Robert Mercer's papers are here: https://dblp.uni-trier.de/pers/hd/m/Mercer:Robert_L=

 

David Magerman's papers are here: https://dblp.org/pers/hd/m/Magerman:David_M=

 

No doubt these ideas were used at Medallion

 

"Jelinek wrote, "The performance of the Renaissance fund is legendary, but I have no idea whether any methods we pioneered at IBM have ever been used. My former colleagues will not tell me: theirs is a very hush-hush operation!"

 

If they didn't use any of it, they would have told their lead co-author Jelinek.

 

His research at least as of 2019 is predominately neural network based.  Transformers, LSTMs with attention that like what is hot now and what was hot 3 years ago in language models with neural networks.  He has some other stuff but I am sure his work is mainly Neural Network based.  I bet by now there is no way Rentech are not using transformers to forecast these time series based on his expertise and what works.  It is likely he ran a giant transformer with like maybe 1billion+ parameters on the time series of every single financial asset and with maybe slight modification is running the same transformer to predict movement in those given assets.  If I had to guess you can copy the structure of the largest transformer, megaton-ln, from here: https://blog.exxactcorp.com/megatron-lm-unleashed-nvidias-transformer-megatraon-lm-is-the-nlp-model-ever-trained/ , steal the transformer base architecture from here: https://github.com/tensorflow/tensor2tensor and with little coding knowledge but 2 or 3 million dollars to spend on AWS you can likely replicate 50% of there returns just by training it on every possible asset class time series.  My guess is that’s the core of the model and all the ml smarts in the world gets only somewhat marginal improvements from there.

 

Oh interesting.  Strange both are in NLP.  Also maybe should have checked because I was surprised a 70 year old is both recently a CEO as well as publishing papers as well as running a Koch style political funding arm.  Either way before I think 90% prob base model is Transformer.  Now probably 80%.  None of his stuff has anything to do with neural networks from a cursory look, which is unsurprising as ANN only got hot 2015.  Still, I’d be surprised if anything could outperform a core transformer model (megatron-ln) trained on basically every asset time series. 

Link to comment
Share on other sites

Until I read the book I thought quant traders didn't make money. It was an eye-opener. But it seems there have been a lot of people copying Renaissance over the last few years. Shouldn't the advantages that quants have get competed away?

 

WorldQuant manages money for or is owned by Millennium management (the same firm that two Renaissance people defected to as described in the book).

 

Citadel and Two Sigma are hiring computer science grads from universities and creating a campus environment.

 

https://www.bloomberg.com/news/articles/2017-03-06/citadel-joins-two-sigma-chasing-quants-in-campus-recruiting-push

 

"Two Sigma will take over “The Bridge,” a space on the new Cornell Tech campus on Roosevelt Island in New York, where engineers and entrepreneurs will work.

 

The hedge fund staff will collaborate with Cornell students and professors on machine learning and data science projects, creating a pipeline of academic talent to the firm. Job candidates will put on virtual-reality glasses to watch a video that explains how the hedge fund sees the world awash in data.

 

The students -- handpicked from 400 applicants -- are competing in a datathon hosted by the $26 billion hedge fund Citadel. Ken Griffin’s firm is upping the ante in the industry’s chase for data scientists and engineers, hosting 18 competitions at universities across the U.S., Britain and Ireland this year. The prize in the final data championship: $100,000.

 

Igor Tulchinsky, the founder of WorldQuant, is pitching a perk that breaks the tradition of hedge-fund secrecy. The $5 billion firm is hiring at least 15 teams of quant managers for its Accelerator platform, offering them the right to keep their intellectual property. The quant hedge fund has also opened more than 20 offices in 15 countries, including emerging markets like Russia and Romania, to find engineers. Talent is distributed around the globe, Tulchinsky said at the Milken conference, “but opportunity is not.”"

Link to comment
Share on other sites

  • 5 months later...
  • 1 month later...

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

So their "employee only" fund does 60% a year...but their "rest of the people" fund, does this...

 

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

The Medallion fund doesn't do equities I believe and has been deliberately kept small to keep the high returns going. Their marketed funds have not been as successful due to size and different asset classes that they're involved in.

Link to comment
Share on other sites

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

So their "employee only" fund does 60% a year...but their "rest of the people" fund, does this...

 

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

The Medallion fund doesn't do equities I believe and has been deliberately kept small to keep the high returns going. Their marketed funds have not been as successful due to size and different asset classes that they're involved in.

 

Not true, Medallion does whatever Medallion wants to do.  I'm sure most of it is equities.  And "small" is $5b+ year in the fund returning 40-50% year after year, if not greater.  Yes, they return all the excess capital annually, but that's still a lot of capital for those kind of returns.  Amazing.

Link to comment
Share on other sites

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

So their "employee only" fund does 60% a year...but their "rest of the people" fund, does this...

 

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

The Medallion fund doesn't do equities I believe and has been deliberately kept small to keep the high returns going. Their marketed funds have not been as successful due to size and different asset classes that they're involved in.

 

Not true, Medallion does whatever Medallion wants to do.  I'm sure most of it is equities.  And "small" is $5b+ year in the fund returning 40-50% year after year, if not greater.  Yes, they return all the excess capital annually, but that's still a lot of capital for those kind of returns.  Amazing.

 

I recall from reading the new book written on Jim Simon that Medallion's bread and butter was not equities; in fact, they had historically struggled to build profitable models on equities. A reason why Medallion pays out so much is that they've found it comparably difficult to reliably generate returns by trading new asset classes.

Link to comment
Share on other sites

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

So their "employee only" fund does 60% a year...but their "rest of the people" fund, does this...

 

https://www.ft.com/content/6bd17811-3205-454e-89e4-953dce6b4dfe

 

The Medallion fund doesn't do equities I believe and has been deliberately kept small to keep the high returns going. Their marketed funds have not been as successful due to size and different asset classes that they're involved in.

 

Not true, Medallion does whatever Medallion wants to do.  I'm sure most of it is equities.  And "small" is $5b+ year in the fund returning 40-50% year after year, if not greater.  Yes, they return all the excess capital annually, but that's still a lot of capital for those kind of returns.  Amazing.

 

I recall from reading the new book written on Jim Simon that Medallion's bread and butter was not equities; in fact, they had historically struggled to build profitable models on equities. A reason why Medallion pays out so much is that they've found it comparably difficult to reliably generate returns by trading new asset classes.

 

According the book, they struggled in equities for a long time and then achieved a breakthrough, which allowed them to scale up to their current size of $10B.

 

I m sure in the current mayhem in the stock markets, they make out like bandits this year.

Link to comment
Share on other sites

I tried looking for an article on Medallion fund but haven't found it. I think it came out last year in case anyone knows what I am talking about they can post it. From my memory, the article explained how the Medallion fund operates (vaguely as much of it is proprietary) much like a casino, placing millions of small trades every day on both sides. That way market moves are somewhat irrelevant and they take a daily vig. Investments never last longer than 2 weeks.

Link to comment
Share on other sites

  • 7 months later...

Renaissance Hit With $5 Billion in Redemptions Since Dec. 1

 

Public hedge funds at Simons’s firm posted worst-ever losses

Unprecedented market swings tripped up quant managers in 2020

 

 

https://www.bloomberg.com/news/articles/2021-02-07/renaissance-hit-with-5-billion-in-redemptions-since-dec-1?srnd=premium

 

Billionaire Jim Simons’s firm, a quant-investing pioneer, is coming off a rough year. Its three public hedge funds posted double-digit losses in 2020 as their algorithms were thrown out of whack by market swings the computers had never seen before. At the same time, its fund for employees and insiders soared 76% last year, Institutional Investor reported.

 

 

Renaissance’s Institutional Equities fund, the biggest of the external vehicles, lost 19% in 2020, the letters show. That fund got the biggest chunk of the redemptions. The Institutional Diversified Alpha fund dropped 32% and the Institutional Diversified Global Equities fund fell 31%.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...