Are Renaissance Technologies just trend followers?

[RuleNumberOne]

From the book:

 

"In January 2000, Medallion made 10.5%. By March 2000, the fund was sitting on over $700 million of profits...Then came true trouble...The tech bubble burst on March 10....Medallion lost about $90 million in a single day in March; the next day it was $80 million more.

 

Medallion's losses now approached $300 million....Medallion trades about eight thousand stocks. There was no way they could quickly revamp the portfolio.

 

After several more all-nighters, a couple of researchers developed a theory about what was causing the problems: A once-trusted strategy was bleeding money. It was a rather simple strategy - if certain stocks rallied in previous weeks, Medallion's system had taught itself to buy more of the shares, under the assumption the surge would continue. For several years, this trending signal had worked, as the fund automatically bought Nasdaq shares that were racing still higher. Now the system's algorithms were instructing Medallion to buy more shares, even though a vicious bear market had begun."

[RuleNumberOne]

I think their returns just came from writing software for momentum trading. By automating trend following, they were able to do it more thoroughly than other hedge funds doing it the old-fashioned manual way.

 

I think it also allowed them to change directions very rapidly compared to the manual approach.

 

Which also means their returns must have gone down in recent years compared to the past as other automated trend-following funds came into existence.

 

Looks like they traded EVERY stock in the market. Out of the 8000 stocks they traded in 2000, I think 7000 would have prices moved by a big fund like Medallion/Renaissance alone. When other automated trend following funds entered who were also able to change directions instantaneously, their returns should have gone down.

 

It makes sense that 90% of the trading in the stock market is done by "algos".

[RuleNumberOne]

Greg Zuckerman, the author of the Jim Simons book says the people at Medallion believed the individual investor who did a buy-and-hold was not losing to Medallion, but instead it was other hedge funds that were on the losing end of Medallion trades.

 

Medallion made 98% in 2008. You needed a machine to make 98% in the 2008 rollercoaster. And I think it is the emotional individual investor who loses to the machines.

 

 

 

[muscleman]

From the book:

 

"In January 2000, Medallion made 10.5%. By March 2000, the fund was sitting on over $700 million of profits...Then came true trouble...The tech bubble burst on March 10....Medallion lost about $90 million in a single day in March; the next day it was $80 million more.

 

Medallion's losses now approached $300 million....Medallion trades about eight thousand stocks. There was no way they could quickly revamp the portfolio.

 

After several more all-nighters, a couple of researchers developed a theory about what was causing the problems: A once-trusted strategy was bleeding money. It was a rather simple strategy - if certain stocks rallied in previous weeks, Medallion's system had taught itself to buy more of the shares, under the assumption the surge would continue. For several years, this trending signal had worked, as the fund automatically bought Nasdaq shares that were racing still higher. Now the system's algorithms were instructing Medallion to buy more shares, even though a vicious bear market had begun."

 

Which book are  you referring to?

[RuleNumberOne]

The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution

by Gregory Zuckerman

 

From the book:

 

"In January 2000, Medallion made 10.5%. By March 2000, the fund was sitting on over $700 million of profits...Then came true trouble...The tech bubble burst on March 10....Medallion lost about $90 million in a single day in March; the next day it was $80 million more.

 

Medallion's losses now approached $300 million....Medallion trades about eight thousand stocks. There was no way they could quickly revamp the portfolio.

 

After several more all-nighters, a couple of researchers developed a theory about what was causing the problems: A once-trusted strategy was bleeding money. It was a rather simple strategy - if certain stocks rallied in previous weeks, Medallion's system had taught itself to buy more of the shares, under the assumption the surge would continue. For several years, this trending signal had worked, as the fund automatically bought Nasdaq shares that were racing still higher. Now the system's algorithms were instructing Medallion to buy more shares, even though a vicious bear market had begun."

 

Which book are  you referring to?

[muscleman]

Thank you!

[Guest cherzeca]

need to read new book, but the best performing quants are arb traders not momentum traders.  see ed thrope, a man for all markets

[cameronfen]

need to read new book, but the best performing quants are arb traders not momentum traders.  see ed thrope, a man for all markets

 

I may be wrong but I don't think arb trading scales all that well, which I think all the big names are trend followers, or at least even the big firms that started off arb trading jave branched into trend following. 

[muscleman]

The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution

by Gregory Zuckerman

 

From the book:

 

"In January 2000, Medallion made 10.5%. By March 2000, the fund was sitting on over $700 million of profits...Then came true trouble...The tech bubble burst on March 10....Medallion lost about $90 million in a single day in March; the next day it was $80 million more.

 

Medallion's losses now approached $300 million....Medallion trades about eight thousand stocks. There was no way they could quickly revamp the portfolio.

 

After several more all-nighters, a couple of researchers developed a theory about what was causing the problems: A once-trusted strategy was bleeding money. It was a rather simple strategy - if certain stocks rallied in previous weeks, Medallion's system had taught itself to buy more of the shares, under the assumption the surge would continue. For several years, this trending signal had worked, as the fund automatically bought Nasdaq shares that were racing still higher. Now the system's algorithms were instructing Medallion to buy more shares, even though a vicious bear market had begun."

 

Which book are  you referring to?

 

 

Thank you! I just picked up the book. I'll read it.

I have a feeling that the fact that they started to publicly talk about their strategies is a hint that their strategies stopped working that well, which seems to be indeed the case for their new institutional equity fund.  :)

 

 

 

[Liberty]

Nothing these guys do is "just" something. Otherwise everybody else would do it.

[LC]

Nothing these guys do is "just" something. Otherwise everybody else would do it.

Seriously. These guys are better than Buffett:

 

[Gregmal]

Well, theyre not traditional value investors, so I'd gander many here would say that their returns dont count or something...

[Spekulatius]

It’s interesting to think about who you are up against, if you try short term or daytrading.

[crastogi]

Well, theyre not traditional value investors, so I'd gander many here would say that their returns dont count or something...

 

I would take those returns any day  ;)

[LC]

It’s interesting to think about who you are up against, if you try short term or daytrading.

 

Even longer term. It is publicly known that Rentech uses value/fundamental strategies.

 

Simple 15 minute example:

 

1) Obtain credit ratings for all companies - proxy for "moat"

2) Determine trailing 1, 3, 5 year earnings/cash flows, growth rates

3) Make some exclusion/treatment rules to throw out highly volatile companies or correct for idiosyncratic events

4) Use 1)2) to build a fundamental model +/- 15% of normalized PEs

 

and trade around that.

 

You then combine this with other models in the firm which are estimating heatmaps of the market - i.e. expected cash inflows or outflows or some other metrics. So your  +15% band expands during periods of expected inflows and your -15% band contracts during outflows.

 

Now you've built a model to automate the fundamental investor.

[Liberty]

Nothing these guys do is "just" something. Otherwise everybody else would do it.

Seriously. These guys are better than Buffett:

 

 

That's after charging 44% fees or something, yes. But they also capped their funds at a much smaller amount of capital than Buffett... It's all very different, but in the end, both are making tons of money and at the top of their fields.

[LC]

Very true. Rentech manages something like 100B so quite heavy but not all in Medallion.

 

One of the other fascinating things is the human component.

 

Berkshire is the product of two people. Simons/Rentech took a group of incredibly smart mathematicians, programmers, etc. (and all the neurosis that come with that) and got them to work and function together. That is one hell of a feat if you ask me.

[Munger_Disciple]

I just started reading Zuckerman's book. So far very interesting. There is a table in the appendix showing Medallion fund results & they are stunning. Medallion however is not the only fund Rentech manages. I have the following comments on Rentech performance:

 

1. The first 10 years of Rentech results (1977-1987) which were not so good were ignored in the performance table (this table starts in 1988).

 

2. Medallion fund is an internal fund. Only founders & employees can invest in it and no outsiders are allowed. They use it as an incentive to hire talent. From the book it looks like Simons kicked out investors and even friends out of this fund. So Rentech's best ideas go to this fund. Given that they don't manage outside money in Medallion, I don't understand why Medallion has massive fees. Who is earning these fees?

 

3. In addition to the $10B Medallion fund, Rentech also manages $55B for outside investors and no performance data for these funds is shown in the book. One has to look at the aggregate performance of all assets under management, not just a portion of it to get a better picture of Rentech capabilities. I heard that the results of the outside money are not that good (perhaps someone more knowledgeable can comment on this), certainly not as good as the Medallion results.

 

4. If I recall correctly, Zuckerman said in an interview Medallion is levered 10 times?

 

 

 

[wescobrk]

I'm reading the book and Medallion was managing $5 billion I believe up to 2005 then it went up to $10 billion but Simons doesn't allow anything beyond $10 billion as it can't scale beyond that.

I've heard the returns on all the other funds are average-ish (certainly less than Buffett).

It hasn't been reported Buffett is being judged on a market cap of $500 billion whereas Simmons is managing $10 billion (he distributes the profits every year).

How the hell can you compare $10 billion to $500 billion? Plus, Buffett's is after tax and Simmons is pretax plus all of Simmons gains are short term (the irs ruled that way as he only holds for a few days although Simmons did appeal it).

 

[sleepydragon]

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

[cameronfen]

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.

[johnhuber]

The other thing to consider about Renaissance is the massive tax bill that LP's paid each year. I know much of their investor base became institutions that in some cases don't pay tax, but if you are a tax paying investor, your results were still incredible, but much, much closer to owning BRK. In fact, if you were a high net worth individual living in NYC, I would bet that your after tax returns would have been higher just owning BRK during its prime three decades than owning Renaissance during their run.

[sleepydragon]

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.

[sleepydragon]

The other thing to consider about Renaissance is the massive tax bill that LP's paid each year. I know much of their investor base became institutions that in some cases don't pay tax, but if you are a tax paying investor, your results were still incredible, but much, much closer to owning BRK. In fact, if you were a high net worth individual living in NYC, I would bet that your after tax returns would have been higher just owning BRK during its prime three decades than owning Renaissance during their run.

 

If all the BRK investors invest in the Ren’s internal fund, it will start posting loses immediately. Like WEB said he could do 50% return in a year if he manages small money.

[RuleNumberOne]

Buffett's principles will always remain valid. No doubt he would win a stock-picking contest with any Renaissance employee even today.

 

But Renaissance is like the Google of the investing world. Someone who has read all the newspapers and encyclopedias of the last 100 years can beat any Google employee in a trivia contest. But they can't beat the search engine.

 

The typical hedge-fund portfolio consists of FAAMG, V, MA, PYPL, BKNG. Whereas the Renaissance 13-F looks completely inhuman, hundreds of positions entered and exited without any regard to market cap. If they were made to file a 13-F every week, we could get a better idea of what the Renaissance machine does.