Forums  > Trading  > You Know Who - Renaissance Watch  
     
Page 1 of 40Goto to page: 1, 2, 3 ... 39, 40 Next
Display using:  

HitmanH


Total Posts: 488
Joined: Apr 2005
 
Posted: 2019-09-04 23:58
With you Rod - interesting approach to joining a new community / forum...

And also - MM deriving those profits for a return on a fund that size.

If only... Or bring on the return of 2008...

Jurassic


Total Posts: 277
Joined: Mar 2018
 
Posted: 2019-09-05 08:50
@HitmanH how can you tell that it cant just be MMing? why cant it derive those profits on a fund that size?

Maggette


Total Posts: 1167
Joined: Jun 2007
 
Posted: 2019-09-05 11:21
That is because HitmanH actually has a rough idea what he is talking about!:)

There might be some MMs elements in their portfolio of strategies....but did you ever ask yourself the question why all the players that openly do NOTHING ELSE THAN MM (like XTX, Optiver, All Options, large parts of Jane Street) are not hedge funds but prop companies?

Because these strategies are highly profitable but don't scale well....hence the size of any of Rentechs funds is far to big to have an majority of work done in MM! As an MM you don't want a lot of investors to share your profit with.

Ich kam hierher und sah dich und deine Leute lächeln, und sagte mir: Maggette, scheiss auf den small talk, lass lieber deine Fäuste sprechen...

gaj


Total Posts: 54
Joined: Apr 2018
 
Posted: 2019-09-05 18:16
If you combine all those MMs, would the combined AUM still be much smaller than Rentec's? My rough guess is no -- I've heard of some HFTs generating 9 to 10 figure profits. So maybe Rentec is just as smart as all the other MMs combined?

kkkk


Total Posts: 13
Joined: Sep 2019
 
Posted: 2019-09-22 01:57
I think it is.

let's take into account that rentec on American stock produces about 150k daily trades. (is an inaccurate figure because we do not know their execution algorithms)

In my opinion there are various types of mm.

Rentec is a slow mm that assumes large positions.

kkkk


Total Posts: 13
Joined: Sep 2019
 
Posted: 2019-09-22 02:28
@Maggette cited companies that have a precise role in the financial industry in fact many companies turn out to be mm members of the stock exchanges.

xtx= execution for customers in fx as virtues for stocks
optiver= mm for European and American options with credit from banks.
all option= as optiver
jane=constructor and mm etf.

We also see giants such as two sigma that make mm in options...

Rentec does not offer any service. They make mm on stock, futures, fx for their cabbages. their models are fine. certainly will not be as simple as making market makers in options.

HitmanH


Total Posts: 488
Joined: Apr 2005
 
Posted: 2019-09-24 02:38
Thank you Maggette ;-)
(I'd hope that c20+ years in the Hedge Fund industry, and almost half in the quant / prop / MM space might give me some clue - but you never know ;-))

i) Gaj - yes it would be a fraction of RenTech. Can argue semantics over which vehicle - but either way - yes, just doesn't add up.
ii) kkkk - there is a nuance to what you say which you may or may not intend, which is that, while maybe not a mm in a traditional / exchange sense, that it is feasible that RenTech provides liquidity into other strategies (that it believes are under some duress). This may not be MM as anyone would define it traditionally (this could be over multiple days for example), but there is some logic in it

Maggette


Total Posts: 1167
Joined: Jun 2007
 
Posted: 2019-09-25 10:34
I totally agree with ii).

The distinction if a strategy is MM or not can IMHO become a bit blurry.

Ich kam hierher und sah dich und deine Leute lächeln, und sagte mir: Maggette, scheiss auf den small talk, lass lieber deine Fäuste sprechen...

Maggette


Total Posts: 1167
Joined: Jun 2007
 
Posted: 2019-09-25 10:34
I totally agree with ii).

The distinction if a strategy is MM or not can IMHO become a bit blurry.

Ich kam hierher und sah dich und deine Leute lächeln, und sagte mir: Maggette, scheiss auf den small talk, lass lieber deine Fäuste sprechen...

Jurassic


Total Posts: 277
Joined: Mar 2018
 
Posted: 2019-09-25 11:08
"The distinction if a strategy is MM or not can IMHO become a bit blurry."

@Maggette in the sense it becomes stat arb instead?

contango_and_cash


Total Posts: 114
Joined: Sep 2015
 
Posted: 2019-10-30 16:57
fyi:
https://www.amazon.com/exec/obidos/ASIN/073521798X/thebigpictu09-20

TSWP


Total Posts: 416
Joined: May 2012
 
Posted: 2019-10-30 17:57
Thanks for the link.

I bet it will probably contain material already discussed in here... Wink

longGamma


Total Posts: 16
Joined: Jan 2019
 
Posted: 2019-11-09 20:42
Entertaining and informative public account of Renaissance's history. Man Who Solved the Market.

Medallion fund averaged ~66% pa, or 39% net of fees, over the last 30 years. Fees are 5/44 since 2002.

The returns on $10 bn AUM is astonishing.

To note,
- fund size was capped at $10 bn
- 1994, 2000, 2007, 2008 were positive outliers
- alpha consists of short term "effects". One example was offering liquidity to MMs who derisk before major economic releases.

- Jim is a character. He once walked into a conference room where sales was closing an investment into RIEF. The investor, a foundation who's mission is to "build a culture of health", was ready to sign a check and mark the occasion with a cake. Simons lit up, filled the room with smoke, and extinguished his cigarette into the icing of the cake.
"As the cake sizzled, Simons walked out, the mouths of his guests agape. The Renaissance salesmen were crestfallen, convinced their lucrative sale had been squandered. The foundation's executives recovered their poise, eagerly signing a big check. It was going to take more than choking on cigarette smoke and a ruined vanilla cake to keep them from the new fund" Applause



ironyman


Total Posts: 3
Joined: Nov 2019
 
Posted: 2019-11-20 04:22
What Rentec has done is build an excellent data processing engine that automatically separates signal from noise. Other more secretive funds have done this too.

kuebiko


Total Posts: 27
Joined: May 2018
 
Posted: 2019-11-20 14:58
Can you name any of these more secretive funds? Do they have anything like the AUM of medallion?

gmetric_Flow


Total Posts: 28
Joined: Oct 2016
 
Posted: 2019-11-20 15:22
@ironyman Sounds rather generic- everyone tries to separate the signal from the noise. Since it seems you're in the know, can you say something more specific?

gaj


Total Posts: 54
Joined: Apr 2018
 
Posted: 2019-11-20 15:41
I'm more than halfway through the book. There's a few meaty stuff that I haven't heard before.

- In their early attempts in trading stocks, the live performance wasn't nearly as good as backtest. It turned out there was a bug where they had used a static value for S&P500.

- When the tech bubble burst, they decided to override the model after experiencing significant losses. Some signals kept suggesting to buy Nasdaq shares, so they had to manually reduced reliance on these signals.

- The book also shares some explicit signals that Rentech used in the early years. E.g., "Monday's price action often followed Friday's, for example, while Tuesday saw reversions to earlier trends".

ironyman


Total Posts: 3
Joined: Nov 2019
 
Posted: 2019-11-21 09:30
@gmetric_Flow Medallion's research platform is key. All firms operating in this space recognize that data is integral to their success and most of them are drowning in the data they're trying to process. Rentec doesn't. A large number of their research team works directly on innovating data processing. They have hired/consulted very specific people who can provide such innovation.

If I recall correctly, David Donaho (https://www.msri.org/people/492) was co-author on many of the papers for the LIGO detection technique ie. separation of gravitational wave signals from a huge amount of noise.

The actual details are proprietary but it's mostly signal processing, information geometry and information theory. You can read the papers rentec authors publish before they left whichever institutions they used to work at (try "@rentec.com" on Google/Google Scholar for starters).

nikol


Total Posts: 851
Joined: Jun 2005
 
Posted: 2019-11-21 12:41
@ironyman

Awesome.

jslade


Total Posts: 1195
Joined: Feb 2007
 
Posted: 2019-11-21 13:58
> it's mostly signal processing, information geometry and information theory

Kind of figured the former and the latter; surprised at information geometry. Didn't turn up any such thing in an @rentec.com google scholar search. Can you point towards something relevant?

"Learning, n. The kind of ignorance distinguishing the studious."

etrader12


Total Posts: 40
Joined: Jan 2010
 
Posted: 2019-11-21 22:06
Simple google search returns the original feed that ironyman has partially referenced from here:
https://news.ycombinator.com/item?id=16649002
Not knocking that, maybe he is the original poster.


jslade


Total Posts: 1195
Joined: Feb 2007
 
Posted: 2019-11-22 09:17
That's worth talking about a little bit, in that I now know enough alumni to know that the statement "Renaissance Technologies has completely automated the process of signal discovery. They don't hire researchers to manually derive novel insights or trading models from data, and they don't really bother with exclusive sources of data. Instead, they hire researchers to improve methods for automatically processing vast amounts of arbitrary data and extracting profitable trading signals from it." is either ... an overstatement, or a misunderstanding of how dimensionality reduction and the research process works. The author was involved in selling alternative data sources. Rentech turned him down. They turned me down too, but I only draw the conclusion this happened because they already had one of what I was selling.

If you look at his other contributions to HN, some of his statements are ... questionable.

Anyway, whether or not the wizards of Setauket use information geometry, it's an interesting basket of ideas, closely related to some stuff I've been doing less formally, and I'm glad it was brought up.

"Learning, n. The kind of ignorance distinguishing the studious."
Previous Thread :: Next Thread 
Page 1 of 40Goto to page: 1, 2, 3 ... 39, 40 Next