Forums  > General  > what are the latest cutting edge strategies  
     
Page 1 of 2Goto to page: [1], 2 Next
Display using:  

etrader12


Total Posts: 39
Joined: Jan 2010
 
Posted: 2010-01-13 00:42
Hi I'm new here and have lurked for a while, but would like to pick some brains and hopefully shed some neurons.

So statistical arbitrage (particularly pairs trading) was very popular for a while, and then common consensus was that it no longer performed as well as the earlier successes. Just curious what the latest bleeding edge type academic or industry strategies are; any comments?

And regarding HFT, I'm hoping the discussion centers around something that retail traders can exploit. I'm of the opinion that the barriers to entry for HFT are just set against retail traders at the outset for more reasons than one.

goldorak


Total Posts: 561
Joined: Nov 2004
 
Posted: 2010-01-13 08:58
The good old one is still very much in use: pick up random stocks and invent a story about your pick! Chew

If you are not living on the edge you are taking up too much space.

Dynamic Turtle


Total Posts: 165
Joined: Sep 2006
 
Posted: 2010-01-13 10:37

It's currently very fashionable to buy high beta assets. Charge investors 2/20. Take performance fees on the way up. Lock investors in on the way down. Re-set your high watermark, or launch a new fund near the bottom of the market. Take more performance fees as the market rallies. So on and so forth....

DT


goldorak


Total Posts: 561
Joined: Nov 2004
 
Posted: 2010-01-13 16:19
You can do better: two funds, opposite trades, high volatility. And you pump up...

If you are not living on the edge you are taking up too much space.

benji


Total Posts: 186
Joined: Feb 2005
 
Posted: 2010-01-13 16:29
I've had a good overview of Fixed Income strategies with the guides you can find in this thread. As a word of warning you will still need some work of your own to bring the sharpe ratios over 2.

mib


Total Posts: 350
Joined: Aug 2004
 
Posted: 2010-01-13 18:01
Yes, statistical arbitrage is dead. An interesting new development is the so called "unstatistical arbitrage". Basically, one is looking for stable statistical relationships in prices and then trades against them. The "reverse Kolmogorov-Smirnov" has been an especially successful one in the options space and it is often accessible for retail traders.

I may have some documents on the subject that I could share on mutually beneficial terms.

Head of Mortality Management, Capital Structure Demolition LLC

white


Total Posts: 66
Joined: Jul 2007
 
Posted: 2010-01-13 18:17
How about: One looks at published strategies on internet forums and then trades against them?

FDAXHunter
Founding Member

Total Posts: 8251
Joined: Mar 2004
 
Posted: 2010-01-13 18:26
Reverse Kolmogorov-Smirnov is a good one. I have a colleague who's made a fortune on it starting in August 2009. He was/is Smirnov'ing (a.k.a.) smurf'ing EUR/USD options until the cows come home.

There's also a strategy known as a "Martingale System" in government bond futures (again, accessible to retail) that many desks have begun to sink their teeth into. I know that Renaissance Technologies just hired three guys who focus on Martingales alone.

The Chaos Army seems suspiciously well-organized.

Martingale
NP House Mouse

Total Posts: 2592
Joined: Jun 2004
 
Posted: 2010-01-13 19:01
does "martingale system" mean the opposite ? 

Martinghoul


Total Posts: 736
Joined: Oct 2008
 
Posted: 2010-01-13 19:09
Like every investor, I swoon as soon as I hear 'martingale system'.

Insofar as I may be heard by anything, which may or may not care what I say, I ask, if it matters, that you be forgiven for anything you may have done or failed to do which requires forgiveness...

sharpend


Total Posts: 275
Joined: Aug 2007
 
Posted: 2010-01-13 19:23
Le Chicago Martingale is a good one: You hold your hand up in the five year pit until the market reverses.

Graeme


Total Posts: 1629
Joined: Jun 2004
 
Posted: 2010-01-13 19:59
Can somebody explain reverse Kolmogorov-Smirnov? Google has exactly one hit (this thread).

Graeme West

Martinghoul


Total Posts: 736
Joined: Oct 2008
 
Posted: 2010-01-13 20:37
Wouldn't it be necessary to refer to "reverse Kolmogorov-Smirnov" methods as "Smirnov-Kolmogorov"?

I just couldn't hold meself back, sorry Blush.

Insofar as I may be heard by anything, which may or may not care what I say, I ask, if it matters, that you be forgiven for anything you may have done or failed to do which requires forgiveness...

FDAXHunter
Founding Member

Total Posts: 8251
Joined: Mar 2004
 
Posted: 2010-01-13 20:43
Graeme, I've sent you our in-house RKS manual. Check your e-mail.

The Chaos Army seems suspiciously well-organized.

Martingale
NP House Mouse

Total Posts: 2592
Joined: Jun 2004
 
Posted: 2010-01-13 20:47
can i get one too? also the "martingale system" ?

Path Integral


Total Posts: 441
Joined: Dec 2004
 
Posted: 2010-01-13 21:26
Hit me with some RKS please.

One of our sales guys is picking up premium like nothing, i.e. free as in lunch, buying options. HTH.

“When you're as great as I am, it's hard to be humble” Muhammad Ali

etrader12


Total Posts: 39
Joined: Jan 2010
 
Posted: 2010-01-13 21:34
Thanks for all the replies, although, I'm not sure how to separate the sarcasm at the moment.Blush

K-S looks a bit like fisher linear discriminant analysis (separating two distributions via a distance metric) version of a t-test. Not sure how to interpret reverse K-S as a trading tool, but would be interested to hear more details.

I sort of like the un-statistical arbitrage concept, although I wonder how reliable it might be. It would be unlikely to find a mean reverting univariate system (unless it was somehow de-trended first), but given that you had a set of pairs that were co-integrated and scaled to have a stable difference over a window, I'm not sure the opposite bet out of sample would be that much more reliable; they could even begin trending out of sync in which case you are flat, or some other bizarre combo as there isn't only one opposite relationship.

I'm not sure if anyone has heard of unscented particle filters here, but I've been meaning to try to build one, as the reported results (rmse wise) tend to blow away kalman, nns, etc... Very interested to see more results in that area. They are pretty complicated to build, however (at least for me).

hedgeQuant


Total Posts: 231
Joined: Dec 2006
 
Posted: 2010-01-14 01:33
FDAX, can I have a copy of the RKS manual please? I have a review coming up Evil Grin

jungle
Chief Rhythm Officer
CSD LLC
Total Posts: 3162
Joined: Jul 2004
 
Posted: 2010-01-14 02:39
There's code for particle filters in Reza's book.

"When God saw that people, instead of turning to God, were turning to the medicine cabinet, God made himself available in the medicine cabinet."

etrader12


Total Posts: 39
Joined: Jan 2010
 
Posted: 2010-01-14 03:28
Hi,

If you are referring to the volatility arb book (Alireza), I've taken a brief look (via goog preview) and he only covers simple particle filters and possibly extended kalman,
he briefly discussed UPF (173), but I haven't built the C++ version he shows and not certain if it is the same as Van der Merwe's algo as I don't see the same transformations, nor did I take the plunge to get the missing code pages. If you confirm it is identical to Van der Merwe's, I just might take the plunge.

The specific filter I'm referring to has reported performance trouncing most others (although it integrates elements of aforementioned filters with transformations).
See the results from Van der Merwe. I'd love to get a workable algorithm to emulate his filter (there are a few papers as well, in visual tracking applications, but I haven't been able to replicate).
http://cslu.cse.ogi.edu/publications/ps/UPF_CSLU_talk.pdf

jslade


Total Posts: 888
Joined: Feb 2007
 
Posted: 2010-01-14 04:07
etrader12: thanks for posting this. I've been kicking around the idea of using particle filters or UKF for some stuff; I hadn't heard of UPF before. Tantalizing stuff.

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

Aleph


Total Posts: 76
Joined: Jul 2006
 
Posted: 2010-01-14 09:21
Would it be possible for me to get a copy of the RKS document as well?

Thanks!

Cheng


Total Posts: 2619
Joined: Feb 2005
 
Posted: 2010-01-14 11:38
FDax, can you also send me a copy pls ?

"Dead eyes / See no future / Falling from grace / We are coming home"

filthy


Total Posts: 1222
Joined: Jun 2004
 
Posted: 2010-01-14 15:34

etrader12,

i don't think you get the joke. i don't mean the RKS stuff. I mean the relationship between "cutting edge" and profitability.

here's the thing: there are only about 3 strategies that actually work (phuck, there might be 4 or 5 but my basic thesis will still hold). they have always worked. they will always work. the ones i am thinkingg of are: liquidity providing, arbitrage, fundamental analysis.

let's look at the specific example of equity option market making. this involves all three of the broad strategies. you try to collect the spread, you rely on (changeable) stat relationships between various part of the implied vol surface and you need to be aware of a company's fundamentals so you can see if it might pay a special dividend or be taken over or whatever.

market makers have good years and bad years. some are better than others. but as a whole, market makers do better and better each year. but not the same ones. the guy who used to make dollar wide markets in one stock in the cboe pit is long gone. now things are far more automated.

this is where the bleeding edge stuff comes in. at the micro strategy level the exact mechanism for adjusting an implied vol based on a new trade for example.

but the basics of market making have been well known for 25 years. baird wrote his book in 1992 and it wasn't cutting edge then.

so start with a wide angle view. understand the basic source of edge. then maybe you will find a home for your UPF.

people will always pay for liquidity.

arbs (statistical incl) will always exist.

understanding shit will always be worth something


"Game's the same, just got more fierce"

white


Total Posts: 66
Joined: Jul 2007
 
Posted: 2010-01-14 15:53
etrader12, congrats on getting your first serious answer in this thread.

Here is a QA with a retail trader who does automated frequency trading:

http://www.reddit.com/r/IAmA/comments/9s9d7/iama_100_automated_independent_retail_trader_i/

IMHO barriers for retail traders doing high frequency trading are becoming smaller and smaller not greater.
Previous Thread :: Next Thread 
Page 1 of 2Goto to page: [1], 2 Next