Forums  > Trading  > Value-Quant research/literature?  
     
Page 2 of 2Goto to page: 1, [2] Prev
Display using:  

rogaine


Total Posts: 67
Joined: Feb 2006
 
Posted: 2006-02-10 15:43
Misfit, thanks very much for the link. In particular, the Kippels article on multivariable factor screens looks very relevant to this thread.

Johnny, my understanding of "industrial" was inline with yours... I think. Confused IMHO the only hope of scaling up from artisanal to industrial lies with 2 familiar factors - (i) good management: e.g. articulate clear research guidelines (ii) good analysts (duh). Otherwise, cue visions of Taleb-trained monkey squadrons on Olivettis.

At the risk of broadening the emphasis from "value-quant" to "equity-quant", I think Jacobs & Levy is an excellent (IMHO) primer to "equity-quant" analysis. Bruner's Global Equity Selection Strategies is interesting too - contains some (oldish) papers from the street, including one from Goldstick's Quant Equity Group.

I haven't read it, but I hear that Grinold & Kahn may be even better than Jacobs & Levy. Any comments?

PS: Anyone know what the returns on the Jacobs & Levy fund(s) is? I can't find them on Bloomberg. I trust it is not a case of Graham-itis (i.e. write great books/papers then fail to cut it in practice)?

Edit: I say "Grinold & Kahn may be even better" in the sense that it may be a better primer to the equity-quant approach, rather than a better ready-to-work-equity-quant model (of course).

Art Vandelay Capital Partners
* Plain Venetian Blind Arbitrage * Latex Swaps *

simonsays


Total Posts: 66
Joined: Jan 2006
 
Posted: 2006-02-10 16:05

I agree philosophically with what you say, but for each of the people that look at the database there is at least one way to model it.

Buffett once said that if the nation were to pair off in a coin flipping contest, with the loser eliminated, and the winner advanced to the next round with double the reward, and you stopped the process when you got to a few hundred contestants, there would be people writing books about their method for winning coin flipping contests.  They would get into arguments with academics saying that "if it was all luck, how come 400 people have been able to be as successful doing it as me?".

That being said, there are many long only and long short funds that "have a system that works", even if those systems change dramatically over time (Third Ave Funds, Manalapan Oracle, and Sabre are examples).

As for reading, try Financial Shennanigans by Schilit.


kr
Founding Member
NP Raider
Total Posts: 3561
Joined: Apr 2004
 
Posted: 2006-02-11 13:16
I'll hit the 'industrial' issue in a moment, but just to clear up:  The framework I offered need not be a 'macro' approach.  In some respects, choosing factors of a factor model - or even the model itself - could be considered an instance of the approach.  I had a conversation about just this with a colleague doing synth CDO - you can imagine how this would work out.  There are too many moving parts in such an instrument, especially relative to dynamic hedging costs, so you can't avoid taking a view on what moves might happen, which ultimately becomes what moves the model allows.  Taking this back to basics, it's not really different from saying the underlying's volatility is constant.  Umm, did we say derivatives hedgers are _investors_?  Well usually not, but the potential is there, and if the desk is making money outside of fees, then apart from fraud one has to admit that risk exposure was taken - somehow - and it worked out well.

This 'industrial' thing is really a great question, and I think 'artisanal' is not a bad way to put it (compare 'Hedgehoggers', now available on swap!)  The one thing that is mechanically different is the control mandate I think.  If you involve a lot of people, you have fiefdoms.  It's like doing a leveraged roll-up:  You buy out all the little stores in order to reduce corporate overhead, and suddenly, earnings at the stores level is dropping everywhere.  Yet as 'Hedgehoggers' points out, pressure not to get involved in a little empire-building is really tough.  If you do succeed, the 'magic formula' that makes your machine work is very hard to keep secret, and barriers to entry for competitors is pretty low.  Need to think more about this circle of ideas, but that's the first scratchy idea.

One more thing that goes for G&D as well as MJW's book is that they are not really textbooks in the way that I suspect the other things mentioned below are.  They are viewpoints, that sometimes work and sometimes don't.  Even TAVFX has its down days - no performance is perfect... it is definitely not science, and one has to be very clear about the subject's aspirations (especially when interpreted by overly quantitative people). 

my bank got pwnd

apine


Total Posts: 1009
Joined: Jun 2004
 
Posted: 2006-02-13 02:51
that's what makes technical analysis so appealing in its various forms (TA, stat arb, etc.).

separately, there is a lot of built in incentives to make the task very not-easy. the reasons have nothing to do with investing but instead because management tends to have advantage with a lack of transparency. even when management is honest, there is a great deal of freedom in accounting without doing anything illegal. plus it is so easy for management that holds large amounts of stock to pretty much do whatever it wants. i'm definitely not an expert but there is a lot of mud that needs to be walked through to get to the other side. i'm still optimistic, though.

i've come to the conclusion that pure quant traders are going to get rather rare. not that quant analysis won't play a part. it will. but it will be in the framework of this discussion: the progression toward an industrial shop rather than a unified theory. we may have pushed things on a pure trading basis as far as things are going for now and need wetware for context.

of course, the best trades are always "customer" driven.

Too many people make decisions based on outcomes rather than process. -- Paul DePodesta

kr
Founding Member
NP Raider
Total Posts: 3561
Joined: Apr 2004
 
Posted: 2006-02-13 21:34

just to make the 'layer cake' analogy, fundamental analysis has been pushed as far as things can go and you have to be an investment banker to know what makes sense...


my bank got pwnd

tabris


Total Posts: 1255
Joined: Feb 2005
 
Posted: 2006-02-14 07:37

Just a random idea, but wouldn't the rewards be better if you look at it as using value/fundamental signals and try to front run all the real money guys and get out/close position instead of using it as a signal for a long/short play? 

If I am not mistaken, these long/short guys using fundamentals are generally long because that is what they are more comfortable doing.  From their perspective, it is easier to find something "cheap" fundamentally to long, than finding something "expensive" fundamentally (and comfortable with) to short. 


Dilbert: Why does it seem as though I am the only honest guy on earth? Dogbert: Your type tends not to reproduce.

simonsays


Total Posts: 66
Joined: Jan 2006
 
Posted: 2006-02-14 17:36

I think that the book hasn't yet been written on statistical methods for fundamental valuation/classification.  There are numerous complications here, and the academic literature is still very active in raising more questions than offering answers.

The question of preference with short and long is an interesing one.  Value investors were pointing out problems with Enron and Microstrategy's fundamentals long before they ever hit their highest share prices.  I view the short side as being more difficult to profit from than the long, and corrections to mispricing on the short side tend to take a lot longer than those on the long side.  This is a conventional piece of Wall St wisdom.

Why is this?  My feeling is that the supply of funds for short positions is so limited that it creates a "positive bias" equilibrium.  Also, stock prices reflect tangible and intangible information.  Earnings momentum very quickly changes tangible info for a security mispriced on the low side.  The adjustment of intangible information is much more unpredictable.


kr
Founding Member
NP Raider
Total Posts: 3561
Joined: Apr 2004
 
Posted: 2006-02-15 00:12
I am getting a lot out of 'Aggressive Conservative' right now.  MJW is extremely anti-statistics - he would say that this is completely missing the point.  It's not about signals so much as fashions.  A certain kind of deal happens to be doable in a certain climate.  The agents' economic alignment is not closely linked to the deal economics.  Huge money is made as deals get done which are strongly against market equilibrium.  It doesn't make much sense but it really is true.  Then the market slowly relaxes.  I tend to agree with MJW that more money is made throwing things out of equilibrium as compared to finding the economic gradient and taking the correct position (which is just applying the maximum amount of statisticomathematical knowledge or whatever).

but re: GOOG or whatever, his 'trader's adage' is also correct upside-down... an expensive stock that remains expensive is - not expensive.

my bank got pwnd

simonsays


Total Posts: 66
Joined: Jan 2006
 
Posted: 2006-02-15 23:25

Reading MJW, he seems to have independently discovered or reinvented economic sociology.  The problem I see is that he doesn't seem to express much curiosity about the origin of the anomalies that he discovers.


misfit


Total Posts: 14
Joined: Feb 2006
 
Posted: 2006-02-17 12:15

how about applying a value/quant model to selecting which market (ie Nikkei, DAX, FTSE etc)- as opposed to individual stocks - to go long/short? Has anyone seen any work related to that?

 


rogaine


Total Posts: 67
Joined: Feb 2006
 
Posted: 2006-02-17 12:30
misfit, I think the NP Italian Stallion (RFMontraz) has done some interesting stuff regarding sector selection models and trading. You might try searching for the old threads.

EDIT: RFM's thread

Art Vandelay Capital Partners
* Plain Venetian Blind Arbitrage * Latex Swaps *

misfit


Total Posts: 14
Joined: Feb 2006
 
Posted: 2006-02-17 16:10

Rogaine thanks for the tip, very useful. But that thread seemed to be more on sector defining, less on sector trading. I was thinking what kind of models are used to trade indexes ie: is it fundamentals (if so what factors), technical, momentum...basically what this thread discusses but applied to markets not individual stocks.

i guess its literally asking the impossible: what market will do best/worst next month?

 

 


simonsays


Total Posts: 66
Joined: Jan 2006
 
Posted: 2006-03-03 02:20
I still get the sense that Aggressive Conservative ignores modern tools, or at least doesn't give them fair appraisal.  Dimensional's funds do about as well as Third Ave, it seems, just by applying Fama/French factors.  LSV probably does about the same or better since they use more factors, although I've never seen performance numbers for them.

tabris


Total Posts: 1255
Joined: Feb 2005
 
Posted: 2006-03-03 02:44
Are there any value added considering their investment philosophy using modern tools?  I am not so sure.  The easiest comparison I can think of with the K.I.S.S approach that has a good track record might be Dodge & Cox.  From what I know, they do nothing fancy and keeps it really simple.  Would their track record start to separate in the future between Dodge & Cox vs LSV?  Maybe, but I still got money in Dodge & Cox and none in LSV Wink

Dilbert: Why does it seem as though I am the only honest guy on earth? Dogbert: Your type tends not to reproduce.

Jurassic


Total Posts: 152
Joined: Mar 2018
 
Posted: 2018-10-05 10:19
@ball_lightning why use ROE and not just the % share price return?

tradeking


Total Posts: 17
Joined: May 2016
 
Posted: 2018-10-05 11:31
@Jurassic, do you realize you are digging up a 12 year old thread? The original posters probably don't browse these forums anymore (or might not even be in the industry anymore...)
Previous Thread :: Next Thread 
Page 2 of 2Goto to page: 1, [2] Prev