Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypotheses

[tech/a]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

Not necessarily.

The motivation for asking could be to demonstrate to others that there is a distinct lack of evidentiary support for deeply entrenched beliefs.

Yes you are right

This IS the case.

 

[sinner]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

Anyway, you are after a system right? A magic formula? One that you can plug numbers in or read the direction and reach El Dorado yea?

Get real man.

Yeah, get real tech, stop trying to scalp us out of our secret formulas, man, why don't you just do some hard work, man.

 

[luutzu]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

Whilst that's not quite how I would define a system, I am otherwise largely agreeable.


I'm not quite sure that I correctly understand the point you're trying to make here.

A formula/system/algorithm/model may lack the self awareness and capacity for independent reasoning, however, its creator and subscribers will have their reasons (irrespective of validity or lack thereof) for their decision to create/operate it.

Couldn't this simply be seen as a rational enhancement of a valid system pursuant to an appreciation of that system's scope and intent?


I'd agree that sometimes when additional layers of complexity evolve, the inner workings can become more deeply obscured.

I'd also agree that many(but certainly not all) do operate with an (at best) limited understanding of those inner workings.

However, the failure of some, to appreciate those inner workings, needn't necessarily invalidate those additional layers.


That may well be true of some people, but I sincerely doubt it could be true of all people.

The two scenarios is to define what I meant as 'system' in the context of my reply to tech/a and your questioning of 'system'.

I agree that any method to do any thing could be define as a system - a way to handicap horses or pushing the pokie have a system to it too.


A system can't be said to be valid when its application have, time and time again, fail miserably. Adding more level of sophistication, in the case of 'modern theory of finance' and DCF, have not work for stock investing.

Take a look at the DCF models I linked to above. Look at how many variables the user must get right. Look at how many models there are and how many variables to hit - the stages and growth rate; the interest rates and costs of capital; the time frame of each stages...

You can say that varying different models, using different WhatIF analysis and scenario mean it's not precise... But it clearly is a mathematical model design to be precise, and precise about the future, the future of a living organism.

If it could be done - predicting the future - it'd be brilliant. History have shown these models has not been able to. And on a theoretical level, when you look closely at the implication of all the sophisticated mechanisms involved, it's designed to obscure and give plausible deniability of nontalent and turn bad management into thoughtful ones more than trying to be scientific and accurate.

In a word, these crazy models are design to create jobs and to scare money out of the public. Even bad performance figures are beneficial to the "experts" - that if they the experts with all the resources and all the crazy models have done poorly, what hope does an average joe have? Would they risk it by doing it themselves?

 

[luutzu]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

Yeah, get real tech, stop trying to scalp us out of our secret formulas, man, why don't you just do some hard work, man.

According to you, I have already given "my" secret formula? No?

Instead of trying to be smart, why not pick a company or two that you got right or got wrong, and show us where you got it right and where you got it wrong.

I bet you would have no idea what goes right and what goes wrong after all that DCF modelling. Which kinda beg the question, why would anyone do something they have no idea and put no faith in it being reliable or not.


Put it another way...

If you use DCF, say a 2 stage model (because 3 stage is insane right?)...
If next year it turns out your estimate is wrong, why was it wrong?

Wrong because WACC was wrong? Wrong due to Earning/CF being higher or lower than you expect? Wrong but still right because it's wrong now but the timeframe is off a bit? All of the above, all equally affecting it?

Oh wait, you wouldn't get to know whether it's wrong or right to do that kind of analysis. You'd just "update" your valuation to "reflect" new changes and new landscape.

That's one way to kick the can down the road. Being paid to forecast the future, does the forecasting, then keep on forecasting and forecasting.


Anyway, I wouldn't get too smug if I don't really know the implication of how my own model works.

 

[sinner]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

According to you, I have already given "my" secret formula? No?

Instead of trying to be smart, why not pick a company or two that you got right or got wrong, and show us where you got it right and where you got it wrong.

I bet you would have no idea what goes right and what goes wrong after all that DCF modelling. Which kinda beg the question, why would anyone do something they have no idea and put no faith in it being reliable or not.

Mate you are a special kind of special. Your crazy assumptions about what people are talking about have completely warped your perception of the discussion going on here.

In this case, you assumed my statement about long term stream of cash flows (which is a fact) means that I use DCF modelling in my investment decisions.

Instead of trying to be smart, how about you leave this thread so it can go back on topic.

:bad::bad::bad::bad::bad:

Anyway, I wouldn't get too smug if I don't really know the implication of how my own model works.

Yes. That is what is going on here, I have no idea about the assumptions, implications and returns of 3 seperate model portfolios which I have constructed through painstaking research and real time trading over 7+ years.

I wish this forum had the option to hide posts from specified users.

 

[luutzu]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

Mate you are a special kind of special. Your crazy assumptions about what people are talking about have completely warped your perception of the discussion going on here.

In this case, you assumed my statement about long term stream of cash flows (which is a fact) means that I use DCF modelling in my investment decisions.

Instead of trying to be smart, how about you leave this thread so it can go back on topic.

:bad::bad::bad::bad::bad:



Yes. That is what is going on here, I have no idea about the assumptions, implications and returns of 3 seperate model portfolios which I have constructed through painstaking research and real time trading over 7+ years.

I wish this forum had the option to hide posts from specified users.

Yup. My mother always said I was special.

7 years? Most non-special people would have figured out they don't know what they're doing a month or two into doing it. Takes talent to keep at it for that long and still can't tell.


Let me be polite and say that if this approach of yours pays for itself for you - as in you use it in investing and no one else is paying you to apply it; that it's so profitable it pays for itself... then great for you. Keep it up.

 

[Trembling Hand]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

I wish this forum had the option to hide posts from specified users.

It does.

 

[luutzu]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

It does.

View attachment 63505

Of course.

We all know great men thinks best when they have around them agreeable men, all saying how great they are.

 

[tech/a]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

So----------Any evidence based papers people would like to recommend?

Interested in comments or any further comments on the veracity of methods used to determine evidence.

 

[sinner]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

If you are actually interested in pursuing this tech, I would advise getting a membership to CXO Advisory. Generating evidence which is robust is difficult and time consuming, so a lot of evidence is not free access.

http://www.cxoadvisory.com/what-investing-approaches-work-best/

 

[tech/a]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

If you are actually interested in pursuing this tech, I would advise getting a membership to CXO Advisory. Generating evidence which is robust is difficult and time consuming, so a lot of evidence is not free access.

http://www.cxoadvisory.com/what-investing-approaches-work-best/

Excellent
Thank you.

 

[CanOz]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

I'm not sure if this belongs here or not Tech/A, but i know you're looking for ideas for systems....I came across this podcast last night, its an interview with the CEO of Systematic Alpha, a NY boutique fund. They use various automated market neutral strategies that take advantage of mis-pricings from liquidity differences, among other things for profit..

I was quite impressed on how much he talked about his strategies, in particular the ES/FTSE/6B spread. This gives you ideas for many more market neutral strategies.

TopTradersUnplugged

CanOz

 

[tech/a]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

I'm not sure if this belongs here or not Tech/A, but i know you're looking for ideas for systems....I came across this podcast last night, its an interview with the CEO of Systematic Alpha, a NY boutique fund. They use various automated market neutral strategies that take advantage of mis-pricings from liquidity differences, among other things for profit..

I was quite impressed on how much he talked about his strategies, in particular the ES/FTSE/6B spread. This gives you ideas for many more market neutral strategies.

TopTradersUnplugged

CanOz

No I'm not.

I'm looking for methods used to give evidence used to support results.
Howard is very clear with his---what is required to verify results.
To prove robustness.

 

[CanOz]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

No I'm not.

I'm looking for methods used to give evidence used to support results.
Howard is very clear with his---what is required to verify results.
To prove robustness.

Well please accept my apologies then, I thought you were also still considering ideas for systems....

 

[tech/a]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

Well please accept my apologies then, I thought you were also still considering ideas for systems....

Sorry that came across as my usual terseness.
I've had a look at the site and another good read.

What I'm looking for is evidence that the results that are achieved positive or negative have been attained
using a method that is not flawed in it self.
I'm sure many tests and in fact many systems people have adopted and traded haven't been evaluated correctly.

Other than the Anecdotal Evidence we have seen here and of course "Appeal To Authority" logical fallacy (sic).
My search continues.

Rubbish in Rubbish out.

 

[galumay]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

My search continues.

Rubbish in Rubbish out.

Do you think its actually possible to have evidence based results that would be meaningful?

It just seems to me that there are some inherent issues in equity markets that make any such confirmation unlikely.

The best that might be expected would be evidence that a given system worked at some point in the past, its logically impossible to then assume that the system would provide the same outcome in the future. Equity markets seem to dynamic with so many complex inputs into their outcomes, this would seem to indicate that only historical assumptions could be drawn from any evidence.

It occours to me that the market itself provides absolute empirical evidence for the outcome of one method of investing, (again only historical with no logical assumption it will hold true in the future.), that is extreme long term buy and hold of the index. The total rate of return has been well in excess of inflation and tax for any longer period of time you choose. Maybe that is a sort of benchmark for what you are looking for?

I have seen plenty of evidence for specific fundamental strategies, but they are in no way up to an academic standard of research and are very specific to particular markets, I dont think that is the rigour you are looking for.

 

[Trembling Hand]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

It occours to me that the market itself provides absolute empirical evidence for the outcome of one method of investing, (again only historical with no logical assumption it will hold true in the future.), that is extreme long term buy and hold of the index. The total rate of return has been well in excess of inflation and tax for any longer period of time you choose. Maybe that is a sort of benchmark for what you are looking for?

You know that Index investing is a very crude trading system. It is basically a weighted trend following system. Rewarding the market cap gainers and culling those that have decreasing market cap.

So if that dumb system "works" so should some other systems. But that still doesn't answer Techs question about robust techniques to test for such success .

 

[sinner]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

You know that Index investing is a very crude trading system. It is basically a weighted trend following system. Rewarding the market cap gainers and culling those that have decreasing market cap.

So if that dumb system "works" so should some other systems. But that still doesn't answer Techs question about robust techniques to test for such success .

It's not necessarily so cut and dried as that.

Cap weighting is passive and scalable, a globally cap weighted portfolio (including credit) is the only portfolio which, theoretically, all market participants could hold simultaneously. With the big assumption, of course, that the market can appropriately price the issuance of shares.

Research and ratings organisations like MSCI and S&P add extra rules (like liquidity filters, or economic representativeness, etc) and it is generally those rules which actually constitute the differences from a raw cap weight benchmark.

 

[Trembling Hand]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

It's not necessarily so cut and dried as that.

Cap weighting is passive and scalable, a globally cap weighted portfolio (including credit) is the only portfolio which, theoretically, all market participants could hold simultaneously. With the big assumption, of course, that the market can appropriately price the issuance of shares.

Research and ratings organisations like MSCI and S&P add extra rules (like liquidity filters, or economic representativeness, etc) and it is generally those rules which actually constitute the differences from a raw cap weight benchmark.

Nevertheless its still a crude system that fundies often say works. If such ideas are good there are more ideas out there that should be better considering how dumb it actually is.

 

[galumay]

Re: Looking for papers on Evidence Based Results for T/A and F/A methods/ideas/hypoth

You know that Index investing is a very crude trading system. It is basically a weighted trend following system. Rewarding the market cap gainers and culling those that have decreasing market cap.

So if that dumb system "works" so should some other systems. But that still doesn't answer Techs question about robust techniques to test for such success .

I guess you could call it a trading system, but I think its gernerally considered as a long term investing strategy.

I am not sure that its logical to assume because one system works, others should. That seems an absolute non-sequitor to me!

It only answers Techs question in the sense that it is an absolutely robust test for the system - 100 years plus of evidential results that can be simply and easily confirmed. The problem is I dont think it throws any light on an evidential methodology for any other system. The evidence for the long term, buy and hold, the index is absolute and complete - but thats all its good for testing. As soon as the system varies then the evidence (index history.) is not specifically relevant.

As much as i hate to invoke the 'B' word, that's why Mr B advocates most investors simply investing in the index and reinvesting the income. The evidence for it's success historically is not open to challenge and it requires no active investment knowledge or skill.

When I think about how most of us invest/trade I cant see evidentially based result reflecting on the systems or strategies being reliable. Who among us does this in a totally automated manner, with no input from our own subjective analysis? I guess some traders apply systems totally automatically, I cant imagine many others would though.