Lyxor Msci Em Discussion Threads

[hooya] 03/8/2003 17:20

I would suggest that human psychology is non-random. It might be a broad subject with a vast number of possible reactions to a certain action but I believe a certain action will bring about a predicatable number of reactions.

If human psychology is random, it would be fair to assume price movements are random and therefore TA, which is based on price movements, is just a random diagram and thus will not create an edge as the probability of the price moving in one direction or another is also random. IMO you cant believe in TA creating an edge for entry unless you believe there is some hidden psychology that is non random.

As for support, resisance....I guess becuase they are so well known in trading
you could agrue that they a useless but they way I would view it is never buy approaching resistance and never sell approaching support.Why? because there will always be someone willing to buy at support and sell at resistance. Whether it is enough to turn the market , who knows but if you avoid it you are putting an extra odd in your favour.

The more people or should I say the more money that know about a particular indicator that "works" will eventually have an effect on that indicator. Purely because people will start to get in early or be so confident that the signal is good that they trade bigger money on it. Thus changing price action
( maybe the size of move or the speed of of the price action) thus a new signal
which previously stated getting in at say a crossover on a MACD at 20 will now equire getting in at a greater or lesser extreme. However, does this mean the TA
becomes useless? NOt really. It changes and the trader changes and then the indicator changes and so on and so on. Thus overall effect is minimal over a
long time scale. Think of the most common tool....a trendline....still works atreat and it's been going longer than any of us :-)

[paulismyname] 03/8/2003 13:14

I have to admit I initially viewed TA from a very simple viewpoint, that is if it was going up it would continue, and (in the bull market of the 90s) if it was going down it would reverse at some point.

However we live and learn as they say. I do broadly agree with the concept of support and resistance as well as the simple use of pivot points, and the use of the 50 and 200 day moving average as a lagging indicator.

Devising a more precise TA system that can spot good trades for a day or swing trader (that is intra day to a few weeks) is much harder. I think TA is part of but not the whole story here - I think share newsflow and fundermentals, world and local events as well as the use of level 2 information all have a part to play

[clem] 02/8/2003 20:34

Zxaxx. Its Zen..... :)

[zzaxx99] 02/8/2003 17:32

-- Scrip,

Do what??

[scripophilist] 01/8/2003 17:53

zzaxx99, Our eyes are open when it suits us to see there are many.

[jbm] 01/8/2003 17:18

btw, dave, check out the historical downloads button on the quote page for ISF. Reason data downloads don't list it is because its not an equity.

Jon

[clem] 01/8/2003 16:35

jdeltablues.

So since when has human psychology been non random. :)

[zzaxx99] 01/8/2003 00:47

Ditto to Dil's comment.

In particular, I think the basis you use for dismissing TA is lacking a degree of rigour. Specifically, the "where are all the billionaires" comment - where are all the billionaires from any method? Well, there's Buffett and...? There may well be a stack of others, and there may well be a stack of TA billionaires, but how would anyone know? Unless they were the subject of articles in Forbes or Market Wizards, who would know what methods they use.

Also, you argue that the correlation of day moves compared to previous day moves follow a normal distribution - but so what? If TA was based only on taking individual days in isolation that makes sense, but it's not, so how does that disprove TA. To take a simple TA concept, support/resistance levels may be developed over years rather than days.

You also argued that tactics will be pre-empted out of existance if they work - but, to take the support/resistance example again, the more time it works, the more reliable it becomes. People may trade against it, but the more people who believe a particular level provides resistance, the more people will trade with it and thereby reinforce it.

[clem] 31/7/2003 17:45

zzaxx99. Im not saying there couldnt be technical signals, only that commonly known ones are unlikely to work. Price series are random and sometimes wildly random, the question is not whether they are, but how to profit from that.

Thats noty to say support and resistance wouldnt work for a stock no one is following.

[clem] 31/7/2003 15:53

zzaxx99. Buffett is one of the reasons I concentrate on value investment. :)
Preemption is a classic problem and the more liquid an instrument the more it happens. Liquidity and randomness tend to go hand in hand and preemption is one of the causes.

Sorry to sound Imperious, but in the end if you have a T/A system that can pull 1% out of Vodafone a day, and VOD has a wonderfully wide daily range of 3.75% on average daily, you could make £1.2m from a capital of £10K, in a year. (Using 10x margin in a CFD account and compounding)

.... and that's only 1% out of the range, not the intraday length which I'd guess, off the top of my head is probably 5% or more.

A quantifiable system can always be preempted and will be preempted. If the system requires "art" then I'm extremely skeptical. If support/resistance 'worked' in a way you could profit from it, the support/resistance would be preempted so that the support/resistance would move away from the preemptable level, to a point where it was no longer preemptable for a profit. This is the basis of an efficient market and why trading floors are full of arbitrageurs.

Another way to look at it is: If the system is science, a computer will do it faster then you, get there first in size sufficient to cancel out the marginal opportunity, and you will not profit.

It is only when a T/A system is not universally known or the instrument is neglected that a system can work. This is why successful traders keep their methods secret and/or focus on neglected instruments.

Most T/A however is exposed and therefore obsolete. The answer is to look to new fields and discover your own systems. My 'random walk' piece was an attempt to throw light on a fertile area.

(As a side note I feel both Forbes and Market Wizards are interesting examples of Survivor Bias.)

In any event, the market turns laws into conjectures, so I simply hope you found the piece stimulating. :)

[clem] 31/7/2003 00:59

Thats how I'd do it. I've sent you a spreadsheet of the Dow to show how I've looked at it.