The DOW seems to be exhibiting two heads, and two sets of shoulders when you look at the charts for the last 6 months. As a longtime Guest on this site, I know any discussion of the markets attracts a wave of disdain and ridicule, but hey, for those of you who are interested and brave enough, can we discuss the highly volatile SMs after yesterday's 360 point fall?

It seems to me that the markets are lurching in one direction then the other, but the downward lurches are bigger. Clearly the PPT are out in force, but I don't think it will take much more for this thing to snap and plunge 20%..........

Anyone know the next bank to report? Is it Credit Suisse?

Hasnt happened yet though, has it?

Got a link for the H%S graph?

Can I ask why you chose 6 months as your region of interest?

Not wishing to upset anyone but I always wonder when people go on about technical analysis type things is there actually any logical basis for it?

To some extent it would seem to become self fulfilling if enough people believe it but otherwise I can't really see any rational behind it.

oh, go on!



I agree. The scale at which these artefacts appear seems to be chosen arbitrarily. At any given scale or granularity there will appear to be patterns, doesn't mean they are meaningful. I would be very interested like read a statistically rigorous report that shows results based on predictions made using technical analysis outperfom those made by random chance.

Unfortunately, I doubt any rigorous statistical analysis is possible because, as you say, the artefacts they are looking at don't seem to be rigorously defined.

It's slightly rippled with a flat underside

Today it looks like a wiggly worm

Heh heh..

Although I think it should be possible to construct a null hypothesis along the lines of "The results of technical analysis are better than random chance alone would suggest" and then do a chi-squared test to see if the hypothesis stands up.

It does seem a bit odd that billions of £s are gambled daily on what appears to be little more than financial voodoo. If it is argued that it is a valid technique because many technical analysts are all thinking the same way then at best "technical analysis" is a proxy for herd behaviour dressed up as real analysis. That's really no way to run a financial system!

As I understand it technical analysts run computer software which searches for patterns in historic data. Basically they back fit theories to fit the historical data so that you get absurd trends like 'the DOW always falls for 20 minutes after lunch on a Thursday when there's an 'r' in the month'

It's all guff but it helps the successful traders feel their success is due to something other than pure randomness

The rational behind it is a very simple one which is that a chart simply represents the price of whatever is being charted in the market, and that the price reflects what is known about an asset to the market. It demonstrates clearly what the market (comprised of buyers and sellers) is thinking. If you combine the price movement with volume then you can see pretty much what the market thinks of that asset.

The technical indicators are then mostly mathematical indicators based on price movement/strength/direction over time. So to answer your question yes, technical analysis is entirely logical. It is the study of what is going on in the market between buyers and sellers of an asset. Yes, it is also clearly the case that since participants in the market are also using the same or similar technical charts then using the charts they are using should help you if you take an interest in them.

If you don't, then I would ask how you would know by looking at any chart/graph etc what has happened in the past and what might happen in the future? How would you know house prices had gone up? or Gold looked expensive? or people don't like Northern Rock? It isn't a guarantee of anything but it is a useful tool in identifying trends and market psychology.

if you could predict the stock market as well as the people on this board...

I've always wondered that - it seems like b*llox to me really. All these patterns probably work with hindsight, but how accurate are they at predicting things? And as you say, if everybody use these techniques, they may work until something different happens or some 'eedjit' doesn't use the charts and screws things up.

I think traders just try and pretend the market is vaguely scientific, whereas infact it's an orderly and chaotic system (at the same time) responding to participants greed and fear (human emotion).

Regards,
crude

I was beginning to think it looked like an ECG trace. If so, the patient just flatlined

No, I have no problem with the mathematical indicators based on well defined things.

It's when people start going on about "resistance levels", "head and shoulders", "flag poles", and start drawing seemingly arbitrary lines on graphs that I start to get a little suspicious.

I could be wrong of course and misunderstood something.

I'm no expert in the stock market as I'm an engineer, but a while ago I was thinking that it would be cool if you could take a set of historical data and perform a fourier transform on the data set to determine a set of frequencies that made up the oscillations of a stock or a market. I thought this might be new so I was quite excited, then did a search on the web and apparently they use fourier transforms for financial frequency analysis already so bang went my idea of thinking I'd discovered something new...

Anyway, I think there must be some unpredictable and undeterminable oscillation that can be exploited, especially on blue chip stocks which should be less volatile. I often think that a stock market could be modelled on an automated control system and thought about trying something but I don't have the time or charting knowledge. Even in a stable control system there is always slight oscillation around the set point due to factors which are often very difficult to pinpoint yet are predictable (in their frequency and thus controllable), I think the same must be true in the financial systems.

I also don't think it's something that many people can easily do, other than those dedicated to its application. It must be heavily resource intensive and surely requires large funds to exploit the small changes.

Yes, it does seem interesting.

My understanding is that they model it using a model based on Brownian motion to price derivatives, etc. I don't really know much about it.

Still no graph, I see.....

Says it all really.