Reply to: Daily Doom

[sasguru]

At least one idiot on this forum has convinced themselves that they can make a living from "technical analysis" of the stock market.
That comes from ignorance of the deep mathematics underlying it - as Doodab and Mich have pointed out there's a difference between taming randomness with statistical methods which allow one to make predictions, and chaotic/turbulent systems which are deterministic but unpredictable.
A salutary read is "A mathematician play the stock market" by John Allen Passos.

[Mich the Tester]

This is not true of the stock market.

thank you

Pseudoscience in the Investment World - SkepticWiki

[doodab]

It depends on hundreds of things of which we have no knowledge; has that person made a profit? Is he trying to cut his losses? Does he want the cash to buy a new TV? Is he just a clueless f***wit?

There are so many dependencies that the behaviour of the system is chaotic; you can never know enough to predict the market's behaviour, because tiny differences in the conditions at any one time can lead to massive fluctuations in the end results.

unpredictable is not the same as random though.

Meaningful predictions can be made about the aggregate outcome of random processes using probabilistic and statistical methods, hence the large body of knowledge related to gambling.

This is not true of the stock market.

[SupremeSpod]

But it isn't random. The probability that someone will sell at a particular price depends on what has happened in the past and what they expect to happen in the future and the tulip that gets printed in the NY Times, FT etc.

FTFY.

[Mich the Tester]

But it isn't random. The probability that someone will sell at a particular price depends on what has happened in the past and what they expect to happen in the future.

It depends on hundreds of things of which we have no knowledge; has that person made a profit? Is he trying to cut his losses? Does he want the cash to buy a new TV? Is he just a clueless f***wit?

There are so many dependencies that the behaviour of the system is chaotic; you can never know enough to predict the market's behaviour, because tiny differences in the conditions at any one time can lead to massive fluctuations in the end results.

[doodab]

But that isn't what happens in a stock market; nothing gets removed; stocks are traded and will change hands as soon as the owner is offered enough money to part with the shares at any particular time; that amount may be more or less than he pad for them.

But it isn't random. The probability that someone will sell at a particular price depends on what has happened in the past and what they expect to happen in the future.

[doodab]

So there he was in 1978, saying there would be a big bull market

When it comes to the stock market, there is always a big market in bull.

The techniques used for predicting stock market movements are educated guessing at best. I suspect that if there was a sure fire method for prediction people would use that instead of inventing ever more complex methods to squeeze a profit from hedging and volatility.

[Mich the Tester]

Monte Carlo falacy is correct when described in terms of roulette - as the roulette wheel has no "memory" of what happened before.

But cards are a bit different, because for each card you pull out - you remove that card from the deck, so the odds do alter. That's why card counting works...

But that isn't what happens in a stock market; nothing gets removed; stocks are traded and will change hands as soon as the owner is offered enough money to part with the shares at any particular time; that amount may be more or less than he pad for them.

[centurian]

Looks like a Monte Carlo fallacy to me. He's saying, basically, that the next five cards will be red because there´s been a long line of red ones.

Monte Carlo falacy is correct when described in terms of roulette - as the roulette wheel has no "memory" of what happened before.

But cards are a bit different, because for each card you pull out - you remove that card from the deck, so the odds do alter. That's why card counting works...

[Mich the Tester]

It suprises me that anyone with the educationa to be expected of IT professionals gives credence to this pseudo-scientific claptrap. It's also a bit disturbing that a paper like the NY Times gives it the oxygen of publicity.

What this man is saying is speculative tosh, derived from selective interpretation of data from the past, incorrect application of fractal mathematics and a habit of counting the hits and ignoring the misses. Apparently, "Mr. Prechter wrote “Elliott Wave Principle,” a 1978 book that predicted the emergence of a great bull market — a forecast that was largely fulfilled. By 1987, he was widely regarded as an expert in technical analysis." So there he was in 1978, saying there would be a big bull market. Either there would not be, in which case, so what, the book was sold, or there would, in which case he'd be seen by eejits as a 'guru'. He had a 50% chance of being right. Getting things right once with a 50% chance doesn't make you an expert and doesn't demonstrate that your theory works. It obviously sells books though.

[doodab]

Looks like people/investors have been piling out of gold in the last week, so where's the money going (apart from down the pan) if not shares? Maybe they are buying Pounds.

Banks are busy refinancing loans made by governments during the financial crisis. I suspect some of it is going there.

[TimberWolf]

Looks like people/investors have been piling out of gold in the last week, so where's the money going (apart from down the pan) if not shares? Maybe they are buying Pounds.

[The_Equalizer]

Yes, and how would that prevent values dropping lower?

It obviously wouldn't, but I doubt the markets would even exist at that level given the upheaval.

[Mich the Tester]

Not quite, I think what he's saying is it's more like you've got this pack of cards, and 26 red ones have been pulled so far, so even if the pack is loaded, then then it's a fairly certain bet there'll be lots of blacks coming up in the future.

yep, correction, because there´s been a long line of black cards. Either way, it´s a fallacy. And there isn´t a ´pack´to be dealt out, just an endless line of random cards.

[doodab]

Bring on the depression - labour will be cheap

You're planning to buy political influence? Tories and Lib dems may well be cheaper by that point.