А вы откладываете на пенсию?

[Oleg]

> Market analysis theories assume that the share price variation at any given period is independent from other periods of observation.

Seasonality and momentum have been well documented. In any case, if the above is true, then it would imply that investing is a lot like gambling. Trying to guess the expectations of others is like trying to guess the next number in the game of ruletka.

[Nikola]

Seasonality and momentum have been well documented.

Forecast models specifically exclude the well-known patterns just because the market by its nature tends to eliminate any predictable behavior as soon as it recognizes a pattern.

[Nikola]

it would imply that investing is a lot like gambling. Trying to guess the expectations of others is like trying to guess the next number in the game of ruletka.

This is true for a day-to-day trade, but not for the long-term trends you mentioned.

[Oleg]

> Forecast models specifically exclude the well-known patterns just because the market by its nature tends to eliminate any predictable behavior as soon as it recognizes a pattern.

I went to scholar.google.com and searched for scientific papers (excluding books) published after 2000. My query was
"technical analysis" stock -book

http://scholar.google.ca/scholar?num=10 ... tnG=Search

It seems that this is a controversial topic. One definitely can't conclude that there is no evidence that technical analysis (defined as the use of past prices to forecast future prices) works. Having said this, I still don't believe in the validity of technical analysis enough to invest in these markets.

[Oleg]

If the following is true

Forecast models specifically exclude the well-known patterns just because the market by its nature tends to eliminate any predictable behavior as soon as it recognizes a pattern.

wouldn't your statement below Not be true? That is, wouldn't long-term trends be eliminated as soon as they become predictable?

This is true for a day-to-day trade, but not for the long-term trends you mentioned.

[Галина]

А у меня знакомый в Торонто, таким образом на пенсию откладывает. Покупает дома и сдает в рент, сейчас у него три дома. Но останавливаться не собирается, но говорит когда пять будет надо тогда с основной работой завязывать. Чем не прибавка к пенсии

[Nikola]

А у меня знакомый в Торонто, таким образом на пенсию откладывает. Покупает дома и сдает в рент, сейчас у него три дома. Но останавливаться не собирается, но говорит когда пять будет надо тогда с основной работой завязывать. Чем не прибавка к пенсии

На каждого если покупать по 3-5 домов, то на всех не хватит. Кроме того, поскольку от домов доход, то с него (вместе с пенсией) - прогрессивный налог, а, значит, пенсию в конце концов как бы отбирают.

[Nikola]

If the following is true

Forecast models specifically exclude the well-known patterns just because the market by its nature tends to eliminate any predictable behavior as soon as it recognizes a pattern.

wouldn't your statement below Not be true? That is, wouldn't long-term trends be eliminated as soon as they become predictable?

This is true for a day-to-day trade, but not for the long-term trends you mentioned.

You are right if we could use both statements together, one after another in order to build a logical chain. But we cannot do that. I even put them in two separate posts, because neither the time frame is far not the same, nor a pattern is synonym for a long-term trend.

I'll try to be more specific. What is a long-term trend? Everyone generally understands that markets tend to grow just because of continuously growing production in decades worldwide. (There is other reasons for that, but here we do not need to qualify them in this context, just ipso facto). That's why the overwhelming majority of pension-related portfolios is composed of market shares, which may yield more, and bonds, and not of just bonds which yield less. The ratio could vary depending on the degree of acceptable risk and shareholders' requirements for dividends.

There is also another assumption that the market information is transparent and is equally available to everyone.

So that when some portfolios just become more profitable than others because of a better long-term forecast related to its structure, their temporary advantage would be quickly eliminated by adjusting the structure of others to a new optimum. This would be a pattern as compared to a trend.There is a very little, if any, place for gambling in the long-term investments which was your point that I was arguing with. The expectations of others investors in this case are fairly transparent to everyone all the times.

Just because it is not the same, long-term trends vs. patterns, it is not possible to use these notions interchangeably in the same logical deduction.

On the other side, gambling is a trivial and the most natural approach to a day-to-day trade because at a fair market nobody can get any sustained gain grâce à the unilateral advantage of using patterns or better forecast from some sort of "secret models" based on the past experience.