Are Stock prices predictable

[naeron]

Bachelier:
Past present and even discounted future events are reflected in the market price, but often show no relevant relation to the price changes. Artificial causes also intervene,the Exchange reacts on itself,and the current fluctuation is a function of not only current and previous fluctuations but also on the current state. These fluctuations depends on infinite factors thus making it impossible to give a precise mathematical signature to them.

Exchange Dynamics will never be an exact science.

The contradictory opinions concerning market changes diverge so much that at the same instant buyers believe in price increase and the sellers believes in price decrease. So on an aggregate the at a given instant of time the aggregate of speculators opinions turns out to a belief in the neither market rise and fall because for each quoted price there are as many buyers and sellers, again at any given instant of time.
hence
THE MATHEMATICAL EXPECTATION OF A SPECULATOR IS ZERO => he/she has equal chance of winning and loosing at each moment in time.

The size of the market fluctuations tend to grow larger as the time horizon stretches out. In a given minute the price change is less than a point. In a day range full points moves are common. As the time frame increases the swings of price changes will range wider. Now the question is how rapid are these changes. "the range is proportional to sq root of time"
---------------------------------------------------------------------------------------------------------------------
Dow Theory is based on the assumption that trends in stock price one under way will tend to persist till the markets will send a signal that these trends will loose the momentum and trend to its reverse. This theory boasts itself to predict the exact time of trend reversals. But users of this theory were clearly divided in finding this exact moment.

Efficient Market hypothesis: based on notion that stock prices reflect all information about companies and markets as a whole. The information is so rapidly reflected in stock prices that no single investor can consistently know more than the market knows as a whole.

some thoughts summary of a book I am reading.
comments welcome...
Nalin ..

 

[kean]

prediction

I don't think anyone can predict the market. Big organizations, financial institutions and big time investors are profited from their internal flow of information. Whoever has more information will have more advantages.

Mathematical modeling is used to structure option deals but it does not mean that the pricing model is valid in the next hours, next few weeks, next few months or next few years. I hope those FE students do not bash me from the back when I make this remark.

I am a supporter of mathematical modeling but I think economic and human factors are uncontrollable. Some people think that working in the front office can make more money (not the salary) via information they collected to profit their own trading. I think it is wrong in some ways that these deals are huge unless one has millions to do it.

I think we can still make money by trading on a few stocks with proper analysis and follow-up over a specific time. I see a need in risk management via legal framework but not via mathematical modeling.

Dow fell nearly 250 points on Friday. Could mathematical models predict prior to Friday? I doubt.

Bachelier:
Past present and even discounted future events are reflected in the market price, but often show no relevant relation to the price changes. Artificial causes also intervene,the Exchange reacts on itself,and the current fluctuation is a function of not only current and previous fluctuations but also on the current state. These fluctuations depends on infinite factors thus making it impossible to give a precise mathematical signature to them.

Exchange Dynamics will never be an exact science.

The contradictory opinions concerning market changes diverge so much that at the same instant buyers believe in price increase and the sellers believes in price decrease. So on an aggregate the at a given instant of time the aggregate of speculators opinions turns out to a belief in the neither market rise and fall because for each quoted price there are as many buyers and sellers, again at any given instant of time.
hence
THE MATHEMATICAL EXPECTATION OF A SPECULATOR IS ZERO => he/she has equal chance of winning and loosing at each moment in time.

The size of the market fluctuations tend to grow larger as the time horizon stretches out. In a given minute the price change is less than a point. In a day range full points moves are common. As the time frame increases the swings of price changes will range wider. Now the question is how rapid are these changes. "the range is proportional to sq root of time"
---------------------------------------------------------------------------------------------------------------------
Dow Theory is based on the assumption that trends in stock price one under way will tend to persist till the markets will send a signal that these trends will loose the momentum and trend to its reverse. This theory boasts itself to predict the exact time of trend reversals. But users of this theory were clearly divided in finding this exact moment.

Efficient Market hypothesis: based on notion that stock prices reflect all information about companies and markets as a whole. The information is so rapidly reflected in stock prices that no single investor can consistently know more than the market knows as a whole.

some thoughts summary of a book I am reading.
comments welcome...
Nalin ..

 

[alain]

some thoughts summary of a book I am reading.

Nalin ..

which book?

 

[naeron]

Today the markets sold off in the morning.. stock markets are like rodeo clowns these days. vol is huge... lets see now what markets will expect of the Fed. ..

 

[GusTsahas]

Regarding the predictability of prices, while one cannot predict what a stock, an index, or a futures contract will do in any one particular day, there are statistical filters that provide above market returns in terms of timing and instrument selection. I am not a believer in the theory that one can only beat the market by using order flow or inside information as we have beat the indexes every year since 2000 by trading systems with a statistical edge.

One book that has a number of easy to understand strategies that one can duplicate in testing is Altucher's book, "Trade Like A Hedge Fund". I don't mean to imply that you will get rich of this book's strategies, but it could convince you to undertake the research necessary to find your own strategies with a trading edge that can make money.

http://www.amazon.com/Trade-Like-Hedge-Fund-Uncorrelated/dp/0471484857/ref=pd_bbs_sr_1/102-4149934-7596914?ie=UTF8&s=books&qid=1189455935&sr=8-1

 

[naeron]

Market itself is a super efficient portfolio. One cannot beat it all the time. The research gives comfort to the decision backed by necessary maths so that one does not miss the obvious information of the numbers. Experience does counts finally.

 

[kean]

Doubt

Statistics provide some ideas how the market reacts in a particular situation - time series.

As you mentioned, one cannot predict the market, how can you beat the market? I don't think anyone can beat the market because market is not "A" player. The market is full of players....it is a zero sum game. Someone is going to lose to someone. Likewise, now some people is betting on rate cut, the opposite side may think that the rate is remain. It is just like playing in casino, you gonna buy big number or small number>>>>you can't bet in between.
Note: Perhaps we can hedge against other financial instruments to reduce risk exposure.

How can you know that inside information is not important otherwise NYSE will not prohibit insider trading?

It is about market sentiment that moves the market. People enter the market at different time frame even we know that this particular stock or index is going down or up at T(t).

I think being a quant, we have to be realistic to consider lots of economic factors when we model the market. Mathematics alone is just not enough.

Cheers,

Regarding the predictability of prices, while one cannot predict what a stock, an index, or a futures contract will do in any one particular day, there are statistical filters that provide above market returns in terms of timing and instrument selection. I am not a believer in the theory that one can only beat the market by using order flow or inside information as we have beat the indexes every year since 2000 by trading systems with a statistical edge.

One book that has a number of easy to understand strategies that one can duplicate in testing is Altucher's book, "Trade Like A Hedge Fund". I don't mean to imply that you will get rich of this book's strategies, but it could convince you to undertake the research necessary to find your own strategies with a trading edge that can make money.

http://www.amazon.com/Trade-Like-Hedge-Fund-Uncorrelated/dp/0471484857/ref=pd_bbs_sr_1/102-4149934-7596914?ie=UTF8&s=books&qid=1189455935&sr=8-1