--------------------------------------------------------------------------------

------------------------------------------------

Quantum mechanics has nothing on trends! In both cases you can be in different states at the same time! You have to decide where you wish to be.

To establish a relationship between the trends of different markets, you have to determine the trend of the individual markets. You can't say how trends compare if you can't identify a trend in the first place. Is it possible to devise a quantitative model to define the trend?

To build such a model, you need to establish the ground on which the model is to be based. Any model provides a detailed structure of the organization being quantified and so delineates the class or classes of information being generated about the organization. We need to specify our objective and say how the aims are to be reached. Indeed, we wish to be long in equities going up, or short in stocks going down.

But what exactly do we mean when we say the price of an equity is going up? The answer may seem obvious, at first, but it isn't apparent when a decision based on the answer has to be made to buy or sell the equity.

Equity prices don't change in a continuous way. They trade in quantum jumps, according to price agreements reached by buyers and sellers, i.e., the strike price. To follow the ideas of quantum mechanics, the price of a stock "jumps" from an indefinite state to a specific state.

Where trading in a particular equity is sequential, the strike price may be higher, lower, or the same as the price reached in the prior trade. The strike price is reached by the specialist by matching his book orders. (The change has been described as a random walk. But it seems more to be controlled by the specialists.) In any case, random or not, we know that prices generally vary from trade to trade and in fact tend to oscillate in value.

The odd fact is that prices tend to oscillate in value no matter what interval is chosen to measure the change. This is the fractal nature of prices and makes it difficult to say whether a price is trending up, down, or staying the same. In fact, the price may be moving up, down, and sideways at the same time, depending on your reference.

To see this, consider that one might take measurements of price change from trade to trade, from one hour to the next, from one day to the next, in weekly intervals, or over any other interval, and still see variations in the price. What's more, the changes in price in the different intervals will usually be different, sometimes being positive, sometimes negative, and sometimes zero. The question whether the price of an equity is moving up may therefore be in doubt, since it may be going "up" in one interval, but going "down" in another, both at the same time.

To decide whether a stock is moving up, or not, we need to establish some criterion for the change. We need a reference to make the decision. This reference amounts to a trader's choice. It depends on the trader's choice of a trading time zone.

Back to Index

---------------------------------------------

While it may be true that the markets aren't for everybody, it's probably equally true that not all markets are for you. It's your choice. And temperament has an important role to play in the choice. Some people prefer a relatively quiet, peaceful life, without major interruptions, and thus they prefer to deal in less volatile issues. They like to be able to sleep at night.

Others, on the other hand, would rather live their days and nights on the edge and so fully enjoy trading the wild market oscillations -- for big profits, sometimes, and big losses at other times. These folks do not like to watch paint dry.

Fact is, a number of long and short intervals are deemed important enough to have names. You can find them here.

Back to Index

---------------------------------------------

An approach that mechanizes your preference is to use a moving average to eliminate the noise (the minor swings). The correct moving average interval smoothes out the price variations at the noise level, below your adopted level, yet it retains the oscillations that are at or longer than your adopted interval. It does this by taking the average of the stock prices during the elected time span and using that average as a proxy for the interval of prices.

Say that you disregard variations in price occurring in intervals of 10 days or less. By using a 10-day average of closing prices, you add the prices for the 10 days and divide the sum by 10. If the successive closing prices were:

3, 4, 3, 2, 3, 4, 3, 4, 5, 4

the average would be:

(3 + 4 + 3 + 2 + 3 + 4 + 3 + 4 + 5 + 4)/10 = 3.5,

and your proxy for the stock price for the 10 day period would be 3.5.

You can now obtain a graph of averages -- or what is called a moving average -- by taking the average values of successive, overlapping chunks of ten days of closing prices. Each day you average the closing prices of the last ten days, including the current day, and plot the result. So each day you would include the current day's price and delete the 11th day's price.

There is, however, a price to pay for the advantage of having a smoothed stream of data. That cost is having to depend on old information. Since you have to wait for the 10-day's worth of data to evolve, the average itself is old.  This can be mitigated somewhat by using an exponential moving average, which favors new data, though doing this has its own problems, in that the average is now more susceptible to news items that have nothing to do with prices.

Back to Index

------------------------------------

Top of Page