Market Data Questions

Trend Indicator calculation

Dear Tom:

I am a current subscriber. I would like to have more information on a few indicators you are using as I do not understand them yet very well:

1) On todays edition[11/09/07]: “sum/10” : this is a moving average using a Summation Index of Price Oscillator: do you mean it adds each day value of the Price Oscillator for 10 days and then does an average of it? If that is the case, what are the values you use for the Price Oscillator. I am only familiar with a Price Oscillator that is the difference of 2 moving averages. What are the values of that Price Oscillator you are using ?

2) On todays edition[11/09/07]: The “Trend Indicator” would then be the difference between the Closing price of the Index (SP500 for instance) with that “Sum/10” and then the “5.8%T” would be an exponential moving average of the “Trend Indicator” with a length of 5.8 - is that correct ?

3) The “10% Trend” and “5% Trend” you use on some charts is an exponential moving average (length 10 and 5 respectively) of the Price Oscillator. Is that correct?

I would really appreciate these clarifications. Thanks!

You pose some good questions about the Trend Indicator. It is a pretty darned good indicator in my experience, although the math involved makes its use almost prohibitive for some people.

First, calculate the Price Oscillator for the index you are interested in. We publish the data for several indices in our big table, and you already know that the Price Oscillator is the 10%T minus 5%T.

Next, calculate a Summation Index of Price Oscillator values. Start with an initial value at the beginning of your data equal to 10 times the initial closing price in your data string. Then add each day’s Price Oscillator value to that Summation Index.

Divide the Summation Index by the number 10 to get the Sum/10 level. Dividing it by 10 puts it back down into the range of current prices or index levels. That Sum/10 line is the moving average line you see in some of our charts. It is a nice slow moving average, and some price series seem to have a more magical affinity for it than they do for other moving averages, hence our interest in it.

As we have discussed, the Trend Indicator is found as the distance between the closing price (or price index level) and the Sum/10 line. We don’t use a raw numerical difference, but instead normalize the calculation to take away the amplitude changes resulting from higher or lower price levels. Use the following formula:

 Close - Sum/10
————————- x constant
       Sum/10

If you leave off the “constant” term, you will get a very tiny decimal number, so we multiply the result by a constant to put it back up into the realm of real numbers. That step also makes the chart scaling much happier. The choice of the constant does not really matter, you can use whatever you want, or omit it.

Obviously, this is the sort of math best done in a spreadsheet program. I would not know how to tell you how to program TS or Metastock, since I don’t play with those platforms.

The 5.8% Trend is an exponential moving average (EMA) that is calculated on the Trend Indicator values. It has a 5.8% (0.058) smoothing constant, meaning that in the calculation, today’s Trend Indicator value gets a 5.8% weighting, and yesterday’s EMA value gets a 94.2% weighting.