TRIX indicator calculation looks wrong

[trapezoid]

Hi NT,

Your calculation of the TRIX looks suspicious. You have:

EMA tripleEma = EMA(EMA(EMA(Inputs[0], period), period), period);
double trix = 100 * ((tripleEma[0] - tripleEma[1]) / tripleEma[0]);
Default.Set(trix);

Whereas I think it should be:

EMA tripleEma = EMA(EMA(EMA(Inputs[0], period), period), period);
double trix = 100 * ((tripleEma[0] - tripleEma[1]) / tripleEma[1]);
Default.Set(trix);

In other words, you are normalising by the current value of the triple exponential, rather than the previous.

Here are three references for you to check

http://support2.dundas.com/OnlineDocumentation/WebChart2005/TRIX.html

http://www.incrediblecharts.com/indicators/trix_indicator.php

http://daytrading.about.com/od/indicators/a/Trix.htm

These are just the first ones that google gave me. I don't know if there is an 'official' reference. I was just looking at the code and it didnt make sense. I didn't see any websites calculating it the way you do.

cheers

[NinjaTrader_Brett]

Hello,

Thanks for posting.

http://fmlabs.com/reference/default.htm

We use this reference for most of our indicators. There isn't any one accepted value here as what is correct. In these cases we always default to FMLabs or stocks and commodities magazine where FMLabs got their forumula for the version to use inside of NinjaTrader.

Let me know if I can be of further assistance.