How would you test for trendiness?

Consider screening for high volaltilty on a monthly chart for trendy stocks. The higher the volatiltiy on a monthly chart the more trendy the lower time frames become, etc daily. They have to "trend" in order to catch up with the monthly.


John
 
Quote from cnms2:

Comparing the two volatilities adjusted with sqrt(time) gives you an indication if the trend is accelerating or slowing: shorter term volatility higher than longer term volatility divided by sqrt(time) indicates that the trend is accelerating.
More...

what if shorter term volatility acclerated because the price is falling in that time frame, but longer term time frame says prices are rising? in this case the measure would say the trend is acclerating, but which way?

let me ask in another way. people always say that the SP market doesn't trend well and is the least trendy. does anyone have any statistical evidence of this? if yes, what measure did you use?
 
Quote from cnms2:

To compare two markets' trendiness you have to compare their velocity of making money: $/time. But to decide which one to trade you have to look at the other factors that define your trading method, including an estimate of their reward / risk ratio.
More...

lets block out the other factors for now (risk/reward, correlation, etc.) lets say all you want to do right now is rank 10 markets from the most trendiest to the least trendiest. how would you do this? what statistical measure would you use?
 
I would select the market that moves the most $ in the unit of time. Tharp's Efficiency Index could be a measure of it.

I probably don't understand what you're looking for. Depending on how you plan to trade it, how you set your stop loss and your target, these may interfere with support / resistance levels, channel boundaries, averages you track, TA indicators you watch, pending news, etc..
Quote from jerryz:

lets block out the other factors for now (risk/reward, correlation, etc.) lets say all you want to do right now is rank 10 markets from the most trendiest to the least trendiest. how would you do this? what statistical measure would you use?
More...
 
Quote from jerryz:

lets block out the other factors for now (risk/reward, correlation, etc.) lets say all you want to do right now is rank 10 markets from the most trendiest to the least trendiest. how would you do this? what statistical measure would you use?
More...

First eliminate all bear markets. Bear markets can unfold in a myriad of patterns, thus they trend for a while and then go sideways.

Second identify all bull markets. Bull markets can only trend one way UP. When they correct its usually either a quick downtrend, or a few shorter term downtrends.

Next evaluate the historical volatility, last 10 years, of each bull market. And where it is in relationship to the mean price of that time span. Markets that are near historic highs have virtually no overhead resistance, i.e. gold 1970's, stock indices 1980 - 90s, bond market 2000+, crude the last couple of years, Google and the Transports now!

With no historical overhead resistance they can trend stronger and higher with large percentage gains in short periods of time.

It's not quantative analysis, it's real life experience.
Good luck!
 
Quote from jerryz:


let me ask in another way. people always say that the SP market doesn't trend well and is the least trendy. does anyone have any statistical evidence of this? if yes, what measure did you use?
More...

Using the simple way I described above, after a quick calculation and assuming I haven’t done a calc mistake :)

Volatility using daily data 0.011942 (annualized 18.96%)
Volatility Using Weekly data 0.024547 (annualized 17.70%)

The market qualifies as mean reverting.
 

Attachments

What keeps you from using your own "intuition" to figure it out?

What does the term "volatility" mean to you?

Markets can be quiet with little movement or volatile with significant swings in both directions. Most markets display both kinds of price action.

How much more "intuitive" do you need the explanation to be?

Steve
 
cbot uses 250 vs 252 for annualizing the dailies

from http://www.cbot.com/cbot/pub/page/0,3181,774,00.html
Annualized Historical Volatility Formula
annualized historical volatility = square root of ( 250 * variance of natural log differences in daily prices for the calendar month)


Quote from gbos:

Using the simple way I described above, after a quick calculation and assuming I haven’t done a calc mistake :)

Volatility using daily data 0.011942 (annualized 18.96%)
Volatility Using Weekly data 0.024547 (annualized 17.70%)

The market qualifies as mean reverting.
More...
 
Quote from jerryz:

gbos, can you please give an intuitive explanation of this method?
More...

The intuition is simple. Imagine an ant that is moving with a speed of 2 meters per minute.
If the ant is moving towards one direction (trending) you expect it to be at a distance of say 1.5 meters after one minute.
If the ant is randomly wondering around (normal market) you expect it to be at a distance of say ¾ of a meter after one minute.
If it is moving back and forth (mean reverting market) you expect it to be at a distance half a meter after one minute.

Daily volatility measures the ant speed.
Weekly measures how far on average you find it from his previous possition.
 
You must log in or register to reply here.
Back
Top