<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Lies, damn lies, and statistics</title>
	<atom:link href="http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/</link>
	<description>Colin McGinley&#039;s journey of forex trading by a thousand cuts</description>
	<lastBuildDate>Tue, 24 Jan 2012 06:42:32 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=</generator>
	<item>
		<title>By: sh77</title>
		<link>http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/comment-page-1/#comment-1078</link>
		<dc:creator>sh77</dc:creator>
		<pubDate>Mon, 13 Dec 2010 12:50:26 +0000</pubDate>
		<guid isPermaLink="false">http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/#comment-1078</guid>
		<description>I&#039;m having trouble figuring out the robust measurements from way of the turtle.  I was wondering if you could post a template with the Robust measurements that you used from your excel file on your blog.

Thanks</description>
		<content:encoded><![CDATA[<p>I&#8217;m having trouble figuring out the robust measurements from way of the turtle.  I was wondering if you could post a template with the Robust measurements that you used from your excel file on your blog.</p>
<p>Thanks</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Colin McGinley</title>
		<link>http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/comment-page-1/#comment-1003</link>
		<dc:creator>Colin McGinley</dc:creator>
		<pubDate>Tue, 02 Dec 2008 21:48:24 +0000</pubDate>
		<guid isPermaLink="false">http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/#comment-1003</guid>
		<description>Hi JD,

Sorry for the delay in replying!

You seem to have figured it out yourself.  Having a quick look at the spreadsheet I setup to calculate RAR% and RRRR I see that I used the INTERCEPT and SLOPE functions.

The RAR% value is then fed straight into the R-Cubed calculation.</description>
		<content:encoded><![CDATA[<p>Hi JD,</p>
<p>Sorry for the delay in replying!</p>
<p>You seem to have figured it out yourself.  Having a quick look at the spreadsheet I setup to calculate RAR% and RRRR I see that I used the INTERCEPT and SLOPE functions.</p>
<p>The RAR% value is then fed straight into the R-Cubed calculation.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jdfagan</title>
		<link>http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/comment-page-1/#comment-1002</link>
		<dc:creator>jdfagan</dc:creator>
		<pubDate>Mon, 24 Nov 2008 06:39:53 +0000</pubDate>
		<guid isPermaLink="false">http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/#comment-1002</guid>
		<description>Ok...I think I just realized I must use this formula to figure out the theoretical last value of regressed return via y = mx + b where m = slope and b = intercept.  So I would plug in X value for last bar of chart and the other values which I&#039;ve calculated and this would then spit back the ending value of this regressed line.  From there it should then be pretty easy to calculate the RAR%.  Is this the approach you took?</description>
		<content:encoded><![CDATA[<p>Ok&#8230;I think I just realized I must use this formula to figure out the theoretical last value of regressed return via y = mx + b where m = slope and b = intercept.  So I would plug in X value for last bar of chart and the other values which I&#8217;ve calculated and this would then spit back the ending value of this regressed line.  From there it should then be pretty easy to calculate the RAR%.  Is this the approach you took?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jdfagan</title>
		<link>http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/comment-page-1/#comment-1001</link>
		<dc:creator>jdfagan</dc:creator>
		<pubDate>Mon, 24 Nov 2008 06:25:30 +0000</pubDate>
		<guid isPermaLink="false">http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/#comment-1001</guid>
		<description>Nice writeup.  I was just trying to figure Curtis Faith&#039;s robust metrics for my own use as well - RAR% and RRRR.  I&#039;ve figured out the AvgMaxDD stuff but before I can accurately calculate RRRR, I was wondering if I was calculating the RAR% correct.

I calculated the slope of the equity curve but not sure how exactly RAR% maps to slope.  For instance, I&#039;ve outputted the equity bar by bar (240 min bars in my case for a futures market) and import this into Excel to perform the calculations of slope, intercept and correlation coefficient.  Does one just multiply slope result by 100 to make it % or is it already expressed as %?

Further, I was wondering if you first take the log10 of the equity so that you take into account geometric growth over time - this would impact the slope I believe and thus the RAR% correct?

JD Fagan</description>
		<content:encoded><![CDATA[<p>Nice writeup.  I was just trying to figure Curtis Faith&#8217;s robust metrics for my own use as well &#8211; RAR% and RRRR.  I&#8217;ve figured out the AvgMaxDD stuff but before I can accurately calculate RRRR, I was wondering if I was calculating the RAR% correct.</p>
<p>I calculated the slope of the equity curve but not sure how exactly RAR% maps to slope.  For instance, I&#8217;ve outputted the equity bar by bar (240 min bars in my case for a futures market) and import this into Excel to perform the calculations of slope, intercept and correlation coefficient.  Does one just multiply slope result by 100 to make it % or is it already expressed as %?</p>
<p>Further, I was wondering if you first take the log10 of the equity so that you take into account geometric growth over time &#8211; this would impact the slope I believe and thus the RAR% correct?</p>
<p>JD Fagan</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Colin McGinley</title>
		<link>http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/comment-page-1/#comment-875</link>
		<dc:creator>Colin McGinley</dc:creator>
		<pubDate>Fri, 09 Nov 2007 02:50:42 +0000</pubDate>
		<guid isPermaLink="false">http://www.forexspirit.com/2007/11/07/lies-damn-lies-and-statistics/#comment-875</guid>
		<description>The inherent nature of trading only in the direction of the main trend means that entries in the lower part of the grid should have lower risk (which is why I&#039;m willing to use higher geared trades lower in the grid).

For example, only 5% of my Q1 entries have been closed out for a loss; 18% of Q2 entries, and around 23% for both Q3 and Q4 entries.

I have not calculated the holding time for the trades.  This is something I can add in.  Thanks for the idea.

I have attempted hedged trades a couple of times but I&#039;ve never been very comfortable with the act of hedging in the classical sense (having two trades open against each other in the same currency pair).  Lately, I have been using what I refer to as &#039;Anti-Hedge&#039; trades, a term coined by Jacko on Forex Factory where I picked this idea up (not that it&#039;s a new idea or anything).  If I think a trade is going to be moving in the wrong direction for a while I will just close it out.  When price returns to the level I exited at I will just put the trade back on again.  This is effectively what you are doing when using classical hedging, but for some reason this way of doing it sits better with me.  

I don&#039;t have exact figures on the maximum drawdown that each trade experienced from the data I have available.  All I really have to go to provide you an answer with is my experience to date.  If a trade goes against me 100-150 pips there is a strong likelihood that is could go against me for another couple of hundred pips.  EUR-USD has a habit of making corrections ranging from 300 to 500 pips.  

I am more inclined these days to use the Anti-Hedging method to bail out of trades within that 100-150 pip window.  I have no problem in waiting for the correction to play itself out.  When it starts to move back in the direction of the long term trend I&#039;ll just jump back on board and look to re-establish those trades I&#039;d closed out on the way down.  They live to fight another day.

I&#039;m just getting to grips with USD-JPY and I&#039;m sure I&#039;ll learn to settle on a similar sort of pip range if things are going against me, but it will take some time to find out what works for me.  For now I&#039;m using the same sort of 100-150 pip range.</description>
		<content:encoded><![CDATA[<p>The inherent nature of trading only in the direction of the main trend means that entries in the lower part of the grid should have lower risk (which is why I&#8217;m willing to use higher geared trades lower in the grid).</p>
<p>For example, only 5% of my Q1 entries have been closed out for a loss; 18% of Q2 entries, and around 23% for both Q3 and Q4 entries.</p>
<p>I have not calculated the holding time for the trades.  This is something I can add in.  Thanks for the idea.</p>
<p>I have attempted hedged trades a couple of times but I&#8217;ve never been very comfortable with the act of hedging in the classical sense (having two trades open against each other in the same currency pair).  Lately, I have been using what I refer to as &#8216;Anti-Hedge&#8217; trades, a term coined by Jacko on Forex Factory where I picked this idea up (not that it&#8217;s a new idea or anything).  If I think a trade is going to be moving in the wrong direction for a while I will just close it out.  When price returns to the level I exited at I will just put the trade back on again.  This is effectively what you are doing when using classical hedging, but for some reason this way of doing it sits better with me.  </p>
<p>I don&#8217;t have exact figures on the maximum drawdown that each trade experienced from the data I have available.  All I really have to go to provide you an answer with is my experience to date.  If a trade goes against me 100-150 pips there is a strong likelihood that is could go against me for another couple of hundred pips.  EUR-USD has a habit of making corrections ranging from 300 to 500 pips.  </p>
<p>I am more inclined these days to use the Anti-Hedging method to bail out of trades within that 100-150 pip window.  I have no problem in waiting for the correction to play itself out.  When it starts to move back in the direction of the long term trend I&#8217;ll just jump back on board and look to re-establish those trades I&#8217;d closed out on the way down.  They live to fight another day.</p>
<p>I&#8217;m just getting to grips with USD-JPY and I&#8217;m sure I&#8217;ll learn to settle on a similar sort of pip range if things are going against me, but it will take some time to find out what works for me.  For now I&#8217;m using the same sort of 100-150 pip range.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

