|
PureBytes Links
Trading Reference Links
|
>Date: Tue, 27 May 1997 23:28:26 -0700
>Reply-To: realtraders@xxxxxxxxxxxxxxxxxxx
>Sender: owner-realtraders@xxxxxxxxxxxxxxxxxxx
>From: "schong@xxxxxxxxxxxxxxxxxx" <schong@xxxxxxxxxxxxxxxxxx>
>To: RealTraders Discussion Group <realtraders@xxxxxxxxxxxxxx>
>Subject: Re: Gann Swing Chart(using Time)
>X-To: realtraders@xxxxxxxxxxxxxxxxxxx
>X-Mailer: Mozilla 3.01Gold (Win16; I)
>
>
>There are several ways of using Gann's Swing Chart
>with time as the filter. Anyone who has anything
>to add to what I've said please feel free to do
>so.
>
>I will be using the default of 2 as the input to
>"filter" hence a two bar swing chart. I prefer to use
>time as the filter rather than points because they
>are more responsive to market actions and also
>because you give less profit away too. Anyone who
>has used swing charts based on points would know
>what I'm talking about.
>
>The input "stats" is suppose to act like a
>boolean field. When it's set to 0 it will
>produce the swing chart that looks like
>zig zags. When it's set to 1 it will produce
>the range of the swing in the form of a
>histogram and the duration of the swing
>as "+" overlaid on the histogram. Duration
>is expressed as the number of bars. Any other
>value for "stats" would have the same effect
>as if 1 have been specified.
>
>
>Like any type of swing charts, it purpose is to
>filter off noise. In this case any swings in the
>opposite direction of the current prevailing trend
>as dictated by higher swing highs and higher swing
>lows for a bull market within the context of your
>trading time frame which last no longer than
>2 bars unbroken in a row would be ignored. Reverse
>the rule for a bear move.
>
>
> You may use a 5, 10 and 30 minutes swing chart.
> Take a trade on the 5 minutes swing chart only
> when all 3 swing charts are pointing to the
> same direction. You wouldn't compare 3 swing
> charts together unless you are trading
> intraday.
>
> If you had used daily, weekly and monthly
> swing charts as substitutes for the above you'll
> be sitting on the fence for most of the time.
>
>
>Now comes the juicy part. Wouldn't it be good
>if you can get an expectation of the move
>in points and/or duration in number of bars
>once a swing has been confirmed which could serve a
>number of useful purposes, .eg to confirm the most
>likely retracement levels that would be applicable,
>or the most likely Energy Point that would work to
>name a few.
>
>
>This is where the other mode of study comes into
>use. You would set "stats" to 1 for this purpose.
>You would then have to perform some statistical
>studies on the duration and range of each swing.
>
>You can get the data required from the plotted
>indicator and export them out into an ASCII file
>which would then be imported into a spreadsheet.
>
>There you would extract the columns for duration
>and range and calculate the average duration and
>range for each swing. This will be used to set
>your average expectation of each swing.
>
>In addition to that you would get the mod of the
>duration and range as well. The mod is simply
>another measure of centre tendency like the
>average. There is evidence to suggest that Gann
>calculated the mod as well. They differ because
>the average sums all observations up and
>divide by number of observation whereas the mod
>looks for clusters in the observation, so in
>essence you may have a number of likely moves.
>Say you can see 3 groupings in the mod study,
>you would say that once the 1st group is
>exceeded the next level to expect is the
>2nd group and so on.
>
>Get a copy of the 45 years in Wallstreet and
>study his rules. You would be able to relate
>to what I've said better.
>
>You will probably want to update the statistics
>every couple of months.
>
>A way to gauge whether the market is showing
>sign of weakness is to look at the range and/or
>duration of each swing in relation to the
>previous alternate swings denoted by the height
>of the histogram for the range and the level of
>the "+" for the duration.
>
>In a bull market, again within the respective
>context of your trading time frame, reactions
>in price should get smaller and smaller, the
>first time a reaction exceed the size of the
>previous alternate range is an indication
>that a top is in the making. You can observe
>this very clearly by looking at the height
>of the histogram and comparing it with
>previous alternate swings.
>
>The probability increases for a top if it's
>further confirmed by a larger duration in
>the reactions. You may assume that the bull
>run of that particular degree of which
>you may use elliot wave or symmetry wave to
>determine could be considered to be over again
>by looking at the level of the "+". This
>is what Gann termed "OverBalancing of Price
>and Time".
>
>
>
>cheers,
>
>Clement Ong
>
>
|