|
PureBytes Links
Trading Reference Links
|
Just to confirm an earlier rter's question about scaling so it is perfectly
clear....are you sure FEAR should be scaled to "screen" and not "price" in
Trade Station? When assigning the FEAR indicators to the same subgraph as
price and scaling to screen will produce different price-indicator
relationships when expanding or compressing the dates in the window as it
does with other indicators. This can be very misleading(lagging).
BobR
At 01:26 PM 7/12/97 -0400, Walt Downs wrote:
>Rt'ers
>
>Just thought I would post an update to let everyone now what's
>going on in the evolution of FEAR!
>
>First, to all of you that are expending time and energy working
>with the indicator, your input is greatly appreciated !
>
>Available information on the indicator is expanding at an
>amazing exponential.
>
>Here is what we are finding out:
>
>As most of you have noticed, FEAR! is designed to adjust to
>current market conditions. To a certain degree, the FEAR
>and FEARfast can be considered support and/or resistnce.
>HOWEVER..... there seems to be a lot of confusion on the
>use of the FEAREnv. The current situation in bonds is a good
>example. What the FEAREnv is telling you is that, at these
>levels, everytime bonds move higher, it must "carry" the
>FEAREnv with it. A lot of traders (the ones who I suspect
>are currently long bonds :) ) have commented that the
>market is moving higher, while blissfully ignoring the fact
>that everytime Bonds have hit the FEAREnv level, it has been
>good for a 6 to 10 point reversal before bonds recover.
>
> It is also my contention that, when bonds finally fold, it will
>be good for 1 to 2 full points. Does this mean that I am betting
>the house on bonds falling? I NEVER bet the house on anything!
>But, I am collecting options. And I have, and probably will
>continue to fade the bonds whenever it pops the FEAREnv level
>intraday.
>
>INTRADAY TRADING:
>
>I personally think it is better to use the plot values from
>the DAILY chart as intraday trading levels.
>
>However, some people that are playing around with the indicator
>on realtime intraday charts are having some interesting results.
>
>I have looked at quite a few offline intraday charts using
>FEAR!, and find that the FEAREnv has some problems holding
>the market back. I believe this is because a person holding
>$300 in profit or loss experiences a lot les fear than a person
>holding a $3000 gain or loss!
>
>I have a query for those looking at fear on intraday "live"
>charts. What kind of bounce are you getting off the various
>plots as the market moves around? I would love to know.
>There are some things I could try in order to strengthen
>the FEAREnv for live intrady trading.
>
>General Information:
>
>So basically, where we are at with FEAR!! right now is that
>it looks cool, and SEEMS to be working pretty well.
>
>The next step is to establish some sort of historical testing
>structure so that we can "prove" the veracity of the indicator.
>
>To that end, Clyde Lee has kindly offered his help. Clyde seems
>to be on the ball in his EL programming, and I am confident that
>if we get enough people working on this, we will be able to come
>up with a "Living" system for a "living " indicator.
>
>In closing, the following is a basic synopsis of the indicator and
>how it functions:
>
>
>The scaling. The idea behind the use of the indicator in
>this scale is to create, in effect, a "living" indicator. When you
>look back in time on a daily chart, you will notice the FEAREnv
>confirming most of the highs and lows, and you will also see the
>validation of the various breakout levels, as the market passes
>through the various plots. The indicator is giving you active
>feedback, saying " My signals are still valid" So, in theory (And
>in practice, for me :) )You know that the market hasn't changed
>and the indicator should still be tradable.
>
>So, how does this differ from curve fitting a system?
> The difference
>lies in the fact that the indicator is allowed to adjust as it
>receives new price data! For example : If I were to look at the
>results the indicator showed over the past 6 months, and based on
>those extrapolations, I then wrote and applied a system, the system,
>now having set mathematical parameters it MUST follow, is
>eventually doomed to fail. When the market changes it's temperament,
>the system won't change with it.
>
>So the idea I'm shooting for is this: Since the indicator is allowed
>free reign to change as it receives new data, if the market starts
>to change, the indicator simply will not isssue a signal based on
>the old parameters. In effect, keeping the trader out of a market
>that has changed it's temperament.
>
> What I am really anxious to see
>is if the indicator, when faced with a market that has changed, will
>MUTATE, and generate a new standard of signals, in effect giving
>the trader a road map to trade the market.
>
>Yes, if the market applies enough force to the FEAREnv, it will
>move. But remember, fear doesn't go away, Every day that the
>market attempts to move higher, it will be forced to "push"
>the FEAREnv with it.
>
>So, the REAL question at this point is whether or not the FEAREnv
>is strong enough to drive back a strong market.
>
>I have been fortunate in that currently there are 4 or 5 markets
>that are testing the FEAREnv: Bonds, OJ, Cotton, the S&P.
>
>If you will look at the dailies, you will have to admit that these
>markets DO seem to be having trouble making headway against it.
>
>I think Bonds will be the sternest test.
>
>If several of these markets do break through, I will increase the ratios
>of the FEAREnv to see if that strengthens it.
>
>The main problem right now is in historically testing it! How do you
>test an indicator that constantly changes!? That is what I am
>working on right now.
>
>The only historical testing to date is my own trading. And "walk
>throughs" going trade by trade on TS charts.
>
>I have traded signals from the FEAR indicators "real time
>and "real money" five times.
>
>Results: 4 winners and 1 loser.
>
>Any comments, input or ideas are encouraged.
>
>Walt :)
>
>
>FEAREnv is dynamic, and
>the market will have to drag it around like a ball and
>chain until it figures out what to do with it.
>
>
|