|
PureBytes Links
Trading Reference Links
|
Carrying this one step further by normalizing the bands facilitates the
buy/sell signals.
---------------------------
{VIX IN DATA1, OEX IN DATA2}
INPUTS: N(300);
VALUE1=((AvgPrice[1] OF DATA1)/SquareRoot(N))/100; {equals one std dev}
VALUE2=((AvgPrice[1] OF DATA2)* VALUE1); {equals one std dev of price}
VALUE3=(CLOSE OF DATA2 - AvgPrice[1] OF DATA2)/VALUE2; {equals the detrend}
PLOT1(VALUE3,"DIST");
----------------------------------------
BobR
At 09:30 PM 11/15/97 -0800, bobrabcd wrote:
>This is very interesting DOC. As you point out one of the keys here is the
>volatility expectation for the next day. If we use the Keyline Pivot Point
>(H+L+C)/3 or (O+H+L+C)/4 or AvgPrice in Easy Language as a projection for
>both the volatility and the OEX for the next day and then run it through
>your calculations using VIX as data1 and OEX as data2 then we come up with
>the volatility projection bands as depicted in the attached .gif. One
>interesting thing here is these bands do not require the averaging of price
>or volatility over an assumed number of days as is done in Bollenger bands
>etc.
>-------------------------------
>{VIX IN DATA1, OEX IN DATA2}
>INPUTS: N(300);
>VALUE1=((AvgPrice[1] OF DATA1)/SquareRoot(N))/100; {equals one std dev in
>percent of data}
>VALUE2= (AvgPrice[1] OF DATA2)*(1 + VALUE1); {upper band}
>VALUE3= (AvgPrice[1] OF DATA2)*(1 - VALUE1); {lower band}
>PLOT1(VALUE2,"");
>PLOT2(VALUE3,"");
>----------------------------------
>
>BobR
>
>
>At 06:20 PM 11/15/97 -0600, THE DOCTOR wrote:
>>It is a pretty simple calculation. I merely took the VIX...currently
>>about 34%......this reprsents an annualized number...and converted it
>>into a one day estimate.
>>
>>The vix assumes about 300 observations a year for volatility.
>>Converting a yearly number to a daily number is simply taking the year #
>>34% and dividing it by the square root of the days. The square of 300
>>is about 17. I'm doing very rough math to keep it easy. 34% annual vol
>>divided by 17(to convert to daily)is 2%.
>>
>>A one std deviation move would be +/- 2% (this should happen about 68%
>>of the time.
>>
>>
>>A two std dev move would be +/- 4%
>>
>>A three sd move would be +/- 6% and this should/could occur about 1% of
>>the time.
>>
>>With a Dow at 7600 this equates to about a 450 point move which we Could
>>see more frequently than once every ten years.
>>
>>The fact that we had one soon thereafter is sort of indication that we
>>are really in a very volatility market.
>>
>>In previous posts I've talked about the "fat Tails" condition which
>>actually distorts implied vol and of course the VIX. Real implied, if
>>the price distribution were normal, would be lower. Except for the last
>>decade we have lived in a period of non normal distributions...the
>>condition of fat tails. Fat tails are probably associated more with the
>>state of technology and trading cost(TOO LONG TO POST THE WHOLE
>>CONCEPT),but essentially because it's cheaper and easier to trade, there
>>is much more trading and therefore the market moves and stock moves are
>>probably bigger than expected.
>>
>>I did a seminar this morning for an Omega users group in 1000 Oaks
>>California and the topic of key interest was vol. Vol is what short
>>term option trading is all about...and one should have a good
>>understanding of it....a real opinion of where it will be on the next
>>day...or they should avoid short term trading.
>>
>>I know it sounds like a tease...but people who understanding option vol
>>have been coining money all year. There were actually a series of
>>parties in Chicago Tuesday night after the market break. The kind of
>>success stories being tossed around all had $1,000,000 plus numbers
>>attached to them.
>>
>>If you want to learn more a must read is Shelly Natenberg's book on Vol
>>based trading.
>>
>>Good luck.
>>A two std deviation move would be +/- 4%
>>
>>
>Attachment Converted: "c:\eudora\attach\vol bands1115.gif"
>
>
Attachment Converted: "c:\eudora\attach\VPDIST1116.gif"
|