[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Corrected Reply to John Sellers



PureBytes Links

Trading Reference Links

Would like to add that with a software capability similar to IRL you may 
compile an index of the S&P in a few seconds. This index will compile the 
accumulated 500 stocks and compute the price for market weighting as you 
prefer. Even so it may not much of an effort given the program stated, 
would the these results of 500 stocks adequately reflect the NY market? I 
am not sure but feel the data on 500 index would be accurately portrayed.

-----Original Message-----
From:	Joseph Ehardt [SMTP:jehardt@xxxxxxxxxxxxxx]
Sent:	Friday, March 19, 1999 7:05 AM
To:	MetaStock Discussion
Subject:	Corrected Reply to John Sellers

Since the NYSE Composite Index represents all the companies listed on the
exchange, the official NYSE volume data seems like a good match with that
index. And you are right, they do publish the advancing, declining, and
unchanged volume data for the entire exchange, but it might be only the OBV
(On Balance Volume) of each equity accumulated together, and if that is the
case (I have been browsing the NYSE Exchange site to see if that is the
case without obtaining an answer), it would misrepresent what investor
money is actually doing. So I made an attempt to perceive volume in a more
meaningful way.  In another message about to be posted, I have attached two
charts comparing the official exchange data and my allocated volume data to
show the differences.

Volume data for the S&P 500 is not published, and I'm not up to the task of
nightly compiling the volume of each of the 500 equities into a total
volume figure for that index, which I could then use in my cumulative
allocated volume algorithm. Surrogates might be used, such as the volume
for the S&P Futures contracts or for Spyders (tradable entity, S&P500
Depositary Receipts), but these also are independent items that can be used
as hedging instruments and the like, so volume in a contract or equity
would not be valid to reflect activity in the index itself. For that,
component volume is needed, and it requires too much work.

At the same time, the idea of CAV could be applied directly to the trading
volume of the mentioned entities. That is, I could use price and volume
data for Spyders, or I could use price and volume data for a particular
futures contract, and from that data make the appropriate chart of each.

And as far as what kind of weighting to use for an index, I would probably
favor capitalization weighted index, because it correlates to what money is
actually doing -- 100 shares of Intel are worth a lot more that 100 shares
of National Semiconductor, and for me being able to translate price *
volume into money would be important. How is money behaving? Is it going
into an instrument or being removed from it?

Joe



John wrote:

>Sounds interesting do you use the total NY volume as opposed to volume for
>the Composite which as far as I know is not available. Also I suppose one
>could use the advances and declines for the NY total to approximate that 
of
>the Composite.
>
>Another approximation would be to generate the S&P index with volume. But
>what type of index calculations would you suggest to be the better. I know
>of five types within my IRL software and they are Average Price Index,
>Price Weighted Index, Market Weighted Index, Equal Dollar Weighted Index
>and Portfolio Weighted Index. Just guessing I would say the fourth one may
>be better. The volume for the index I believe is calculated by summing the
>individual volumes of each stock included within an index. The prices are
>calculated by different formulas for each of the indices. As the titles
>imply the weighting is in general is accomplished to some extent in each 
of
>their own formulas.
>