The conflict in what you guys are describing is that exploration runs
once per symbol. In other words, you can have at most one row per symbol
in the output.
For example, you cannot have the following because it would assume more
than a single row per symbol:
XYZ ABC 0.867
XYZ JKI 0.932
...
If you want to see the correlation between one symbol and every other
symbol in the watchlist, then you must use a table format as pointed out
by Herman. You can just zero out the columns that don't meet your
minimum criteria, and even use a different color for those uninteresting
cells.
If you only want to see predetermined pairs, then write your code to
specifically check for those combinations.
periods = 252;
correlated = 0;
other = "";
if (Name() == "MSFT") {
other = "AAPL";
correlated = Correlation(Close, Foreign(other, "C"), periods);
} else if (Name() == "ORCL") {
other = "IBM";
correlated = Correlation(Close, Foreign(other, "C"), periods);
}
Filter = correlated > 0.8;
SetOption("NoDefaultColumns", true);
AddTextColumn(Name(), "Ticker 1");
AddTextColumn(other, "Ticker 2");
AddColumn(correlated, "Correlation");
Mike
--- In [email protected], "ezbentley" <ezbent...@...> wrote:
>
> Hi Herman,
>
> I am aware of the correlation table. However, that table will show the
correlation among ALL stocks.
>
> What I am trying to accomplish is something like this(pseudo code only
to illustrated my intent):
>
> Correl = Correlation(StockA, StockB, 250);
> Filter = Correl > 0.8;
> AddColumn(StockA, "Ticker1");
> AddColumn(StockB, "Ticker2");
> AddColumn(Correl, "Correlation");
>
> And the output of the Exploration should look something like:
>
> Ticker1 Ticker2 Correlation
> XYZ ABC 0.867
> DFG JKI 0.932
> ..........
>
> In other words, I want to filter for high correlation pairs and output
them in a tabular form as supposed to a matrix.
>
> Any help will be greatly appreciated,
>
