Use a function to get the time interval:

Function tmIntvl(vtime)
Dim hr
Dim min
vtmntval = "NONE
if hr<10 or hr>=16 then exit function

hr = Hour(vtime)
min = (Minute(vtime) \ 10) * 10
tmIntvl = TimeSerial(hr, min, 1) & " - " & TimeSerial(hr, min + 10, 0)
End Function

Use
 tmIntvl([Time])  as period
 instead of your IIF

This query gives you basically the same answers:

    TRANSFORM Last(B.Rate)
    SELECT B.Symbol
    FROM broadcast AS B
    GROUP BY B.Symbol
    PIVOT CDate(Int(B.Time*144)/144);

That is, no answer at all, because Last() does not what you expect it to do. Last gives you any one value from the group -- the last encountered, and there is no syntax to specify which order you want to apply for the Last() function.

In other words, your query will get even slower by several orders of magnitude. For example, this is correct but terribly slow:

    SELECT B.Time, B.Symbol, B.Rate
    FROM broadcast AS B
    WHERE Not Exists (
        Select True From broadcast
        Where Symbol=B.Symbol
            And Time > B.Time And Time < (Int(B.Time*144)+1)/144
        );

Even with indexes on Symbol and Time, the calculation hurts the query, and nothing gets optimized. Instead, you can add a new field:

    Slice: Number, Integer; Indexed

And run this:

    UPDATE broadcast AS B SET B.Slice = Int(B.Time*144);

If needed, the equivalent time is: CDate(Slice/144) to CDate((Slice+1)/144). With all three indexes (on Symbol, Slice, and Time), this is reasonably fast:

    SELECT B.Time, B.Slice, B.Symbol, B.Rate
    FROM broadcast AS B
    WHERE Not Exists (
        Select True From broadcast
        Where Symbol=B.Symbol
            And Slice = B.Slice
            And Time > B.Time
        );

However, it uses a sub-query, so that it can't serve as source for a cross-tab. That's one of the annoying limitations of Jet SQL. You can now either use a make-table query and use the new table as source, or run a second update query, for the new field LastInS: Yes/No

    UPDATE broadcast SET LastInS = False;

    UPDATE broadcast AS B
    SET B.LastInS = True
    WHERE Not Exists (
        Select True From broadcast
        Where Symbol=B.Symbol
            And Slice = B.Slice
            And Time > B.Time
        );

LastInS indicates which records we want to see (113 among the 8888 you showed us). And now we have a fast cross-tab, for example:

    TRANSFORM First(B.Rate)
    SELECT B.Symbol
    FROM broadcast AS B
    WHERE B.LastInS
    GROUP BY B.Symbol
    PIVOT CDate(B.Slice/144) & '-' & CDate((B.Slice+1)/144);

or (looks better and is even faster)

    TRANSFORM First(B.Rate)
    SELECT CDate(B.Slice/144) AS [From], CDate((B.Slice+1)/144) AS [To]
    FROM broadcast AS B
    WHERE B.LastInS
    GROUP BY B.Slice
    PIVOT B.Symbol;

Cheers!
(°v°)

> after every 10-15 minutes

In that case, work only with the latest data. Keep track of the time you run the procedure and limit your queries to only the "time slices" that can still change.

If you added the fields Slice and LastInS, run the first update query for only "Slice Is Null" and the second for only "Slice >= (the corresponding number)". It will be even faster. But the update queries are already quite fast, and you probably tested the cross-tab: it's now almost instantaneous, isn't it?

Do not use the make-table query, as that would force you to regenerate the entire table every time, and it can't be faster than the update query.

(°v°)