I would expect Worktable1 to contain only those columns referenced in the query.
At a minimum the following columns would be in Worktable1:
marketDataSet.M_ALIAS
marketDataSet.M_LABEL
Also in Worktable1 would be any marketDataSet columns mentioned in the select/projection list (defined as 'xxx' in your SQL code snippet).
As for the clustered index ... *shrug* ... not sure if it would be (M_ALIAS) or (M_ALIAS, M_LABEL). [Does optimizer build index solely on join columns, or does it also include SARGs?]
