NYSE TAQ
NYSE TAQ on WRDS
You may find the WRDS overview on NYSE TAQ data here (WRDS sign on required).
WRDS developed the intraday indicator dataset (IID) to improve efficiency and lower the computational barriers when researching with NYSE TAQ data.
WRDS IID homepage is here; web query is here; data manual is here; WRDS IID is accessible based on your NYSE TAQ subscription.
WRDS IID SAS Studio Demo
For SAS programmers, you may use SAS Studio with your WRDS account: https://wrds-cloud.wharton.upenn.edu/SASStudio/
You can try the following sample code that replicates the main finding of Holden and Jacobsen (2014, Journal of Finance).
Note this example uses Tesla's TAQ data for 2014 only. It takes about 5 seconds to run.
libname iidm '/wrds/nyse/sasdata/wrds_taqms_iid'; *extract pre-calculated simple average percent effective spread from Daily TAQ;data work.iidm_test;set iidm.wrds_iid_2014;where sym_root='TSLA';keep sym_root date EffectiveSpread_Percent_Ave; run;
libname iid '/wrds/nyse/sasdata/wrds_taqs_iid_v1'; *extract pre-calculated simple average percent effective spread from Monthly TAQ, previous second method and interpolated method;data work.iid_test;set iid.wrds_iid_2014;where symbol='TSLA';keep date symbol ESpreadPct_Avg1 ESpreadPct_Avgi; run;
proc sql;create table compare as select a.*, b.ESpreadPct_Avg1, b.ESpreadPct_Avgifrom iidm_test a left join iid_test bon a.sym_root=b.symbol and a.date=b.date; quit;
proc sgplot data=compare;title "Replication- Holden and Jacobsen 2014, JF, Improvements Using the Interpolated Method";series x=date y=EffectiveSpread_Percent_Ave / lineattrs = (color = navy thickness = 1 );series x=date y=ESpreadPct_Avg1 / lineattrs = (color = green thickness = 1 pattern=shortdash);series x=date y=ESpreadPct_Avgi / lineattrs = (color = orange thickness = 1.5 pattern=shortdash); run;
The spread generated using the interpolated method (with the Monthly TAQ data) delivers a very close proxy to the spread generated from the Daily TAQ using the previous millisecond (trade-to-quote matching) method. The interpolated method is a significant improvement to the previous second method using the same Monthly TAQ data.
- - - previous second method (Monthly TAQ)
— interpolated method (Monthly TAQ)
- - - previous millisecond method (Daily TAQ)