Yes, that is definitely something we plan to offer.
Yuval,
Is it an issue of Factset not offering short interest data and institutional ownership data or aggregating it differently or from different sources? One thing I’ve noticed is the Core Sentiment ranking is actually pretty dependent on short interest. When you remove that portion of the ranking sentiment doesn’t work as well. At least I have noticed that in my backtesting. My biggest concern would be whether sentiment would be harder to assess if short interest factor is either not available or of lower quality.
Jeff
I will add my two cents in on this issue. Short interest data is critical to my models!
Short interest data is (recently?) tightly controlled by NASDAQ and they want a lot of money to allow sites like ours to display it. It’s at least $60k/year not including other exchanges which probably want similar amounts. We are getting it through S&P probably from some old licensing agreement between them. It’s the only explanation otherwise our data costs would be double. Factset has no such agreement. They have the data but require us to get a license directly from exchanges.
We’re still exploring ways to continue to provide short interest, or similar.
-
We’re exploring getting it through a price data provider included in a package
-
Another option is for us to import this free data: BZX Short Sale Daily Reports . Data goes back to 2008. It’s all the short transactions , not the total current short position. It’s also not clear if the cover transactions are included (doesn’t look like it). But maybe , simply analyzing the trend of the daily short transaction totals (like an OBV trend) , might turn out to be very correlated to what we currently offer , the "Short Interest, Current Month Position " and related “Short Intererst Ratio”. Or maybe better. I have never see this data used anywhere. But I think it’s very interesting.
-
With our new Factor Import function you could subscribe to shortsqueeze.com and import it yourself for your own use (we will also have an API to do that as well) Their memberships run about $40/mo. I plan to contact them to see if they want to resell it through us as well.
If you have any other suggestions let us know
Thanks
Marco,
Thanks for the update and information. Appreciate hearing you are working towards a solution going forward. I used to subscribe to InvestorsEdge and they used Factset and did not have access to short interest. Suffice it to say they didn’t have many if any sentiment based strategies because it was something they couldn’t much capture. As has been said short money is smart money.
Are there other indicators which are highly correlated with short interest that have a causal relationship? Indicators that might substitute?
Thanks,
Jeff
According to some academic papers, share turnover is highly correlated with short interest. In the S&P 500 right now the correlation is 0.62 and in the Russell 3000 it’s 0.60. I’m comparing AvgVol(65)/Float to SI%Float.
Hi Yuval, Can you share the process you used to calculate the correlation?
Walter
I ran a screen with those two rules and on those two universes, downloaded the screen report into Excel, and used the CORREL function on the resulting columns.
Yuval,
Thanks. It’s not perfect but it does reach similar results albeit slightly lower returns in backtesting. As it does not perfectly correlate one could expect some lower predictive power compared to short interest. Would still prefer short interest but does show a way forward if short interest is absolutely unattainable.
Jeff
This already sounds problematic…
Are we able to sign up for CapitalIQ ourselves and run their database through the P123 front end?
Interesting! Thank you for detailing your method.
So, literally you have a Spearman Rank correlation of these 2 factors (for one period), I believe. I assume you are correlating the ranks.
And easy to get.
Best,
Jim
I expect a lot of things will break. It will be interesting to see what happens to the DMs.
Walter
I hope not much will break but we will see. If it does break it will be interesting to know what breaks it.
It could be better. And we could find that it was broken all along and we just did not know it. Remember we can get PIT earnings estimates from the data provider now.
If we have had PIT earnings estimates here already it is only because Marco has made a special effort to do snapshots.
Fundamentals change only once a quarter whereas estimates can (and do) change much more frequently.
Anyway, I am optimistic that it might even be better.
Regards,
Jim
Thanks! So that correlation is for one particular date. Is there a method to see it as a time series?
how about momentum. is it correlated also?
I asked earlier but never got an answer.
I believe factset does have Put and Call price data but I don’t know if they have volume data. If they do then maybe P123 could make a Put/Call ratio if that date is included in the P123 package or a low cost addon. It’s not exactly short interest but it does have correlation and would be better than nothing IMO.
We’ve had this question from a few users and one university. We could have the capability to support clients that have their own S&P license if it makes sense financially. S&P ballpark, list price for data licenses is around $24K/year but we’ve seen them make deals with some professionals at 10K-12K strictly for access to their data with P123. If this is interesting to you contact joseph.smith@spglobal.com and let him know.
Hi Yuval and Marco,
I was wondering if you have any updates (or preparatory materials) to help subscribers make best use of the 3 month transition period beginning April 1st during which both S&P and FactSet data is available?
It seems like a one-time opportunity to make adjustments. After June 30th, I don’t know how one could make the kind of comparisons that should be possible during the upcoming window.
Thank you very much.
Hugh
I second Hugh’s request.
Thx
Me too!