it looks like “all the #SPEPS functions” are being redone for Factset.
These are used often for market timing rules.
Any idea on their status? thanks.
I don’t know if this is a bug. In my spare time, I am trying to compare outputs of my sims using the FS engine.
For most part, the simulations are similar, however, I seem to get consistently higher turnover Factset. What could the reasons be? How Final statements are handled?
Also, ranking is almost always similar, but different.
How do I count how many instances of a particular factor in the ranking is actually available in the population data. (ie FCF, Net equity issuance, etc)
One possibility is a larger universe.
In your ranking system, click on the Ranks tab; in the columns box choose “Include Composite and Factor Ranks”; then click Run. Any figures that are N/A will be colored white.
Hmm… I don’t think Larger universe is a possibility as I often limit my sim to R1000 constituents.
I will look at the ranking outputs. Thanks for your help, Yuval
wow, real good performance now on running sims, great, thank you!!!
Best Regards
Andreas
-
I still get some variance in my sims (actually for the better since a week, 39,58% on old data, 39,17% on new data with my best system, last
10 Years, very, very encouraging, I might not have to change anything :-)))))))))))))))))))))))). That said, I would like to get a rough estimation from p123 when this will actually “freeze” (It will never freeze, I understand this, you get the idea), e.g. my (or our all)
sims do not vary anymore.
It would be really great if P123 Team would start a new general thread (“Low hanging fruits with the new data provider”, something like that) on this board about best practices on how to adapt the ranking systems to the new data provider. There might be stuff, that the new data provider is doing much better (insider data?, better industry data whatever) and that we could profit on.
Also, p123 could point out what worked good on the old data provider, but does not work well on the new data provider and how one
can fix this easily. There might be also some real nice low hanging fruits that we all could implement in a fast way.
I know this is a thin line bc. we need to protect our models, so that can not be done with a whole ranking system, but
maybe p123 (and its users) could give line item examples, that would be really great.
Or p123, just give away your best secrets, that would be fine too 
(you get the irony)
Thank you for the million tons of work to all the p123 people and users who helped (I have to admit, I did not!!) to get
this transition done!
Thank you in advance!!!
Best Regards
Andreas
What works significantly worse on Factset → P123 Greenblatt ranking (which is a shame as it was allowing to get good results with minimal efforts).
- This is true on a wide range of curated large caps universes that I use.
- This is particular true for financials. I understand that Compustat was doing things to try to make Fin Serv fundamentals act like normal industries. Maybe they actually succeeded by luck or design since the P123 GB ranker was doing quite well
Jerome
I would like to know the date when P123 dumps Compustat and switches only to FactSet.
I get better results with FactSet for most of my sims. So not having Compustat data is not a concern for me.
Thanks.
Thank you Jerome!!!
Do we know when Ownership data will be posted on the new FactSet engine?
Thanks.
Hugh
any updates right before the cutoff date 26th of June?
Isn’t the cutoff July 1st? Can we have an update for the deal with S&P before the cutoff date too.
Please??? Feels like full radio silence the last few days and the deadline is fast approaching.
I got a call back from Tim Love (S&P Canada), he says the agreement is getting finalized so users get a seamless transition. He confirms the cutoff date is actually June 26th. He says there will be a grace period between June 26th to June 30th for those who plan to sign the new agreement terms later this week, everyone else is losing access this week end. He’s going to send me a draft to review tomorrow on the terms and conditions.
its 12k a year right?
It’s gonna depend on what your company does, etc. You have to call to get your quote.
Just to reassure you:
We’ll be switching everyone over to FactSet on June 26th, as planned. There should be no interruptions or loss of data. We’ll make sure to bridge any gaps. We have found a provider for short interest data and will let you know more details soon.
For more details, please see this document: https://docs.google.com/document/d/17Q0cVnGdGwKGEXFZOuwf297itKdqUC_KGN2UHcMTdYg/edit?usp=sharing
Today, 2 days before the switch, my sims running on the FactSet engine show holdings allocation for GICS sectors and not RBICS sectors.
Also will SecCount and SecWeight use RBICS correctly as of Friday?
Georg, we’ve been working on getting front-end support for RBICS finished up.
All diversification functionality is using RBICS for screens and simulations run using FactSet; since those servers don’t even load GICS, there’s no other possibility.
Unfortunately, we won’t be able to launch the front-end RBICS support prior to the cutoff, so users who have not procured an S&P data license will start seeing RBICS as of Friday morning.
Aaron, thank you for the reply. I am looking forward to RBICS; this classification system has 11 sectors, much improved classification.
Will the backtest period also be extended from the current max of 15 years to 20 years or longer as previously announced here:
https://www.portfolio123.com/mvnforum/viewthread_thread,11932#!#69910.