Factors Missing

Several 5 year growth factors have disappeared since last week. So far I’ve discovered:

CF5YCGr%
PMgn5YCGr%
SalesPS5YCGr%

Was this announced? Were they marked as deprecated? Why have they been dropped?

And CapSp5YCGr%

And BV5YCGr%

and GMgn5YCGr%

Look like they are missing from the reference. We redid the reference to support the auto-complete feature. I think anything with a ‘%’ is missing from the reference but it’s still available if you use it in rules. Thanks for reporting it . We should have a fix tomorrow.

Thanks.

SGA2SalesTTM also appears to be missing. They are not accepted when entered into the Raw Editor for systems. They generate an error: { “message”: “ERROR: In ‘Factor’. Invalid value (SGA2SalesTTM)!”, “line”: “21” }

IntInc(0,TTM) is also missing.

So it’s not a bug… we’re still dealing with the remnants of Reuters definitions and naming conventions.

We removed all ratios that were in the DEPRECATED category in the rule reference as they interfered with auto-complete in the rules. They can still be used , but we would like you to transition over to the new versions (and when missing ones become available next week). The raw-editor will not support the old names, sorry about that.

  • SGA2SalesTTM was renamed SGA2Sales%TTM. See the help on SGA2Sales% , all the variations are there

  • IntInc(0,TTM). Compustat is not giving us any interim data on this, all N/A’s. Only annual is available.

For the 5Y compounded growth rates: you can still use them but they will not reappear in the reference. Some of the main issues with the ones that disappeared are:

  • Bad names
  • One off, like PMgn5YCGr% based on net income which include xor items (not recommended, use Net Profit Margin next week)
  • Inconsistent implementation. For examples Sales growth has a full set implemented but SalesPS growth only had 5Y

Of course you can write the growth rates yourself , but it’s very time consuming and error prone. For example:

Sales5YCgr% => ((SAles(0,ANN) -Sales(5,ANN))/Abs(Sales(5,ANN)))^1/5

We’re working on a permanent solution for next week. We’re adding functions to easily get annual growth rates for any fundamental function.

Thanks for bringing this up

Thank you for the update, is it possible to get an email announcement once this issue is resolved?

Any deprecated 5Y growth rate can now be recreated using Gr%5Y. See this post , thanks

Marco -

I appreciate and applaud the effort to standardize. However, it is important to take care in how these changes are implemented & phased in.

(1) 80 - 90% of the P123 Ranking Systems use growth rates or % change. Presumably all of the P123 ranking systems will be updated with functionally identical growth factors? In other words, don’t substitute P/E excluding Ext. Items for Including Ext. Items, etc. (These ranking / rating systems are used all over the place and in some very fine-tuned systems.)

(2) What about R2G models? Are we required to update the R2Gs or can we continue to use the old factors even though they may be undocumented? We are only allowed to make changes every 6 months. The last change to industry factors was about 4 months ago, and 2 months before that. I have difficulties keeping up, given the 6 month rule and would prefer to leave the models as is unless I see a reason to change the model, then of course I would update the growth factors.

(3) How are the industry growth factors going to be handled?

Thanks
Steve

These were deprecated first by placing them in a DEPRECATED section of the reference. Now they have been removed completely from the reference so its not confusing… But they still work if already in a system. If we decide to remove them permanently from the code base (unlikely anytime soon) we would first have a substitute, then do a find/replace, so your systems would be automatically updated.

It sure would be nice if the “ DEPRECATED section of the reference” was permanently available.
As a new user, I spend a lot of time looking through the public Ranking systems and screens.
They are full of these deprecated formulas and I have no way of looking them up.

Why deprecate the “deprecated section”?

Because at some point those old formulas may stop working. They are unsupported for a reason. We don’t want people using them and then having their strategies fail.

This was an issue recently with $EV, for instance. See https://www.portfolio123.com/mvnforum/viewthread_thread,12468 . . . A lot of people were using $EV rather than EV and their results were terribly inconsistent.

I understand. But that was five years ago. Even if the formulas stopped working today, keeping them in the reference with a “Deprecated” flag couldn’t possibly hurt, and would certainly help newbs.

I checked my ranking, all factors seem to be be based on the new ones, besides the following

(mktcap + DbtTotQ - (CashPSQ * ShsOutMR)) / Eval(EBITDAq>0,EBITDAq,NA)

→ ShsOutMR seems not to be there anymore…

Should I change to SharesFD(0, QTR, FALLBACK)?

e.g. (mktcap + DbtTotQ - (CashPSQ * SharesFD(0, QTR, FALLBACK))) / Eval(FCFQ>0,FCFQ,NA)

Thank you and best regards!!!

Andreas

pop

ShsOutMR is now Shares(0,Qtr,KeepNA)

Just use SharesQ.

SharesQ is the pre-built factor that uses the function Shares(offset,type[,NAHandling]) and corresponds to Shares(0,Qtr,FALLBACK), but much easier to write. It automatically falls back to previous quarter value if the value is missing during preliminaries. A very reasonable thing to do with the # of shares.

Using KEEPNA is really hard and should be done only when doing something very specific. The reason it’s hard is because in a long formula like yours some factors may fallback other not. So you really have to think of all the possibilities

Btw, We planning a way to make fallbacks easier to use (more consistent) in long formulas

PS . We’re introducing EV2EBITDA factors very soon so you can just use pre-built one and not worry about all the technical things.

Just one question, why are you using EBITAQ & FCFQ ? It akin to saying PEQ which I’ve never seen anyone use.

When we implement EV2EBITDA we were not planning a Q version. But if we did we would annualized the denominator to make it comparable to the TTM version (annualizing the Q is a bit trickier than it seems since you have to do it right for companies with semi-annuals)