There is a variables cepsd.s. When I run list:.s, this variable is not shown. I am trying to determine which is the actual variable in the SIPro screener and I am struggling. There is also another variable, dpsind.s that does show in list:.s but I am unsure of the SIPro variable.
Do you happen to know what the correct SIPro variables are to match the screen.
To find the SIP equivilents of any .s:
In GTR1 Miscellaneous Command line-only options: Enter list:.s
Then Select “Run Backtest”
The output will be a list of all SIP commands and their GTR1 mappings.
Filename Type Description Default Value Start Date End Date Last Modified
dpsind.s 115 “[SI Dividend - Indicated]” -999999 19970829 99991231 20230303
RAM, for dpsind.s does SI Dividend Indicated mean trailing 4Q of dividends, or four times the last dividend.
cepsd.s does not show when I run list:*.s which is strange, as the screen works using the variable. I think it means current earnings per share, but I am unsure of what the “d” means.
RAM, for dpsind.s does SI Dividend Indicated mean trailing 4Q of dividends, or four times the last dividend.
From SIP Field Definitions:
Dividend, indicated
Dividend, indicated
Data Table Name: DPS_IND
Data Category: Income Statement - Annual
Field Type: Dollars per share (0.00 to 9999.99)
Percent Rank: No
Industry/Sector Median: No
The indicated yearly dividend per share. The cumulative per-share dividend a company expects to pay over the next four quarters. Typically calculated by multiplying the latest per-share dividend paid by four.
and the top pick is ASC. From GTR1, cdv and dpsind.s are 0.45. Looking at the annual financial statement for ASC on SIpro, it shows no dividend was paid.
You mention Data Table Name: DPS_IND. I do not know how to access the data table. Is it through the file stockinvestorinstall_20230331 file?
You can get to DPS_IND several ways. From the SIP program itself you can choose a stock, select Fin-Ann’l and scroll down to Dividends per Share(Total) which shows the Current and list of previous dividends.
SIP stores the data in the si_isa.dbf (income statement annual) table which you is organized by a company_ID and in this case the DPS_IND is one of 644 fields in the isa table. Every SIP daily data set contains around 2000 fields for >7000 stocks > 250 MB of which last time I looked all our MI screens use less than 250 fields.
$0.45 dividend. Looking at the Fin-Ann’l section (which I showed above), dividend is listed as zero. If not too much of a bother, could you please check the dbf file and see if is $0.45.
Ardmore Shipping Initiates New Quarterly Dividend Policy, November 2, 2022
“Commencing with the Company’s performance for the quarter ending December 31, 2022, Ardmore intends to pay a quarterly cash dividend equivalent to one-third of Adjusted Earnings (as defined below) for the applicable quarter. The declaration and payment of dividends is subject to the discretion of our Board of Directors.”
Just to recap my understanding, using the SIPro screener DPSIND is either not available or not updated in a timely manner as all the dividend payments are listed as $0 for ASC, but it is in the database. In the SIPro screener “Multiples”, the quarterly yield is shown as 3% which matches the dividend payment of $0.45c which has an ex dividend date of Feb 27.
Can we add YEY to our weekly si pro picks? Also Jim, stated that he thinks that strategies that have more picks in them work better. i.e., buy the top 20 yey picks and drop at 25.
Apologies I’ve been traveling and so haven’t logged in recently.
cepsd.s is a derived variable from SI-Pro sources. The GTR1 runs a bunch of calculations to create afield - from other SI-Pro fields - to more-or-less try to match the original VL field. And so if you were able to match the GTR1’s calculation exactly it might be possible to replicate cepsd.s outside the GTR1.
For my own purposes in the SI-Pro screener I started using epsey0.s which is the [SI EPS Est Y0] field in SI-Pro. It’s EPS estimated for Year 0. It worked fine, but cepsd.s worked a little better overall, but could only be used within the GTR1 as I could not replicate the GTR1’s calculation well enough (Robbie posted the code but it relied on other derived fields and I took the simple, easy path at that point).
I just pulled up SI-Pro and loaded my YldEarnYear custom view & the ‘Dividend, indicated’ field shows $0.45 for ASC. So the information is there, but not shown in the backwards looking regular tables.
So to implement YldEarnYear in SI-Pro I created a custom ‘Screen’ & custom ‘View’. The ‘Screen’ did the initial easy screening & the view only supplied the fields I was interested in. I then exported the ‘View’ into a s/sheet every time I needed to generate the screen & did a little further calculation & sorting.
For cepsd.s, after much anguish and frustration, I have come close enough to Robbie’s cepsd.s to be able to implement the screen using SIPro.
There is a lag in the data used for the cepsd.s calculation, it appears to be 10 days, but may be 12, it is not more than 12. This lag is similar to what Valueline has which I think was what Robbie was trying to replicate. So using the screen picks from 3/10/2023 and the SIPro file from 2/28 I found the following:
Generally cepsd.s is the summation of the forward two qtr estimates (epseq0 and epseq1) plus the most recent two continuing operations quarters (epsconq0 and epsconq1).
If epseq0 is null, then sum the previous 4 quarters of continuing operations.
I could not find a stock with epseq0 not null and epseq1 null, so I can’t check that line of code in Robbie’s formula.
The remaining mystery to me is if the most recent reported quarter has old data. There were several stocks in the FEB 28 file (TNP for example) that had the most recent quarter reported as the end of September (most stocks had Dec 31). In this case, cepsd.s summed the forward two quarter estimates, plus the most two recent reported quarter earnings and subtracts the third most recent quarter. I chose to ignore this line and the CAGR increased slightly.
Using items 1 and 2 from above, and not lagging the information, I was able to get a 0.2% increase in CAGR from cepsd.s. So not exact, but comparable.
