I received an email from a new user at one of our client schools. They were asking about the checkbox control in the CIK filter option. This is an amazing feature. We were the first to offer CIK filtering, the first to offer CIK/date filtering and I think still the only search application that has this Historical CIK feature though we have had it for almost 8 years. It was hard to implement but I think it important. Let’s take a look at it. The check-box can be seen in the image below.

If you are running a search and are filtering on CIK or CIK/DATE this button reviews your CIK and or CIK/DATE file and looks to see if you have any CIKs that we have identified as a successor or predecessor to another filer. If so then your file is augmented with the successor or predecessor (whichever is missing).

To illustrate the value – suppose you intend to identify 8-K filings made in the 2016-2020 for your research. The firms in your sample were identified using some other data source and you have a list of their CIKs. For the purposes of this demo I have four (4) CIKs listed as you can see in the image below.

Oh-oh – I had no results at all.

If I rerun the search and this time check the Include Historical CIKs button the results will be different. Here is a screenshot after submitting the same search but selecting the Include Historical CIKs.

Notice that the application reports that 8 CIK(s) were submitted. But we only had 4 in our list. That is because of the mapping process that occurs when you start the search. The application reads your list of CIKs and checks for the existence of related CIKs. We add additional rows to the filtering list based on our mapping. In the case of CIK/Date filtering for each CIK-MATCHDATE pair you have included if the CIK has one or more related CIKs we augment the file with a new row with the additional CIK-MATCHDATE pairs.

The next question of course, how to see the mapping that was made. There are two ways to do this. If you have a large number of CIKs it might be easiest to review the CIK_MAPPING database that is accessible from the Query Databases viewer. Just select the database from the Query List and hit execute to access the entire list. Scroll through the list to find your CIKs and identify the values for RELATED~

In the next image I pointed arrows to the CIK_MAPPING database and the Execute button. I then boxed around the BASE_CIK for 912752 and 1971213 – you can see that we are asserting that one of these is the successor entity to the other.

The second way to identify the augmented list is to access the filter file that was generated by your selection of CIKs. If you click on the Set CIK File button after you have completed a search the Select Existing CIK File box will be populated with the name of the file that was actually used in the search. Even if you used your own file, the application will read it and then store a copy (augmented or not) in the D:\PhotonUser\AppData\Local\Temp\somenumber directory. Note this is a session directory not a permanent directory.

While it is a bit painful to navigate to that directory you can and open the file to examine it. One of the problems though is it does not map the relationship as you can see below. Further, you can’t predict the relationship based on order.

One of the sanity checks we perform when we make these links it we compare the last balance sheet in the last 10-K for the predecessor to the same balance sheet in the first 10-K in the successor. In this case the issues are even more complicated because CIK 912725 has a continued filing obligation because of some debit securities. Nonetheless, based on my review of all of the facts available, I believe the time series of data related to CIK 1971213 (Sinclair Inc) extends to the filings made by the entity that has the current name Sinclair Broadcast Group LLC with the CIK 912752.

Names are not always predictors of the predecessor-successor relationships. For example, CIK 1999001 Six Flags Entertainment Corporation/NEW is mapped as the successor to CIK 811532 (CEDAR FAIR L P) despite the fact that the name of the filer with the CIK 701374 is Six Flags Entertainment Corp/OLD.

The last 10-K filed by CIK 701374 for the year ended 12/31/2023 reported total assets of 2,711,474,000. The first 10-K filed by CIK 1999001 for the year ended 12/31/2024 listed comparative values for 12/31/2023 with total assets of 2,240,533,000. While these don’t match, CEDAR FAIR’s last 10-K reports total assets of 2,240,533,000 for the fiscal year ended 12/31/2023 which matches the amount reported by Six Flags Entertainment Corporation/NEW.

We created this because early on clients and potential clients would lament the fact that they were losing observations – the most common example they would cite was Oracle (777676 -> 1341439). These are often hard to identify. The best check is when we can compare the balance sheets. However, knowing that these events even happen is hard. These entities do have some reporting obligations but my estimate is that only about 60% of the filers do everything correct. The good thing is that they are relatively rare.

Last note, we don’t automatically update your mapping dictionary. You have to update it. To do so from the File menu on the application select Options. From Options select Update Historical CIK and then press the Perform Update button.

I cannot make a guarantee that we have identified all mappings. The next time I get some bandwidth I am going to see if I can use our compensation databases to identify others. My hypothesis is that if I have a filer that has executive (or director) compensation in 2023 and none in 2024 and then another that has compensation in 2024 but not 2023 and there is overlap in the PERSON-CIKs of the executives then we should research those to see if the 2024 filer is the successor. However overlap alone will not be sufficient. There are a number of executives and directors that had reporting obligations because of their roles at World Wrestling Entertainment Inc (WWE) and have similar reporting obligations with TKO Group Holdings, Inc but TKO acquired WWE, it is not a successor entity.

As an aside – the reason I selected the 2016-2020 span is that none of the successor filers had filings in that span. Both the successor and predecessors had filings in the 2021-2024 span.

I was contacted by a research assistant this morning who was trying to collect some interesting data. They had a list of CIKs and with a date variable. What they wanted was to determine/identify any 8-K filed by those companies in a ten day window around the date variable. I was going to point them to our CIK/Date Filter capability (CIK_DATE_FILTER) except I noticed that their list covered the entirety of EDGAR. Further, we have not finished standardizing the metadata at the document level and they wanted the SEC FILING date as well as the acceptance date time. The first constraint meant that the post-query filtering was going to be a drag – much better to search one index at a time. The second issue meant that the inconsistent metadata would mean that some of their time would be wasted.

I ended up writing two bits of code, the first took in their original list and added additional rows and a new field. The code checked each CIK and determined whether or not the CIK was in our CIK_MAPPING database. This is where we try to archive CIK changes. The student had CIK 1532063 with an event date of 6/30/1996. The problem was that CIK 1532063 is the successor to CIK 885721. CIK 1532063 made their first filing 10/6/2011. Therefore without augmenting their CIK list to add an entry for CIK 885721 with the date 6/30/1996 the entry for 1532063 for 6/30/1996 would be missing.

The first bit of code is called augment_cik_data.py and it takes an input csv file with a column labeled ‘cik’ and ‘last_date’ and compares the CIK values to determine if they are in the CIK_MAPPING database. If so, the entry is duplicated with the value from the RELATED_CIK_1 field in the database as the value in the ‘cik’ column and the original value for ‘cik’ added to a new column named ‘BASE_CIK’. Below is a screenshot of the csv file. The entries in red text are the augmented entries.

I then used this file as the input to a second script named pull_matching_from_8K_meta.py. This reads the newly augmented csv file and then finds matching entries based on the CIK date criteria.

To run the first script – log into the platform and upload a csv file with your sample. To follow this example exactly you need a column labeled ‘cik’ and a column labeled ‘last_date’. The cik needs to be an integer and the last_date values need to be in the form YYYYMMDD. from the Catalog icon at the top of the frame select IDLE – see the image below.

When Python opens from the File menu select Open as illustrated in the next screenshot:

Navigate to the platform share and the folder PythonCode to select the augment_cik_data.py script.

When the file opens you will need to modify the input and output paths as illustrated in the next image.

You need to save the file. However, you don’t have write privileges on the share so you will want to save to the same folder as your csv file. Once you have saved it with the edited paths (and other edits you might make) use the Run Module command to run the file.

When the code finishes (very quick) there will be a completion message in Idle. Now that you have supplemented the original sample with any matched CIKs you want to use that output file as the input to pull_matching_from_8K_meta.py. Follow the same steps as above to find the file on the Share, modify the path and save it to your Temporary Files folder and then run it.

The first time I ran this it took about six minutes so I added an index for SEC_FILING_DATE and it took less than a minute.

The Wall Street Journal just reported that President Trump is considering eliminating quarterly reporting for US issuers (WSJ Article). Wow! It is an interesting argument – will companies somehow perform better if they were not having to report earnings four times a year? What about investor uncertainty? I will try later to post links to some academic research that has established that information in 10-Qs has significant price implications. I spent much of the weekend trying to sort out how we can quickly get the new data that I expect to see when companies start filing under ASU-2024-03 but frankly I was wondering in the back of my mind if the SEC would somehow lessen the impact of that standard. I did not imagine this.