According to CH registration, companies should only be used the Condensed List of SIC codes which each contain 5 digits (some with leading zeros). But I count almost 18,000 companies in the download csv files with only 4 digit SIC codes. Is there a reason for this?
Short answer: Registration with mandatory 5 digit SIC codes was only introduced mid 2016.
Longer, more speculative answer: Prior to 2016 (and the implementation of the The Small Business, Enterprise and Employment Act which also introduced PSCs and CS01s) you used to have an unvalidated free-text field for “nature of business” or SIC codes on formation.
Some of them might be a holdover from pre-digital times people tended to drop the leading zero from SIC codes where it had one (you see this done on Company Registration Numbers in a lot of older documents as well).
In some cases people would use the US versions of the codes, which were more readily available online (eg: LONGHAUL STEWARDS LIMITED overview - Find and update company information - GOV.UK). Unfortunately this means you can’t just add a leading 0 to all the 4 digit ones as this one, as an example, would be invalid.
Beyond that the history of SIC codes in the UK is pretty convoluted in general and I think adoption of 5 digit codes is actually relatively recent.
However, in theory, I think any legacy codes would have to be ammended when the company filed their first CS01, so my guess would be that those companies were all dissolved before 2017 or so?
^^^^^ After having a look at the XML for submitting CS01s this might not be the case as you only need to report changes in SIC codes, so while technically people should probably update them to valid 5 digit codes that’s probably not going to happen a lot of the time!
Thank you. Interesting.
I’d guessed that the 4 digit codes were from from 2007 registrations (but not US versions, LOL!) and on further manual searching, I can see that the companies are either dissolved, in adminstration, or just haven’t filed accounts for many years - so I guess stopped trading but did nothing about it WRT Companies House, or can’t bother filing accounts at all!
Was researchng something else and just noticed the constants enumeration file (https://github.com/companieshouse/api-enumerations/blob/master/constants.yml) has all the sic codes in it including a bunch of 4 digit ones. Might be useful.
Thank you. That is a really useful file.