Yes, there are economic reasons for this. I can give a few, but it really is impossible to answer completely.
The ultimate aim can be unrelated to the IP at all, but simply for tax avoidance.
Broadly, companies are taxed on their profits. Expenses reduce profits, and therefore reduce the tax due. But if the expense comes from within the corporate group, the corporate group is not worse off as a whole: it is just shuffling money around.
So, say a group has two companies, Company A (in a high tax jurisdiction) and Company B (in a low tax jurisdiction). All the IP is transferred to Company B. Company A pays a licence fee to company B for use of the IP. This reduces the taxable profits for company A, so company A pays less tax to the high tax jurisidiction. The licence fee is income for Company B, so it may be taxed in company B's jurisdiction, but at a much lower rate. But the group as a whole is still better off.
This is the basic concept, but in reality it will be much more complex.
Competitors may want to know your IP portfolio, to work out what your future plans are. This is because patent applications are published at 18 months from filing, whereas the official announcement may still be years down the line. So by transferring the IP to one or more subsidiaries which are not obviously linked to the parent company, it can make the job much more difficult for competitors than a simple name search.
Each subsidiary may be generally standalone. So it makes sense for the IP to be held by that subsidiary, rather than the parent company. This is so that if the subsidiary is ever spun out or sold off, the IP goes with it. This is quite a simple reason, and is generally for ease of administration than anything else.