I have had a new paper, co-authored with Prof Adam Leaver, published by the University of Sheffield Centre for Research into Accounting and Finance in Context (CRAFiC). The paper is available here. The paper looks at the phenomena that we describe as ‘hollowed out firms'. These are companies where it would seem that financial engineering has a higher priority than actually generating profits for payment to shareholders.

One of the characteristics of these groups of companies that we have identified is an apparent ability to pay dividends in excess of the level of profits reported in the accounts of the group as a whole. This is possible because the capacity to pay dividends is not, bizarrely, based on the group accounts, but is instead based upon the accounts of the group parent company, considered in isolation. This means that the profit available for distribution depends upon the ability of group subsidiary companies to distribute realised profits that they have made, calculated in isolation, to that group parent company without having to take into account the losses of other group subsidiaries that might otherwise dilute those distributable profits.

A group of academics I work with have spent some time considering how this situation might be exacerbated by deliberate financial engineering, facilitated by the ability of many groups to borrow based on the security that they can offer rather than the profits that they declare. Much of this consideration previously related to various issues such as mark-to-market accounting; the realisation of profits in internal sales of assets, and similar related issues. However, in this new paper Adam Leaver and I present another possibility. This is that the technique of transfer mispricing, which uses artificial prices to record transactions between companies under common control, might be used to generate such profits by incorrectly apportioning profit between the parties to a transaction.

This technique of transfer mispricing has been subject to considerable attention in recent years because of its tax consequences. In that context I have been looking at it for almost two decades. It motivated my work on country-by-country reporting. The critical point to note is that although the artificial allocation of profit can be adjusted by most tax authorities for tax purposes there is no reason why the accounts must be adjusted to reflect that fact. As a result it is our suggestion that it is entirely possible for a group to use transfer mispricing to generate profits for distribution in some subsidiary companies that might, in turn, lead to losses in others that might, if the group was carefully structured, be ignored for the purposes of calculating distributable profits for the purposes of payment to shareholders of the group, whilst allowing the losses to be ignored. In other words, a group parent company might appear to have distributable profits available to it that have not really been earned within the group as a whole and which are, instead, the result of financial engineering using transfer mispricing.

To illustrate the consequences we have in our paper constructed an admittedly exaggerated example as to how distributable profits might be created in this way. We used this research methodology because data on intra-group transactions is almost entirely hidden from view in the accounts of multinational companies. That is because accounting standards deem such transactions to be of no interest to shareholders when our work suggests that the whole current edifice of many shareholders' rewards in many companies might actually be based on their abuse. For precisely this reason, and because we have no budget to take this further right now, we have had to rely on this particular research technique. We do, however, think it appropriate in this circumstance.

We suggest, however, that the implications of this research finding are significant. Firstly, and to state the obvious, we accept that more research is required. We think that important. If we could show this simulated finding reflects reality then it reveals a significant flaw in our current accounting framework that is being exploited to permit financial engineering that is in itself deeply dangerous to the companies that might be using this technique, and to those who rely on them for an income stream.

We made the following recommendations in the paper to address the potentially serious risks that might well be widely prevalent inside the current structure of globalised business:

In the first instance, and most obviously, a simple change to the rules determining the extent of distributable profits within group parent companies is required, so that these may not, whatever other accounting might imply, be used for the purposes of determining legal dividend and share buyback distributions that exceed the retained reserves of the firm as a whole as evidenced by its consolidated group accounts.

Second, for the purposes of ensuring that this requirement is complied with it must be required that disclosure be made of the retained reserves of the firm split between those that are realised, and therefore available for distribution to the members of the parent company by way of either divided or share buyback, and those that might not be realised e.g. because they arise as a result of fair value asset revaluations. Critically, any reserve that requires imputation of future valuation in its calculation should not be distributable.

Third, so that the reserves of the parent company are properly understood it is important that the accounts of the parent company acting in its own right (published as a part of, but distinct from the consolidated accounts of the firm as a whole for which it is responsible) must be always be made available, in full. In addition, it is important that disclosure of the name, place of incorporation, place of trading and trading activity of each of the subsidiaries of the group (without exception) be made available on the group website together with an organisation chart so that the manner in which the group parent company might get to access the retained reserves of subsidiary companies might be understood.

Fourth, the directors of the firm should be required to state what their dividend policy might be, and how it might be considered sustainable given the constraints now imposed by the above suggestions. This review should make explicit the assumptions inherent within it, including on future group funding and the reasons why that policy has been adopted.

Fifth, it should be a requirement that the group's auditors to explicitly review and report on that statement, making clear whether they consider it, and the assumptions underpinning it, sustainable given the overall circumstances of the group.

Sixth, the auditors should be explicitly required to note that if the retained realised reserves of the parent company exceed those of the group as shown by the consolidated balance sheet then that excess if not distributable.

Finally, it should be required that disclosure be made of all the transfer pricing adjustments required of the group in the period to which the accounts relate with regard to past financial statements, with their cause being given, so that the risk of misstatement of current realisable reserves can be appraised based on past evidence supplied by third party tax authorities.

Both Adam Leaver and I would stress that a simulation cannot prove that a situation exists. But that there are hollowed out firms is a fact. And that they exist is in no small part because of their ability to distribute profits that prime facie it appears that they have not earned when calculated on a group basis.

What we raise then are questions about how accounting, accounting standards, groups and those dependent upon them have created structures that might let this happen, and have simultaneously then ensured that those structures can quite legitimately remain hidden from view. It is hard to believe that there is no reason for that opacity, and that even if there was not, that no attempt to address it has been made. The time for transparency has, we suggest, arrived.

This issue matters enormously. I have never seen this suggestion that transfer mispricing has been used for this purpose made before. If it has been, I would be pleased to hear of it. But whether this is genuinely original research (and I think it is) or not is not the point. What matters is that if any companies are using such arrangements then the fact that this might happen should be if massive concern. This is, quite literally a matter of taking returns now that should not be paid, and which may well belong to the future. This undermines the whole reporting edifice on which modern financial capitalism is built. And that does, because of its significance, potentially undermine our society. I don't want to overstate this issue, but it's hard not to. This matter needs serious discussion and investigation. The credibility of financial markets depends upon it.