Associated companies: Making sense of the control tests By Nichola Ross Martin (2024)

Accountingweb Tax Editor Nichola Ross Martin takes issue with some of the "extreme views" of “control” in circulation in respect of associated companies and finds problems with some basic Revenue interpretation post-Newfields

For the purposes of small companies relief under s 13 ICTA 1988, a company is an associated company of another at a given time if at that time:

• one of the companies has control of the other, or
• both of the companies are under the control of the same person or persons

.
“Control” is construed in accordance with s 416 “A person shall be taken to have control of a company if he exercises or is able to exercise, or is entitled to acquire, direct or indirect control over the company’s affairs”.

S.416(2) lists the various tests for control, which are broadly by entitlement of an individual:

• to acquire the greater part of the share capital, or voting power
• to receive the greater part of distributions to shareholders
• to receive the greater part of assets on winding up (this tests means that one looks also at the entitlement of loan creditors.

S.416(3) says that where two or more persons satisfy any of the conditions above, they shall be taken to have control of a company.

In order to determine who is entitled to what, the rights of a person’s nominee are taken into consideration - s 416(5) and the rights of a person’s associates may also be attributed to a person to establish control - s 416(6).

A person’s associates are

• Husband and wife, including separated spouses, but not divorced spouses, including civil partners
• Parents, grand parents and remoter forebear
• Child or grand child or remoter issue
• Brother or sister, including half siblings, but not step’ nor aunts or uncles or cousins.
• Partner (as in a business partnership)
• Settlements and will trust associates
• Trustees are associates where the individual, or any living or dead relative is or was the settlor,
• Where the individual is interested in a settlement, then beneficiaries, remainder men and trustees are associates.

Conventional wisdom
The Inland Revenue, as they were then known issued a press release back in 1983 to confirm that two companies are under the control of the same person or persons only if an irreducible group of persons having control of a company is identical with an irreducible group of persons having control of the second company. An irreducible group of persons is one which has control but would not have if any one of them were excluded.

The example frequently used to illustrate this interpretation is as follows:

A and B can control X Ltd, but B with 51% of voting power controls the company on his own – he is the irreducible group of persons.
B does not control Y Ltd on his own and so the two are not associated. The irreducible group who controls Y Ltd is A and B.

The concept of the irreducible group seems reasonable, control achieved outright by one person or more than one person acting together to achieve it.

Attribution of associates may be necessary to try to establish the controlling party, as follows:

The issued share capital of Cat Ltd comprises of 100 £1 Ordinary shares, held as follows:

Dog Ltd is owned 60% by C and 40% by an unrelated individual.

C is deemed to control Cat Ltd, because the rights in the shares of Dog Ltd are attributed to him and he therefore controls Cat Ltd outright. You do not calculate his control in Dog Ltd here as being 60% of 30%, you attribute the entire 30%, so C is deemed to control 65% of Cat Ltd.
A & B are associates, being husband and wife, although they cannot achieve control even if your attribute each other’s shares to give one 35%.

Newfields
The leading tax case on the topic of associated companies is R v CIR ex parte Newfield Developments Ltd,, which went all the way to the House of Lords. Mrs W was a beneficiary of a trust created by her deceased husband, which meant that its trustees were her associates, and he was the settlor of another trust, and she his wife, the trustees of that trust were also her associates. Each set of trustees controlled a company and so HMRC attributed their share ownership rights to Mrs W, and claimed that the two companies were therefore associates per s.13.

The case explored the wording of s.416(6) and whether HMRC had a discretion not to attribute the rights of associates as Mrs W appeared “to control” a company when in reality she had no influence in the running of either company. The court found that irrespective of the logic, HMRC may attribute the rights of associates if that is necessary to create a controlling party to satisfy s.416. Adding together all the shares of Mrs W’s associates allocated to her the greater number of shares and it was deemed that she “controlled” the companies.

Problem 1
In HMRC’s Company Tax Manual at CTM03730 is an example which takes “control” a stage further (I have changed shareholdings to percentage shareholdings for comparison). It goes as follows:

A, B, C and D are not associates, so HMRC say that CB Ltd is controlled by A and B together, or A and C together, or B and C together, that is by any two of them together.

They say too that AA Ltd is controlled by A, B and C together, or A, B and D together, or A, C and D together, or B, C and D together, that is by any three of them together.

Fair enough, everyone understands that rule, but if you stand back, does it make any sense? S.416 provides a number of different tests so that control can be achieved by the person or persons as a result of looking at the results of the three basic tests, who has the greater voting power, and or share ownership and or rights on winding up?
If that cannot be found, then (6) allows attribution (if it is needed), perhaps as a tie breaker. In the example above, all the groups have equal voting power or share ownership and I cannot see how it can be claimed that any of the combinations have “greater” rights than another. Indeed, in a company where shares are owned equally by non associated individuals, no one sub-group can “outcontrol” another, there exists a stalemate; only A, B & C as a group control CB Ltd, and A,B,C & D control AA Ltd.

Likewise in two man company, where shares are held 50:50 then both control per s 416.

This “problem” tends to be avoided in CTM60210 where HMRC detail tests for control and definition, only appearing in one example. Close company examples all tend to feature companies which are controlled outright according to one test or another.

The associated company rules “cast the net very wide”, but what is actually meant by that comment “very wide”, which comes from R v CIR ex parte Newfield Developments Ltd, is the fact that s.416(2) lists various mutually exclusive tests so that you may have different groups of persons controlling a company; one group according to votes, another according to rights or one according to rights on winding up. HMRC (and writers like me, I hasten to add) seem to have decided over the years since Newfields that each test can also have a range of outcomes which means that many different people may control for each test, this is what I refer to as “the extreme view” of control.

Problem 2: The “extreme view” and more irreducible questions
The “extreme view” seems to be that there can also be mutually exclusive outcomes for each control test, so that you could have in theory any number of people who can control a company by shareholdings, not necessarily the person with the most shares, or the smallest group of people. It has evolved from the judgments in Newfields, and I have doubts about its practical application as like so many of these things in tax, it appeals to lecturers and writers (it is interesting to play around with the rules), and it could be used to bamboozle small companies if HMRC tried it on. Nevertheless, it is worth considering as it takes the whole concept of “control” into a totally new ball park.

It is worth pointing out that lecturer Mark Morton discusses it in an article in Taxation magazine in May 2005 saying that he heard the arguments from HMRC’s head office, having never seen then at District level. It is interesting to note that very little seems to have been heard about this since, beyond the odd lecture note and this is probably because the “extreme view” only hits a very limited number of companies, that is if it really affects them at all.

The decision in Newfields established that HMRC may attribute the rights of a person’s associates if it so wants to, or not. This has rather a curious effect on the concept of the irreducible group and in so doing means that if correct HMRC’s examples of control in the CTM need revision and should stop being linked to the close company examples which illustrate s.416, if any sense is to be made of the conundrum.

It is easiest to illustrate the “extreme” view by way of another illustration (this example comes Mark’s article):

Who controls Y Ltd? It could be, according to HMRC:

• Either Mr A and Mr B – 51% or
• Mr A and Mrs A – 51% or
• Mr A (attributing his wife’s rights) – 51% or
• Mrs A (attributing her husband’s rights) – 51%.

The alternative view would be to discard surplus persons and create an irreducible group, which in each case can only be done by attribution rights so that only one person can control this would mean that either Mr A or Mrs A could control with the total of 51%. This to me is the most sensible outcome, as we avoid the impasse of having so many controlling combinations. HMRC's answer creates stalemates if no one has a greater shareholding than another group, how can any one group control?

In passing, we go back to the situation that if Mr A or Mrs A controls the company then one could say that no one person can control it outright, but I think that in the case of attribution (which has its own subsection) that the intention is that it is used to decide control and so Mr A and Mrs A may be held to be mutually exclusively in control as rights have been attributed.

A further “extreme” example given by Mark in his article is as follows:

He suggests that HMRC might take any combination to create control in this case, but I don’t agree. Looking at s.416(2) we are after a “greater part” of voting power or share capital, and so I suggest that the “extreme” argument does not work here. Mr A and Mr B combined (without attributing rights) have only 51% of voting power, whereas Mr (attributing Mrs A) has 98%.

.Conclusion
“Control” in s.416 is not about control in a “real world sense” but this claimed only due to the effects of attribution. In order to establish the basic form of control a person or group of persons must be capable of exerting greater powers than the next person or group. If they cannot there surely exists a stalemate and no party can be claimed to control the other.
The Revenue’s “extreme” view, if it still holds it, is that s.13 looks at any controlling combination that can apply to two companies, and so in passing, by not attributing, if necessary, one ignores any other combination who actually controls each company individually. I can’t see this, as for starters s.13 doesn’t say anything about a new order of control it just says that control is construed according to s.416.

Looking back to the rationale behind s.13 it is in place to stop people artificially splitting their business, it seems that if HMRC ever did follow up the “extreme view” they would have made a huge leap in using Newfields to create artificial groups who do not and cannot control in accordance with even the basic tests in s 416.

In the words of Peter Gibson L.J. I think that my solution here “is the least worst solution of a problem created by a mode of drafting which rewards readers who desire clarity by offering them riddles”. These rules are crying out for a re-write.