Home > Publications > 2022 > Dividend integrity and personal services income attribution – a Government discussion document > Chapter 3 – Proposal to tax a deemed dividend portion of proceeds from selling shares

Dividend integrity and personal services income attribution

A Government discussion document

Hon Grant Robertson
Minister of Finance

Hon David Parker
Minister of Revenue

March 2022

Chapter 3 – Proposal to tax a deemed dividend portion of proceeds from selling shares

• General scope of a rule
  • A. Shareholder sells shares of a controlled company but retains economic ownership of the company
  • B. Shareholder sells shares of a controlled company to an unrelated company
  • C. Shareholder sells shares of a controlled company to an unrelated individual
  • To which scenarios should a generic recharacterisation rule apply?
• Details of a rule
  • What share sales should the rule apply to?
    • Type of shareholder
    • Type of company
    • Shareholding size and control criteria
    • Look back rule
  • How much of the sale price should be taxable?
  • Consequences to seller
  • Consequences to the company
  • Consequences to the buyer
• Other issues
  • Should the amount of the deemed dividend be limited to the gain on the sale?
  • Why apply the rule to sales by companies?
  • Corporate groups
  • Sales of shares in a controlled company to another company controlled by the same shareholder
• Questions for submitters

General scope of a rule

3.1 The objective approach proposed in this chapter involves recharacterising share sales as dividends in the hands of the selling shareholder. A fundamental question is which transactions should be subject to the proposed recharacterisation rule.

A. Shareholder sells shares of a controlled company but retains economic ownership of the company

3.2 The clearest case of dividend stripping arises when a shareholder sells shares in a company it controls or owns for cash (or cash equivalent) but retains control or ownership. This is because, like a straightforward dividend, the shareholder has realised cash from selling the company but still owns or controls the company, as is illustrated in Example 1.

3.3 The type of transaction illustrated in Example 1 is the clearest case of converting a dividend into a capital receipt, as the shareholder has both received cash and retained the company, as it would have if Bullseye had paid a dividend. However, the shareholder is better off because it did not have to pay the “top-up” tax or use the imputation credits. In this case, current law will already treat the transaction as a dividend stripping transaction, and $72 of the sale proceeds should be taxable as a dividend. However, such transactions may not always be self-assessed and may also not be known or pursued by Inland Revenue.

B. Shareholder sells shares of a controlled company to an unrelated company

3.4 Another dividend stripping scenario, Scenario B, is the essentially the same as Scenario A except:

• An unrelated person seeks to acquire all the shares in the target company.
• The unrelated person forms a new company to acquire all the shares in the target company.
• The unrelated person contributes an amount of capital to the purchaser company that equals the amount the purchaser company will pay the shareholder of the target company to acquire all the shares in the target company.

3.5 The overall result obtained in Example 2 is that the shareholder of Bullseye has received the $72 value of Bullseye’s retained earnings as a capital receipt and has not paid the top-up tax on it. In addition, even though Bullseye’s imputation credits have been eliminated, it can distribute its $72 of retained earnings to Purchaser and eventually to Purchaser’s shareholder (via a share repurchase or a liquidation) without any additional tax being paid.

C. Shareholder sells shares of a controlled company to an unrelated individual

3.6 his scenario is the same as Scenario B, except the shares of the target company are purchased by an unrelated individual instead of a company.

3.7 Scenario C is not as favourable to the new shareholder as Scenario B. In Example 3, the new shareholder has paid $172 for Bullseye, but the ASC of Bullseye is still $100. In addition, Bullseye has lost its $28 of imputation credits. If the $72 of retained earnings of Bullseye are distributed, they would be taxed as an unimputed dividend. This would also apply if Bullseye were liquidated.

3.8 In practice, the sale is more likely to be carried out as described in Example 2 (Scenario B), in order to realise the tax benefit of that approach. On the other hand, if Purchaser acquired less than 100 percent of the shares in Bullseye, carrying out the transaction in the manner described in Example 2 would not protect Purchaser from tax on a distribution from Bullseye since the inter-corporate dividend exemption would not apply.

3.9 Alternatively, to avoid an outcome which benefits the seller but hurts the purchaser, the parties to a sale could agree, for example, that the target company would distribute a dividend before the sale (the shareholder of the target company would pay the 11% top up tax), and the purchaser would pay less for the company to reflect the fact that its earnings had been distributed. A second option is for a dividend to be declared and reinvested in the company – this means that the retained earnings are converted into additional ASC, an amount distributable to the purchaser tax free on a share repurchase or on liquidation. Another option would be to negotiate a purchase price that reflects that the exiting shareholder of the target company gets a benefit in not having to pay the top-up tax on a dividend, and the purchaser will have to pay a penalty in the future for the possibility of deriving an unimputed dividend. If the purchaser thinks this possibility is far enough in the future they may be able to agree on a price that makes them both better off.

To which scenarios should a generic recharacterisation rule apply?

3.10 Of the possible scenarios for a rule to recharacterise a portion of the receipts from selling shares as a dividend, A is the narrowest in scope and C is the broadest in scope.

3.11 Scenario A is the one that most clearly looks like a transaction that results in the same outcome as if a dividend was paid. The shareholder receives value for the retained earnings in the company in the form of cash or debt or some other consideration, and the shareholder retains economic ownership of the company (although the legal structure of the ownership changes). There are also significant tax advantages to this as not only are the earnings made available to the shareholder without any top-up tax being paid, the full imputation credits remain in the company and are available to be used later.

3.12 This sort of related party transaction is one where Inland Revenue often uses section BG 1 or GB 1 to treat the sale as a dividend stripping arrangement. This is also similar to a United States rule that treats part of the consideration for a controlled company’s shares as potentially a dividend from both companies if the shares are sold to another company that is also controlled by the selling shareholder.[14]

3.13 Scenario B is also a way the selling shareholder can receive the value of the target company’s retained earnings in the form of a non-taxable capital receipt. The buyer obtains no great advantage but is also not disadvantaged, except to the extent it loses imputation credits so that it cannot distribute pre-acquisition earnings without the distribution being taxed as a dividend unless it is done as a share cancellation or liquidation (although it can pay dividends out of post-acquisition earnings which generate imputation credits). Applying the recharacterisation rule to this type of sale would expand the concept of recharacterisation beyond what is now thought of as “dividend stripping” subject to section GB 1.

3.14 Scenario C has the broadest application. While it provides the selling shareholder the same benefit as Scenarios A and B, it appears to disadvantage the purchaser so it is more likely the purchaser would use a structure, such as in Scenario B. However, the tax advantages in Scenario B are only available if all the shares in the company are being sold. In Scenario C, the purchaser may also seek to have the seller pay itself a dividend before the sale, with the sale price being adjusted for that. If this were to happen the dividend would be taxed, so there would be no dividend recharacterisation required. However, it is also possible the parties could try to negotiate a purchase price which would give them both an advantage resulting from the seller not realising a taxable dividend.

3.15 It is proposed that a dividend recharacterisation rule may be applied to Scenarios A, B and C. All scenarios have the same consequence for the seller, although they have different consequences for the buyer. It seems appropriate that the shareholder who owned shares in a company when the company earned income would be taxed on the income when shares are sold for a price that includes the value of the company’s retained earnings. Failure to tax the retained earnings component of the sale price would also allow deferral of the top-up tax to be extended for a potentially lengthy period.

3.16 Such a recharacterisation rule could encourage the parties to the sale to agree to terms such as the company paying a dividend to the selling shareholder before the sale, in exchange for the buyer paying a lower price for the shares. However, such a practice would produce an appropriate result from a tax perspective.

3.17 The remainder of this chapter will discuss how a recharacterisation rule could be designed, keeping in mind that it should apply to all three scenarios.

Details of a rule

3.18 The following criteria should all be considered in establishing a deemed dividend on share sale rule.

• To what sales of shares should the rule apply?
• How is the deemed dividend amount determined?
• What other consequential implications flow from the rule?

What share sales should the rule apply to?

Type of shareholder

3.19 The proposal will only apply to shareholders who are New Zealand resident natural persons (including natural persons who recognise the income as beneficiary income), trustees, and companies. The proposal will not apply to sales by shareholders, other than companies, that are taxed at 28%, such as PIEs, superannuation schemes and group investment funds. The reason for applying this rule to shareholders that are companies is discussed under other issues.

3.20 If the shareholder is a partnership, including a limited partnership, this rule will apply as if the partners directly owned and sold the shares in the company.

Type of company

3.21 Because the rule will require reference to the imputation credit account (ICA) balance to determine the deemed dividend amount, the Government proposes that this rule would apply only to sales of shares in companies that maintain an ICA. This would be New Zealand resident companies, and Australian resident companies that elect to maintain an ICA.

Shareholding size and control criteria

3.22 The proposed recharacterisation rule would only apply when shares in a company are sold by the controlling shareholder. An important point to bear in mind is that a shareholder who owns more than 50 percent of the voting interests in a company controls that company. The Government does not propose that a recharacterisation rule should apply to sales of shares in listed companies and sales by portfolio shareholders. This is because the shareholder may not have sufficient information about the tax attributes of the company in order to determine how much of the sale price should be recharacterised as a dividend (these will be discussed later in this chapter). Also, listed companies tend to have a high dividend payout rate anyway. Companies with low dividend payout rates tend to be closely-held companies.

3.23 The proceeds from a share sale would be recharacterised as a dividend if the shareholder (together with associates and other shareholders acting together) controls the company immediately before the sale. It should not matter how large the block of shares sold is, as long as the control criterion is met. This is so there is no ability to avoid recharacterisation by selling shares in small “drip feed” parcels.

Look back rule

3.24 A specific anti-avoidance rule is also proposed, which is intended to capture someone selling first a controlling interest in a company (which would be subject to this rule), followed by selling more shares to the same person (or an associate) after they have divested control. To cover this, it is proposed that the rule would also cover share sales when all the following apply:

• The seller (together with associates) did not control the company immediately before the sale.
• The sale is made within two years of a previous sale of the company’s shares by the seller, or an associate of the seller, to the same buyer (or an associate of that buyer).
• The control criterion applied at that time to the earlier sale of shares (that is, the company was controlled by the seller, together with associates, immediately before the earlier sale).

How much of the sale price should be taxable?

3.25 When shares of a company are sold, a number of different factors affect the value of the shares. These include the value of the underlying assets, including goodwill or capitalised expected future earnings. The value of shares in an operating company may be the higher of capitalised future earnings or the value of the company’s net assets (liquidation value). Assets of a company can be funded from a number of financing sources including paid-in capital, company debt, and retained earnings. Retained earnings may include taxable income, capital gains, and other forms of income that are not included in taxable income.

3.26 The tax treatment that applies on the liquidation of a company is instructive for determining how much of a share sale receipt should be treated as a dividend. When a company is liquidated, the shareholders dispose of their shares for the assets of the company (which could be cash if the company has already sold its assets). When a shareholder sells shares to another person, the shareholder disposes of their shares for, usually, a combination of cash and debt.

3.27 In a liquidation, the amount that is a dividend is determined as a residual amount, meaning that all amounts that are not a dividend are determined first and the rest is treated as a dividend. The following are subtracted from the value of the company’s property that was distributed in liquidation.

• Available subscribed capital (ASC).
• The portion of retained earnings that are from realised capital gains.
• The value of property that represents unrealised capital gains.
• The difference between the (accounting) retained earnings derived from foreign portfolio shares subject to the fair dividend rate (FDR) regime and historic undistributed FDR income.

3.28 What remains is deemed to be a dividend. Imputation credits may be available to use against tax on this income.

3.29 Putting it another way, the dividend amount is taxable income plus income (other than capital gains and the FDR adjustment) that is not included in taxable income. In other words, it includes a clawback of preferences (other than capital gains) as does an operating dividend. The clawback of preferences is usually the amount of the dividend that is not fully imputed. With capital gains and an FDR adjustment specifically excluded from this amount in a liquidation, the potential clawback of preferences may not be significant. It may include things such as dividends from non-portfolio shares in a foreign company.

3.30 The Government does not propose duplicating the liquidation calculation for determining the dividend amount from the sale of shares. This is too complex and uncertain. Determining the unrealised capital gain component if shares of an operating business are sold would be especially difficult, as there would likely be a significant goodwill component which could have increased in value over time.

3.31 Using an accounting concept of retained earnings is possible. This could be adjusted to remove capital gains. Accounting earnings could potentially also capture more earnings than taxable income and so pick up a clawback of preferences amount. However, a disadvantage is accounting standards are more flexible than tax rules, so the standards could vary by taxpayer. However, the degree of variability is still limited by the application of accounting standards, such as International Financial Reporting Standards (IFRS).

3.32 Another option is to refer to the company’s ICA balance at the time of the sale in order to calculate the amount of undistributed taxable income that it represents. For example, a $28 ICA balance could be deemed to represent a $72 net dividend or a $100 gross dividend implicit in the price paid for the shares.

3.33 We recognise that this is not a perfect measure. The ICA balance is broadly tax paid less tax refunds and imputation credits distributed. Typically, imputation credits are generated when provisional tax payments are made, and these could be a little higher than the actual tax liability if the uplift method is used to ensure the taxpayer is not charged use of money interest before the final instalment. On the other hand, the payments could be lower than the actual liability. Also, some adjustments, such as the shareholding change debit, could cause the grossed-up ICA to be unrepresentative of undistributed taxable income more significantly.

3.34 Both the accounting retained earnings and grossed-up ICA approaches have strengths and weaknesses. The Government therefore proposes to make the undistributed earnings portion of the grossed-up deemed dividend amount the higher of the grossed-up accounting retained earnings (less non-taxable capital gains) or the grossed-up ICA at the time of sale.

Consequences to seller

3.35 When a shareholder that (together with associates) controls a company sells shares in the company, a portion of the sale proceeds would be treated as a dividend. That portion is:

• the grossed-up undistributed earnings of the company at the time. This is the higher of:
  • the accounting retained earnings (less non-taxable capital gains) grossed-up to a pre-tax amount (by adding the ICA balance), or
  • the ICA balance divided by the company tax rate
• pro-rated to the proportion of shares sold (the income interest of the shares sold as a percentage).

3.36 This is the amount of gross dividend income included as part of the sale price. If the sale price is less than the corresponding net dividend amount, then the sale price is the net dividend, and that must be grossed up by the amount of imputation credits that are or could be attached to determine the gross dividend.

3.37 The allocated ICA amount is then available to the shareholder to use as a credit. This is illustrated by way of examples later in this chapter. The effect is the shareholder must pay the top up tax if its personal tax rate is higher than the company tax rate.

Consequences to the company

3.38 Since the retained earnings out of which the dividend is deemed to be paid remain in the company, the seller is treated as if it received the net dividend and immediately returned it to the company as additional capital. To reflect this, the ASC of the company is increased by the deemed net dividend.

3.39 If the ICA balance is not forfeited on sale (which would be the case if the shareholder sells the company to another company that it controls or the sale does not result in a 34 percent change in ownership since the credits arose), then the ICA balance is reduced by the amount of imputation credits used by the seller to reduce its tax on the deemed dividend.

Consequences to the buyer

3.40 There are no special consequences for the buyer. The buyer is treated as buying the shares for the price it paid for them, but it does get the benefit of additional ASC in the company generated by the deemed dividend.

Other issues

Should the amount of the deemed dividend be limited to the gain on the sale?

3.41 The Government does not propose to limit the deemed dividend to the gain on sale (rather, it should be limited to the total sale proceeds). This is because limiting it to the gain on sale would effectively allow a deduction for a capital loss.

3.42 If a company has accumulated $100 over a year, you would expect it to appreciate in value by $100 over the year. However, if the company only appreciates in value by $50, it means some other asset or part of the business (such as goodwill or land) has depreciated by $50 over the year. This means the deemed dividend amount should not be limited to the gain, as that would implicitly allow the deemed dividend (which should be the entire $100 of retained earnings) to be reduced by the capital loss.

Why apply the rule to sales by companies?

3.43 If the rule did not apply to sales by companies, it may be possible to prevent a deemed dividend from arising by using a holding company structure. Consider the following fact scenario:

• A shareholder may own a holding company which owns an operating company.
• The operating company has retained earnings, but it never paid a dividend to the holding company.
• The holding company sells shares in the operating company, realising a capital gain (not taxed).
• The shareholder then sells the holding company (which holds the value of the company that was sold). The ICA balance of the holding company is nil, so this rule would not deem a dividend to arise to the shareholder from the second sale.

3.44 Applying the rule will address this issue because the first sale would result in a deemed dividend to the holding company, and this would transfer the ICA balance to the holding company. When the shareholder then sells the holding company, this rule would apply to deem a dividend arising to the shareholder.

Corporate groups

3.45 If a shareholder is selling shares in a parent company of a corporate group (a group being at least 50 percent common ownership), the ICA amount on which the deemed dividend is based should be the net of all ICA debits and credits of each company in the group, as well as the consolidated imputation group if there is one. For companies less than 100 percent owned, the individual ICA amounts should be pro-rated by the portion of ownership.

Example 6: Calculation of deemed dividend and change in ASC for group of companies

 

Parent, a company, has an ICA balance of $1,000. It has subsidiaries with the following ICA balances:

• A company it owns 80 percent of with a $500 balance.
• A company it owns 100 percent of with a $100 debit balance.
• A company it owns 50 percent of with a $500 balance.
• A company it owns 49 percent of with an $800 balance.

If a shareholder sells its shares in Parent and this rule applies, the ICA balance used for determining the deemed dividend amount is calculated by summing the following amounts:

• $1,000 (being Parent’s ICA balance)
• $400 ($500 × 80%)
• -$100
• $250 ($500 × 50%)
• Zero (the 49 percent-owned company is not taken into account).

The above calculation gives a total of $1,550. If Parent has consolidated its accounting earnings with its subsidiaries, this amount should reflect the retained earnings of the group (so that would be the amount of retained earnings used without looking through to each underlying company).

ASC adjustments would apply for all companies that contributed to the calculation of the deemed dividend amount because they had retained earnings or a non-zero ICA balance. This would be the net dividend paid from their own retained earnings or grossed-up ICA, plus any amounts deemed on-paid by subsidiaries. The parent company would have an ASC adjustment of the entire net dividend amount deemed to have been derived by the shareholder.

Another approach may be to take the total deemed dividend amount, and pro-rate it among contributing companies in proportion to their respective ICA balances or retained earnings (with parents of subsidiaries also including their subsidiaries’ amounts).

Sales of shares in a controlled company to another company controlled by the same shareholder

3.46 If shares in one controlled company are sold to another controlled company that also has retained earnings or an ICA balance, the amount received is potentially in substitution for dividends paid by both companies. Not only is the amount received by the shareholder potentially a payment for the retained earnings of the sold company, the purchasing company may pay for the shares out of its own retained earnings.

3.47 United States legislation provides that in this case, a dividend is deemed to be paid first by the purchasing company, and if that does not account for all that is paid for the shares, another dividend is deemed to be paid by the target company. This is because the cash or other consideration is being paid by the purchasing company to its shareholder, so it is the purchasing company’s earnings that are being transferred to its shareholder. To the extent the purchase price exceeds this amount, it would be paid from the capital reserves of the purchasing company. However, as with the more general rule, if that amount is compensation for the retained earnings of the target company, that would be treated as a dividend from the target company.

3.48 In this case, it is proposed to follow the order used in the United States law (the amount received is first deemed to be paid out of the retained earnings of the purchasing company). This is calculated by grossing up the ICA balance or retained earnings of the purchasing company (or group) as described earlier in the context of determining the amount of a deemed dividend with respect to the sold company. As the purchasing company is paying cash or equivalent to the seller (consideration paid in the form of shares in the purchasing company is not taken into account in determining the amount of the deemed dividend), no amount is deemed to be recontributed to the purchasing company, and there is no adjustment to the ASC of the purchasing company for the deemed dividend amount.

3.49 However, the shareholder should still be entitled to the protection of the ASC of the target company in the case of a liquidation or share repurchase of the group, including the purchasing and target companies, as that ASC would have been available to reduce those amounts if the target company had continued to be owned directly by the shareholder. To reflect this, the ASC of the purchasing company should be increased by the lesser of the market value of the shares it acquired in the target company, and the ASC of the target company.

3.50 If the deemed dividend from the purchasing company does not account for the entire amount paid for the shares, then the amount of a deemed dividend from the target company is determined as described earlier, but this time it is limited to the amount paid after subtracting the deemed dividend from the purchasing company. The ASC of the target company is increased as described earlier. This increase in the ASC of the target company is deemed to happen immediately before the acquisition by the purchasing company, so the increase may also apply to the ASC of the purchaser (that is, an ASC increase of the lesser of the market value of the target company and the ASC of the target company).