Image by Ahmad Ardity from Pixabay

Tax-driven off-market buybacks were first used in 1997 by the Commonwealth Bank. Since then, TOMBs (as we call them) have been used nearly 60 times, usually by large Australian Stock Exchange (ASX) listed companies. TOMBs are a complex capital management technique that involves a buy-back of shares in which the payment consists primarily of a franked dividend component with the remainder designated as a return of capital. BHP has done five and Commonwealth Bank and Rio Tinto four each, with the largest by BHP in 2018 distributing almost $7.4 billion to shareholders.

Overall, almost $50 billion of cash has been returned to a very specific type of shareholder, together with around $17 billion of franked dividends. All shareholders can bid into the tender process used to participate, but those on low tax rates to whom the franking credits are of most value will logically put in the lowest bids and win the right to participate. Because of those franking credit benefits, the buy-back price determined in the tender is anomalously below the price at which the shares are trading on the ASX.

For example, if a company’s current share market price is $10, the buyback price determined in the tender generally (for reasons explained below) would be below market, at $8.60 split between franked dividends of, say, $7 and $1.60 of return of capital.

Government loses billions in tax revenues because of TOMBs

The participating shareholders in TOMBs are either in the zero or 15 per cent tax brackets, including superannuation funds, retirees and charities. The buy-back achieves a ‘streaming’ of franking credits to these shareholders at a cost to government revenue, relative to a pari-passu payment of dividends to all shareholders. The low tax-rate participants receive a refund of franking credits.

There is also a potential capital gains tax benefit to participants (another cost to taxpayers), since the capital return component is low, leading to capital losses for tax purposes – although seemingly arbitrary tax determinations have slightly reduced the significance of that effect.

We estimate a cost to government revenue in 2018 of around $2 billion, driven largely by the $7.4 billion buyback by BHP.

That revenue cost to government should, by itself, be sufficient to initiate action to prevent this type of transaction – particularly since there are equally effective, and much less complex, ways of distributing cash and franked dividends to shareholders. A special dividend, together with a separate return of capital to all investors could be structured to pay out the same amount of cash and franked dividends, with much less administrative complexity and cost. There is no reason, other than streaming franking credits to preferred shareholders at the expense of the taxpayer, for this elaborate piece of financial engineering.

There are other reasons for banning the practice

Another reason for banning TOMBs is their complexity, involving the need for special regulatory and tax treatment and resulting administrative cost. The Australian Securities and Investment Commission (ASIC) has to give  special dispensation relief from its rules to allow such buybacks. The Australian Tax Office (ATO) has had to make a range of special tax determinations to deal with TOMBs. These include: imposing a maximum 14 per cent limit on the discount to current market price; special treatment of determination of capital losses for tax purposes; a special deduction of the company’s franking account balance designed to offset the franking credit streaming effect on government revenue. That deduction, which accounts only for streaming away from foreign shareholders, we argue is inappropriately calculated.

Another reason is the issue of equitable treatment of shareholders which has gained press coverage over the years. As explained, zero or low-tax investors benefit most from receiving franked dividends and are willing to sell shares into the buyback at a discount to current market price. That discounted sale price is to the benefit of non-participating shareholders, but whether the discount is adequate to compensate for the loss of franking credits is unclear. The ATO regulates a 14 per cent maximum discount. Low tax rate investors would be willing to accept a much lower price for their shares (our calculations suggest around a 20 per cent discount) so this maximum discount limit is to the disadvantage of non-participants.

The maximum discount of 14 per cent also means that there is a massive excess supply of shares offered at that discount, leading to large scale-backs of amounts bought back relative to those offered by participants (often in the vicinity of 70-80 per cent). That creates more uncertainty for participants about likely success of their tenders, and means that most participating shareholders will have some of their shares accepted and others not. Participants ‘win’ on those shares accepted and ‘lose’ on those shares not accepted, making the distributional effects of TOMBs very hard to identify precisely.

Finally, the mechanics of TOMBs are inconsistent with the spirit of other tax integrity rules especially the 45-day rule designed to prevent trading of franking credits. The company usually allows, say, 46 or 47 days between announcement and the actual buyback date, permitting investors to buy after the announcement and tender so as to access the distribution of franking credits. That increases the amount tendered and increases the size of the scale-backs, and creates uncertainty of outcome for those attempting to use this strategy, and for other shareholders.

We are also critical of the transparency with which legal and administrative decisions around TOMBs have been made. We are unable to find any public explanation from the ATO of why 14 per cent has been chosen as the maximum allowable discount.

Inside the company franking account, the additional debit to the company’s franking account balance is, we believe, not appropriately justified. By focusing solely on the streaming of franking credits away from foreign shareholders, this misses the potentially more important streaming away from high-tax-rate domestic shareholders.

COVID-19 may halt TOMBs temporarily, but in future they should be banned

There have been attempts, including changes proposed by the Board of Taxation, to improve the tax treatment of off-market share buybacks. The draft legislation to effect the changes was dropped by the Government in 2013 with no public explanation. Those changes were, in our view, an improvement on the treatment of TOMBs.

The COVID-19 crisis is likely to halt buybacks at least temporarily as companies hoard cash to strengthen balance sheets. But a far better approach would be to make legislative or regulatory changes to simply kill off TOMBS. In the longer term, the possibility of the demise of dividend imputation, which drives the use of TOMBs, may have the same effect.


This is based on our paper ‘Tax-driven Off-Market Buybacks (TOMBs): Time to Lay Them to Rest’, Australian Tax Forum, 35(2), 2020, 232-257.

This article has 1 comment

  1. If the following propositions are correct:
    – the progrssivity of tax rates should be relative to the circumstances of real persons not legal enities.

    – given the above, company tax should only ever be, in effect a withholding tax to ensure payment. like all withholding tax it is then credited later against the individual’s personal tax rate which is adjusted for that person’s ability to pay tax

    This is how franking credits work except, contrary to the above principles, company tax is also paid on retained earnings not just distributed dividends. This leads to many companies having pools of undistributed franking credits. These retained earnings and their franking credits could be paid out in full (and in some cases have been) funded by the company issuing new shares to the same value. This would return the tax rate of each individal shareholder to their personal rate on the full pre-tax earnings of the company.

    The problem described above really arises from the fact that the personal tax rate of some individuals (via super) is unnecessarily low not because of the over-generosity of the franking credit system.

    On the contrary, the franking credit system could be improved by applying company tax only to distributed earnings like the old British Advance Corporations Tax. This would then ensure that company tax operated purely as a withholding tax and make sure all shareholders paid tax on their pre-tax income derived from corporate vehicles at their own tax rate and would eliminate pools of franking credits which are the basis of elaborate streaming structures. It would still tax foreign shareholders as the dividends are still taxed in Australia before any receipt.

    The usual objection that the capital gain resulting from retained earnings is escaping tax is incorrect. As all share prices are the NPV of future dividends are aleady deflated by the tax rate so gains in share prices are also already deflated by the tax rate. The government also receives its full value of tax revenue when those initially untaxed retained earning together with the investment earnings on that reinvestment is eventually distributed to shareholders. the Government is in effect a co-investor with the shareholders and equally patient for its return as its co-shareholders.

    The present vast complexity (10,000+ pages of it administrered by armies of tax lawyers and accountants and ATO personnel) of the determination of a company’s taxable income is also then of minimal importance as it does not drive the tax payment the company’s own decision to pay a dividend does.

    This might help reduce the relatively high cost of equity capital in Australia as demonstrated by the number of foreign acquisitions of Australia far out numbering Australian foreign acquisitions and a branch office economy.

Leave a comment

Your email address will not be published. Required fields are marked *