There is currently considerable interest in the practice of “dividend washing”. This refers to the practice of investors being able to trade shares cum-dividend for a period after the ex-dividend date has passed, under special arrangements provided by the ASX.
It is suggested that some investors are “double dipping” into the pool of franking credits distributed with dividends, by selling shares they hold once they go ex-dividend and then buying replacement new shares in that special trading, which they receive cum-dividend.
They get, for example, dividends (and attached franking credits) on 2000 shares despite only owning 1000 shares at any point in time. If foreign investors are selling the shares, there is a cost to tax revenue because they would have been unable to use the franking credits attached to the dividend.
The issue of dividend washing can be addressed in three steps. First, should the practice be permitted? Second, why might cum-dividend trading be allowed after the ex-dividend date? Third, if the rationale for allowing such trading reflects current institutional arrangements, is there some simple adjustment to those arrangements, which is a superior solution to others proposed?
It is clear that dividend washing is at variance with the objective of preventing trading in imputation credits, which transfers those credits from those unable to use the tax benefits (foreign investors) to domestic investors. Dividend washing also enables some domestic investors to essentially gain a dividend “twice” by selling their current stock holding ex-div and then buying cum-div in that period of trading permitted by the ASX after the ex-div date.
But this is only a cost to government tax revenue if it increases the total amount of sales of cum-dividend stock by foreigners. It may be simply a redistribution of the timing of cum-div sales by foreigners from before the ex-div date to this later period. That is an empirical question.
Cum-dividend trading after the ex-dividend date arises because of the institutional arrangements associated with trading of equity options on the ASX. Specifically:
- writers of call options are required to deliver stocks if options are exercised against them
- buyers of call options might exercise an option on the last
cum-dividend day - the allocation of exercised options against writers (in the event that less than 100% of outstanding options are exercised) is done randomly because of novation
- that process occurs overnight, so option writers will not know they have been exercised against and need to deliver stocks until the ex-dividend day.
Consequently, the stocks they are required to buy to deliver (if they are not hedged by already holding such stocks), will not have dividends attached, whereas the option holder is entitled to receive stocks with the dividend attached. By allowing a short period of cum-dividend trading after the ex-dividend date, this problem is resolved – but creates the opportunity for dividend-washing transactions unrelated to option trading.
The first point to note, is that the reason for allowing cum-dividend trading is unrelated to the issue of franking-credit use. If the same institutional arrangements for options trading and settlement existed in a market without dividend imputation, the same issue would apply. The option writer would still face the problem of being exercised against and not being able to buy shares with the dividend attached.
Therefore, there are two pieces of empirical evidence that are worth examining. First, is there any evidence of cum-dividend trading being greater for stocks paying franked dividends than stocks paying unfranked dividends in Australia? If so, this may suggest that cum-dividend trading is being requested primarily for trading of franking credits. Second, do similar arrangements for cum-dividend trading apply in overseas markets? If not, perhaps there are alternative institutional arrangements for options trading that obviate the need for allowing cum-dividend trading.
What other institutional arrangements for options trading could be considered, which would obviate the need for cum-dividend trading?
One arrangement would be to allow settlement of options contracts in cash rather than requiring physical delivery. Then an option writer exercised against would be liable to pay the cash amount equal to the difference between the strike price and market price at the time of exercise, plus the grossed-up value of the dividend. (The grossed-up amount is the cash dividend plus the amount of the franking credit.) They would not have to buy stock cum-dividend.
An alternative approach would be to rewrite the option contract terms so the option holder would be entitled to receive the share cum-dividend only if the exercise occurred one or two days prior to the last cum-dividend date. This would provide the option writer with time to purchase the stock cum-dividend for delivery. Traders would need to know which was the key date one or two days prior to the last cum-dividend date and this may make arrangements for option trading slightly more complicated. However, option traders are expected to be financially sophisticated enough that this should not be a major issue.
Alternatively, if physical delivery is the only option, and a rewriting of the exercise conditions is not made, it would be possible to make a simple amendment to the tax laws to prevent dividend washing. Specifically, there would be a requirement that purchasers of shares cum-dividend after the ex-dividend date are not entitled to claim the franking credits. They would be able to deliver the shares cum dividend, inclusive of franking credits, in settlement of option obligations to option holders.
That approach might appear to give rise to dividend-washing opportunities by an investor buying both stock and call options in the same stock at least 45 days prior to the ex-dividend day. The strategy would be to exercise the options on the last cum-dividend day (prior to the ex-div day) and sell the stock already held on the ex-div day (using the proceeds to make the payment required on exercise of the call option). The investor would appear to be entitled to receive the dividend and franking credits on the stock and would then also receive the dividend and franking credits on the stock received from exercise of the call options.
However, the last-in-first-out rule would preclude this, as long as the date of purchase for the shares due from the call option is the date-of-option exercise (rather than delivery date). Those shares would then be deemed the ones sold on the ex-dividend day, and thus held for less than 45 days, precluding use of the franking credits attached to the dividend.
To the extent that particular institutional arrangements for ASX option trading give rise to the problem of dividend washing, it is perhaps appropriate to examine whether those arrangements can be changed in a simple manner.