What Dividend Imputation Means for Retirement Savers

Abstract

We examine the implications for Australian retirees of full access to dividend imputation credits using a stochastic life-cycle model. We find that the availability of imputation credits can justify a significant bias towards Australian equities in retirement portfolios, largely at the expense of world equities. We also generate estimates of the value of imputation credits to retirees, finding it could potentially support increased consumption during retirement of 5-6 per cent, or equivalent value to a 8-9 per cent higher balance at retirement. Our study enhances the understanding of equity home bias, and provides insights relevant for public policy. 1 Estimated as 30/70 = 0.429, with reference to a fully franked dividend of 70 cents treated as being paid out of pre-tax earnings of $1 with 30 cents of tax 'pre-paid' by the company. 181

The ‘right’ level for the superannuation guarantee: identifying the key considerations

“…Superannuation funds pay a tax rate on investment returns of 15 percent on regular income and 10 percent on capital gains pre-retirement, and zero tax in retirement. Superannuation funds also qualify for imputation tax credits (Butt et al, 2019). Imputation credits reduce the effective tax rate on the risky asset, which we calculate to be 6.74 percent in the pre-retirement phase and − 6.05 percent in the post-retirement phase.…”

“…Some thought might be given to allowing for more flexibility around the contribution rate, perhaps by positioning the SG more like a default than a compulsion, allowing members more scope to apply their own optimal rate. Khemka, 2015;Butt et al, 2019). The approach involves discretising the state variable W t into increments of $20,000 over the range $0 to $2 million.…”

“…Our baseline assumption is ρ = 4, which sits within the range used in the literature (e.g. Ameriks et al, 2011;Yogo, 2016;Butt et al, 2019). ρ = 2 and ρ = 6 are examined under sensitivity testing.…”

“…The latter is captured by weighting outcomes during the retirement phase by the probability of survival based on the Australian Life Tables (2015-17) 2 including mortality improvements. We apply the numerical dynamic programming method outlined in Butt et al (2019), which is similar to that used by . Table 1 summarises the main elements of the model, with baseline assumptions appearing in the middle column and the alternative assumptions in the last column.…”

“…While reference-dependent utility functions have been applied in a life-cycle context by Blake et al (2013) and Butt et al (2019), our particular application presents a further contribution. Reference-dependent utility aligns with the view that the primary concern of the superannuation system relates to adequacy of income.…”