Dividend ratio's shift over time

Andrew Smithers
Financial Times - Letters , 14th March 2009

Your editorial “The double edge of dividend cuts” (March 7) claims that “dividend cuts hit long-term returns”. The standard view (Miller-Modigliani theorem) is that returns are unaffected by payout ratios. The apparently stable long-term real return on equities of around 6 per cent could be achieved either by average dividend yields of 3 per cent with dividends per share rising at 3 per cent a year, or by 5 per cent yields and dividend increases per share of 1 per cent a year. You quote the evidence that the second has been the long-term pattern with the inference that, if payout ratios had been lower, the rise in dividends per share would have been unaltered and the total return therefore lower.

It is highly unlikely that payout ratios alter long-term returns on equity. They simply affect the level of new issues and buy-backs. Increases in retained profits reduce the need for new issues. Are you about to publish a refutation of the Miller-Modigliani theorem?