Signaling theory and Miller and Modigliani’s (MM) dividend irrelevance theory are two approaches to payout policy that oppose each other. Through analysing the recent payout policies of major firms in today’s financial market and investigating the literature on the theories we can see some of the assumptions of MM’s theories can’t be true. Asymmetry of information and investment policy are a part of the financial world and prevent MM’s theories from being relevant while helping to prove the basic concepts of signaling theory.
MM’s theory on payout policy was developed due to other theories at the time being incomplete. Prior to MM most economists believed the larger the dividends a firm declared the greater the value of the firm (Allen and Michaely 2003). This would result in some firms issuing shares to fund large dividend payments (Brealy, Myers and Allen 2013) regardless of any future repercussions. MM introduced the idea that dividend payout is irrelevant as long as investment policy remains constant. This theory doesn’t change as repurchasing becomes more popular as it doesn’t affect investment policy.
The fundamental concept of MM is that firms cannot create value for stockholders over and above the income generated from their investment policy (DeAngelo, DeAngelo and Douglas J 2009). Payout policy is important in that managers distribute 100% FCF generated over the financial period as not doing so would decrease shareholder wealth. Payout policy becomes irrelevant as managers under the MM model must pay out 100% of FCF generated, they have no choice. A profound example of the misunderstanding of this theory is that 58.4% of CFO’s that had not paid dividends or repurchased shares in the previous three years had said they may never do so again (Brav, et al. 2005). Effectively minimizing shareholder wealth- the exact opposite of a firm’s goal.
For MM’s theory to be relevant there are some key assumptions that have to be met and it is here that MM struggles to be applicable: * Investors are rational * Contracts are complete * Information is symmetrical * There are no taxes * Markets are complete * Cash flows from investment policy must remain constant
When analyzing these assumptions the two that pose the biggest problem for MM is that cash flows remain constant and information is symmetrical.
The assumption that information is symmetrical is rarely applicable. Even today, with the abundance of laws requiring transparency of information from firms, a manager will always hold greater knowledge of value of a firm compared to an investor. This is the core concept for dividend signaling theories. As an alternative to models such as Gordon (1959) which proposed that dividends increased firm value by having a more secure, immediate payment, signaling theories develop the idea that dividends affect share price due to the information they represent (Allen and Michaely 2003).
Battacharya (1979), John and Williams (1985) and Miller and Rock (1985) all developed dividend signaling theories. These models help explain the market shifts that occur with a change in dividend payment or the announcement of a share repurchase, and they make basic sense. It is unlikely that managers and investors will share the same level of knowledge about a company in the real world and the symmetry of information is therefore an assumption of MM that often falls flat.
A survey in 2004 (Brealy, Myers and Allen 2013) highlighted 3 ideas that were important to financial managers: 1. Manager are reluctant to make dividend changes they may have to reverse. 2. Managers smooth dividends 3. Managers focus more on dividend changes than dividend levels e.g. a rise of 15% is more significant than dividend levels staying steady.
These 3 points can be supported with the dividend announcements of BHP, Rio Tinto and FMG in tables 1 and 2. Rio recorded a