Dividend Irrelevance and Homemade Dividends

Imagine a company that pays no dividend at all. Can a shareholder who wants cash income every quarter still get it? Yes: they simply sell a few shares each quarter. Miller and Modigliani's dividend irrelevance theorem says that, in a world without taxes, transaction costs, or information asymmetries, this "homemade dividend" is a perfect substitute for a real one, so a firm's choice of dividend policy cannot affect its share price.

The logic follows directly from value conservation. Paying a dividend of $1 per share reduces the company's cash and its share price by (roughly) $1, the shareholder is no richer, just holding cash instead of equity. If instead they wanted that $1 without a dividend, they could sell $1 worth of shares themselves, ending up in exactly the same place. Either way, total wealth, shares plus cash, is unchanged.

Worked example. A stock trades at $50 with no dividend. An investor who wants $500 of cash from a 100-share holding sells 10 shares, leaving 90 shares worth $4,500 plus $500 cash, total $5,000, same as before. If the company had instead paid a $5-per-share dividend, the stock would drop to about $45, and the investor would hold 100 shares worth $4,500 plus $500 cash, the identical outcome.

The theorem is a benchmark, not a prediction: real markets have taxes that treat dividends and capital gains differently, brokerage costs that make homemade dividends imperfect, and signaling effects where a dividend change conveys information about management's view of future earnings. Those frictions are exactly why dividend policy debates exist in practice, irrelevance is the clean starting point against which every real-world complication is measured.