I’ve had many arguments with stock-market enthusiasts about “total return”. They see the stock market cap growing with GDP, say 6%, note that stocks also pay dividends, say 3%, and decide that their own stock of wealth can grow at 9% if they re-invest dividends, or, equivalently buy debt or the stocks of other firms with their dividends.
At this point, I try to explain that nothing can grow faster than GDP, over the long term. Like any outperformance strategy, dividend-reinvestment works only when not too many people try it.
When too many people try it, all that happens is that the dividend yield falls, and the cost of capital falls as well. When buying an asset, it is because you believe it is undervalued, and you are willing to accept a lower return. It may be a useful approximation to view everyone as a price taker — it’s just a small error. But when many people repeat this error, then the sum total of the small errors is a material error.
In this case, the investors do not really want the return on capital to decline — their return demands are not lower than the market demands, but higher. They expect to get the “full” return in which both principle and dividend are compounded, instead of just having the principle compound, paying a constant yield each period. But if the firm could increase its capital stock while still maintaining the current earnings, then it would not have paid out the dividend. As investors push the money back to the firm, they are signaling to the firm that they want lower earnings — a lower cost of capital — when in fact they are not willing to accept a lower cost of capital.
As the cost of capital declines, firms swell in size, buying more capital goods, and the price of capital goods goes up, creating capital gains during the growth spurt.
After the growth spurt, the capital stock will go back to growing with the rate of GDP, but now with a lower yield, so that “total return” will be less than before.
And when this happens, the investors will sell the stock, as they were not willing to accept the lower returns. In the process, firms will liquidate capital, up until total return is restored to what it was before, at least until the next Dow 36,000 book is published, and it becomes vogue to dream of 8% annualized returns.
Here is an illustration of the process, in which capital goods are not produced perfectly elastically, so that the price of capital goods varies with demand.
We’re going to model the return from purchasing a share as that from purchasing a capital good. The return from purchasing a capital good is the sum of the change in price of the capital good + dividend supplied. In instantaneous terms, this is depicted below:
But when the original dividend yield is low, then the change in price dominates and returns become negative. In that case, only an excessively high dividend yield will be sufficient to get investors to invest — wages must fall below MRP in order for investment to continue during the shaded period, as no one will purchase capital that is declining in price unless the (instantaneous) dividend is greater than the rate of decline.