Disclaimer: There are a lot of interesting aspects to this question. Shareholder voting rights, control over the company, etc. are all interesting things to consider. Here, I only focus on a few parts of the question.
Excess Volatility
Claim: Are actual stock prices much too volatile to be explained by dividends? Is this evidence that the price of a stock depends on more than just future dividends?
It's perhaps not much of an exaggeration to state that the paper that won Robert Shiller his Nobel prize in economics is the paper "Do Stock Prices Move Too Much to be Justified by Subsequent Changes in Dividends?" In this paper, Shiller looks at historical prices and dividend payments and asks the question, suppose market expectations of future dividends were on average correct? What should the prices of a stock been historically if those expectations were on average correct? He argues that we can do this by looking at the dividend payments and prices and constructing and "ex-post rational price" at each point in time, called $p^*$. He then compares that price to the actual historical prices, $p$. This is shown if Figure 1 below for the S&P Composite Stock Price index from 1871 to 1979, after detrending both series. The result is that $p^*$ appears to be far too smooth relative to $p$. He argues that this is a problem because it implies that actual stock prices are too volatile to be explained just by movements in future dividends.
References and other resources:
Important Note: In the current asset pricing literature, there is a strong belief that this "excess volatility" puzzle is not a puzzle at all. Empirical and theoretical investigation seems to indicate that this "excess volatility" arises from variation in discount rates. Shiller constructed this ex post ration prices by assuming that the discount rate was constant over time. Empirically, this seems to be far from the truth. For a good discussion of this, see the following:
Rational Bubbles
Claim: Perhaps stock prices can be explained by a belief that prices will continue to grow in the future, even if dividends do not? That is, suppose the price of a stock is not derived from a belief in future dividends but in the belief that prices will continue to growth without bound.
There is a literature that investigates the possible effect of "rational bubbles" on the price of a stock. This is summarized nicely in the "Discount Rates" paper cited about. One way to investigate this question, as described in the Discount Rates paper is to use the Campbell-Shiller (1988) present value identity,
$$
dp_t \approx \sum_{j=1}^k \rho^{j-1} r_{t+j} - \sum_{j=1}^k \rho^{j-1} \Delta d_{t+j} + \rho^k dp_{t+k},
$$
where $\Delta d_{t+j} = d_{t+j} - d_{t+j-1}$, $dp_t := d_t - p_t = \log(D_t/P_t)$, $D_t$ is the dividend, $P_t$ is the price, $r_{t+1} = \log R_{t+1}$ is the log return of the asset, and $\rho \approx 0.96$ is a constant of approximation. This is an identity based on the definition of returns $R_{t+1} = (D_{t+1} + P_{t+1})/P_t$. This says that the normalized prices of an asset ($dp_t$) is derived from three possible terms:
- Future returns on the asset (discount rates)
- Future dividend growth paid by the asset (cash flows)
- The future price of the asset (bubble term)
Investigation of the contribution of each of these terms to the volatility of normalized prices indicates the nearly all of the volatility comes from variation in the discount rates term and essentially none comes from the dividend growth and rational bubble terms. This also explains the above "excess volatility" puzzle. Note that this still means that dividend explain the value of a stock. It just rules out that changes in the price-dividend ratio of a stock are not explained by changes in future dividends nor by belief in rational bubble growth in prices.
You can view more of the details of the empirical investigation in the paper or the same videos linked in the previous two section.
Conclusion
The attack against the present discounted value model of stock prices embodied in the "excess volatility" puzzle seem to have an explanation in the fact that discount rates vary over time. An investigation of the contribution of future dividend growth, discount rates, and "rational bubbles" to dividend-price ratios indicates that "rational bubbles" do not seem to explain prices. Rather, dividend yields seem to be entirely explained by discount rate variation. Further investigation seems to indicate that prices and dividends seem to be cointegrated. All of this combined is consistent with the theory that prices reflect the present value of expected future dividends.
However, as noted in my disclaimer above, this leaves open many interesting questions. There is still much to do and learn in this literature.
