David Greenwald and Sydney Ludvigson of New York University and Martin Lettau at the University of California Berkeley have estimated that more than all of the rise in US share prices since 1980 (adjusted for inflation) have come because profits rose at the expense of wages. By contrast, fluctuations in productivity growth in this time had little effect, although slower productivity since 2008 has depressed prices.
There's a simple reason why productivity growth doesn't much matter for shares, says Professor Ludvigson. It's that variations in them benefit or hurt workers more than shareholders; faster productivity growth means faster real wage growth, and slower productivity growth means slower real wage growth.
In this context, it's not so paradoxical that the S&P 500 is close to an all-time high while real incomes for the typical American household are 10 per cent lower than they were in the late 1990s. Shareholders have gained from a shift in incomes away from workers and towards themselves. Rising real share prices since the early 1980s and increased inequality, says Professor Ludvigson, are part of the same story.
This corroborates earlier work by Jean-Pierre Danthine of the Swiss National Bank and Columbia University's John Donaldson. They've found that shifts in the distribution of income are a "risk factor of first-order importance" for equities.
There's a simple reason why this isn't sufficiently appreciated. Stock market pundits focus upon short-term moves in prices, and these are mainly due to changes in risk appetite and optimism. In the long run, however, these shifts tend to cancel out. Instead, at long horizons it is changes in the share of profits that matter.
Greed and fear might set the weather for share prices, but profits and wages - the balance of class power - set the climate.
Now, this doesn't mean that shares are a safe long-term investment for working people because share prices are a hedge against wage incomes. For one thing, some of our wage earnings are in fact a share of profits; powerful workers - not just bankers and footballers - can grab for themselves a share of what would otherwise be profits. And for another thing, there's idiosyncratic risk in labour incomes: even if workers generally gain at the expense of shareholders, it doesn't follow that every individual worker will.
What it does mean, though, is that if you are a long-term investor - say because you're saving for retirement - your future equity wealth depends heavily upon the distribution of income between wages and profits.
Here, there are two especial risks.
The upside one (for shareholders) has been described by Erik Brynjolfsson and Andrew McAfee. They say that technical change could see workers replaced by machines, which would force down wages to the benefit of profits.
The downside one is that there might eventually be a political backlash against increasing inequality and the growing wealth of the top 1 per cent.
It is impossible to tell which of these forces will prevail. What we can say is that they will heavily determine the future of equity returns. In this sense, those returns are not just risky, but fundamentally uncertain.