Monday, December 15, 2014

Compensation and Risk Premiums

Maybe I'm just repeating myself, but I wanted to revisit my recent post about how low risk for equity is related to high compensation share.  This is really the same idea I played around with before, when I described labor as having some of the characteristics of debt, in that equity owners earn a premium from both of these categories of inputs in exchange for accepting fluctuations in income.  This risk trade would create a discount in the cost of labor.  When risk premiums are high, labor would have to take a large discount and when risk premiums are low, labor would take less of a discount.  Historical data on compensation, debt, and equity seems to broadly suggest support for the various interrelationships of interest rates and risk premiums.

Here is a graph that makes the relationship a little more clear than the graph from the recent post.


I have included rental income in compensation, because this is generally owner-occupied rent to homeowners, so it is income widely distributed among households, and its recent increases have pushed compensation down in a way that is unrelated to corporate capital income.

I am using the annual Equity Risk Premium (ERP) from Damodaran (not unlevered).  We can see that there have been periods with generally high ERPs and periods with generally low ERPs.  (It is inverted in the graph.)  When ERPs have been low, compensation share has tended to rise.  When ERPs have been high, compensation share has tended to fall.

As a reminder, here is the graph from the recent post.  Note that the compensation share is coming out of the profit share.  Profit shares are going up when compensation shares are going down, and vice versa.  This could lead to a simplistic interpretation that labor and capital are locked in a fight for incomes, with gains for one coming at the other's expense.  One thing that is so interesting about these topics is how very subtle changes in how we look at the data can flip the interpretation on its head.

If this was a story of simply fighting over shares of income, profit shares would decline when compensation increased, but we shouldn't expect ERPs to decline.  We would expect profits to fall and valuations to fall along with them.  What we find, instead, is that a fall in ERPs is the factor which coincides with rising compensation.  In fact, equity prices were very strong in the late 1980s and late 1990s, when compensation was growing.  Equity gains in the 1960's were lackluster.  Here are corporate tax rates over the period, from the Flow of Funds report.  There is no obvious pattern there.  These periods span the Johnson, Reagan, and Clinton presidencies - three different eras of fiscal governance.  But, these periods had something in common, which was that equity holders demanded less of a premium during these periods for holding equity instead of debt.

One other similarity is that the periods with low ERPs tend to have longer business cycles.  There have been three recoveries that lasted the better part of a decade over the past 60 years, and they coincide with these three periods of low ERPs and increasing compensation shares.  And, these periods had moderate inflation, generally in the 2%-4% range.  The high ERP period in the 1970s coincided with high inflation, and the recent high ERP period has coincided with low inflation.

So, I propose that what we see is not simply a redistribution of income.  It's more of a growth in risk-adjusted income, which is not easily measured with nominal dollars.  The increase in compensation is a product of the lens through which equity holders view their income, which is as a premium on risk.  When there is less risk, or at least less risk aversion, equity bids away less nominal income, leaving more for labor and debt.  And, further, there is some evidence that a causal factor is stable aggregate demand.

No comments:

Post a Comment