Sunday, August 12, 2012

GMO Defends Equities Inconsistently

Pimco maven Bill Gross recently warned that both equities and bonds look headed for a bad intermediate future.  He makes the interesting observation that as labor income has declined over the past century, this allowed capital to reap more reward than what is merely in GDP growth (see chart below).  Yet, this cannot continue any more than bonds can appreciate from lower yields.  The result is pretty scary, because as Gross points out, a 4.75% asset return premium needs about a 7% growth from equities if bonds are going to generate 2% returns, and if this won't happen, CalPERS and many of its ilk are in serious trouble.


Ben Inker of GMO responded by noting two points. First, one doesn't see much relation from developed or developing countries between equity returns and GDP growth, over many decades. This is really interesting because most general equilibrium models of asset growth would show a relationship, and so to my mind this highlights the inappropriateness of 'Lucas tree models' for understanding the aggregate stock market. Equities should not be thought of as residual claims on an economy, but rather the peculiar returns to an asset class in a very complex equilibrium.

Consider the first-day returns on IPOs, which have averaged about 12% historically. This is obviously a huge return for those fortunate enough to get such shares, but unless you are Nancy Pelosi, or an investor who perpetually overpays for commissions, you don't get that one-day pop. That is, the investment banks capture this return by making investors overpay (or grant regulatory favors), so non-politicians don't capture any of this net net. The net money always goes to insiders, not passive investors, in hedge funds, corporate equity, corporate bonds, etc. Those big passive equity funds are like kids showing up to a trendy club and finding out that you need more than money to get in, you need connections, something unique to exchange.

As commissions and spreads have declined, the insiders are not making as much as they used to, so they can't afford to give away big returns to top-line stock indices. On top of low current long-term bond yields, this suggests weak equity returns going forward.

But then Inker returns to the intuitive theory of karma, and argues that equities have to generate a 6% real premium to compensate for the fact they are riskier than cash, noting their high volatility, and draw-downs during famously bad times like the Great Depression, WW2, and various financial crises. Yet it is pretty clear that risk is not correlated with higher returns within a variety of asset classes, such as equities, corporate bonds from BBB to C. Even GMO accepts this fact. Many people believe in a fundamentally inconsistent asset market: broad asset classes generate risk premiums that don't exist within these asset classes.

My solution, out in a book soon, argues this is because passive investing has zero risk premiums everywhere, when you look at assets in total. In any case, whether you believe in a large equity premium or not, there appears little reason to be in those highly volatile classes, unless you are like most people and overconfident about your stock picking abilities. This little conceit facilitates a rather large cluster-puck of mistaken assumptions, and I'm afraid most people will take the wrong lesson (eg, Joe Nocera's latest screed against shareholders), which if heeded will just aggravate our tendency towards more giant corporations with governmental influence, privileges, and resulting decreases in efficiency and competition. That will hurt GDP growth.


ed said...

I don't think you're being very fair to Inker's paper. The main point isn't about risk premia, and he doesn't make a big deal about karma or anything. Nor does he defend the CAPM, or predict higher returns for high-beta stocks, or anything like it.

He's simply interested in looking at where aggregate stock returns come from. Bottom line, returns to the index should be the earnings yield, minus a percent or two, plus anything that comes from changes in the P/E multiple. Thus he predicts that real returns over 10 years should be around 5 percent if there were no P/E change.

But he goes on to predict that P/Es will revert closer to their historical average, based on his guess that people will demand higher returns to own stocks, consistent with historical experience, and that this will bring returns down over the next decade, before they revert to a higher long-term level. This is pretty much a guess, and I don't even agree with it, but it is not the main point of his paper.

What I'm confused about is what YOU think future returns to be. As far as I can tell, you think that next-decade returns will be low, which would seem to imply you think P/E will shrink. But you also seem to think that long run returns will be low, which would imply that P/E will expand. Or maybe you just think that accounting earnings are completely unreliable anyway? Or something else? I don't get it.

Tim Worstall said...

I think Gross is using the wrong numbers there.

Wages and salaries is not the labour share of national income. You need to add in employee benefits (health care for example) and also employer paid taxes on employment to get that.

Then the labour share of income plus the profit share does not sum to national income. There's taxes minus subsidies and then also mixed income to consider.

So you cannot take the share of wages and salaries as declining and thus assume that the profit share is rising.

It might be, sure, maybe it even is, but that's not the right set of numbers to be using when trying to show so.

Eric Falkenstein said...

Tim: Yeah, that's always a big adjustment. It was the first I heard the argument, so I thought worth a post, but the labor-share argument is probably not related to long term equity returns.

cig said...

On the shareholder value story, I feel there's an unfortunate misunderstanding here. In the long term shareholder value and other stakeholders' interest are more or less aligned, a business that doesn't deliver stuff to customers and mistreats its employees and partner businesses usually won't go anywhere; it's an excessive focus on short term shareholder value -- where not delivering to customers and cheating suppliers matched with good PR can work, for a while -- that causes damage to everybody but (some) news traders.

Jack Brown said...

"Yet it is pretty clear that risk is not correlated with higher returns within a variety of asset classes". Perhaps it's the vantage point?

There are almost universally strong relationships between risk (premium) and (future) returns if someone measures risk by things fundamental yields (be it earning/price for equities or yield for fixed, etc.).

When fundamental risk premium is high enough (i.e. a low P/E market), the payoff profile is typically asymmetric to the upside. In this light, S&P 500 in 1999 is a vastly different risk than 1981.

Finally, markets with higher "average" fundamental risk premiums, higher average returns are typically linked.

Not sure I'd call it karma, but I think this is part value, part anti-EMH thinking.