Beware of bubble talk, for now

In a wide ranging speech delivered overnight in New York, Glenn Stevens, provided more colour on his prior communications in which he has bemoaned the low level of entrepreneurial risk taking across Australia's corporate sector and implored businesses to invest for future growth prospects.

I have long argued that animal spirits remain dormant in Australia (and across most of the developed world) thanks to persistent revenue headwinds and high discount rates that companies are assigning to expected future cash flows for potential new projects.  Moreover, companies continue to cater to investors' insatiable appetite for income by lifting payout ratios.

At first glance, the persistence of a high cost of capital is difficult to reconcile with the risk free rate or yields on sovereign bonds being at record lows.  But Mr Stevens has acknowledged that lower returns on safe assets have not pulled down the cost of capital because there has been an offsetting rise in the equity risk premium.  The RBA Governor argues that the stability of earnings yields in the face of declining long term interest rates implies that the risk premium has lifted, reflecting more risk being assigned to future earnings and/or lower expected growth in future earnings (see chart).  Mr Stevens also highlights anecdotal evidence of stickiness in hurdle rates used by corporate Australia.

I have previously argued that long term interest rates are low precisely because the equity risk premium has shifted up (see the blog post from 26 November 2014 - Low interest rates are here to stay but don't expect stocks to re-rate).  That is, investors have gravitated towards safe assets as they have perceived a still high level of risk in the world.  From a capital budgeting perspective, the risk free rate is low due to factors that are restraining corporate investment: low expected earnings growth and a high expected risk premium.  Consequently, stock multiples should not get an assist from a lower risk free rate given a higher ERP and lower growth prospects.

A behavioural perspective on the risk premium and cost of capital

The common misconception that record low interest rates should be associated with a re-rating of stocks stems from the availability heuristic that is a cornerstone of behavioural finance.  Events that are memorable, directly observable and have a high valency are more likely to resonate with us.  The risk free rate is directly observable on a daily basis, so our familiarity with low long term interest rates informs our mental framing of the cost of capital.  But the equity risk premium is not directly observable, and so we are likely to neglect it as a source of variation in the cost of capital.  Just because the ERP cannot be directly observed, does not mean that it should be ignored.  

In fact, from the perspective of monetary policy, the ERP is a key channel in which a central bank can influence the psychology of risk taking.  If a central bank is seeking to revive animal spirits in the corporate sector, it must do so by reducing the expected equity risk premium (as well as lifting expectations of revenue growth).

If the lift in the risk premium has broadly offset lower returns on safe assets, then the effect on the multiple of the market ought to be neutral.  But this masks important sources of cross variation between high and low beta stocks.  High beta stocks are associated with greater sensitivity of their cost of capital to shifts in the risk premium.  For instance, a lift in the risk premium from 4% to 5% causes a high beta stock's cost of equity to rise by 150 basis points, while the cost of capital for a low beta stock rises by less than 50 basis points (see chart). 

While not shown here, mathematically, the percentage increase in the multiple of a high beta stock exceeds that of a low beta stock in an environment where the risk premium is rising, other things being equal.  Moreover, it is reasonable to think that low beta stocks have experienced a larger compression of their betas in recent years than their high beta counterparts, as investors have assigned lower risk to the future earnings of defensive stocks. 

Indeed, the pattern confirms that the defensive sectors in the Australian market are trading at a premium (based on dividend multiples) to their 10 year median estimates, while cyclical sectors are trading at or below their historical medians (see chart).  While lower expected earnings growth has contributed to this trend, we believe that the higher equity risk premium has played an important role in driving a wedge in the multiples between high and low beta stocks.

Beware of bubble talk, for now

It might be tempting to think that either defensive stocks or high yielding stocks with strong and sustainable sources of competitive advantage are exhibiting bubble like valuations.  But shifts in the risk premium and cost of capital matter for portfolio construction, and our analysis suggests that they have contributed to the relative rise in valuations for low beta or defensive stocks.  I expect those high valuations to persist while the corporate sector's animal spirits remain dormant.  The unwinding of the defensive trade will eventually happen abruptly, but that remains way off as long as output remains below potential, there is deficient aggregate demand, unemployment is well above the natural rate, and inflation remains subdued.