Friday, June 16th, 2017
This week’s commentary outlines the investment consequences of the ultra-low interest rates that have prevailed in the United States (and much of the developed world) over the past 15 years. These interest rates have been set by the US Federal Reserve (the Fed) in response to rolling financial crises from 1999 to 2010. Since that period, the maintenance of ultra low rates have created consequences – some of which we explore here.
One unintended consequence of sustained low interest rates is the faster growth and advancement of technology. Simply stated – lower sustained interest rates have greatly benefited the development of IT companies. This trend of the last 15 years is unlikely to change any time soon.
The higher growth of technologies (from medical sciences to physical metallurgy to renewable energy et al) is real and investors need to pay more attention to these companies. We perceive that some technology companies may grow at exceptional rates as a “positive feedback loop” develops.
We believe that inflation will remain low for longer and thus both interest rates (generally) and bond yields across the globe will also remain low for longer. Central banks have been and will be too slow in adjusting rates to at least meet inflation. This delay is now compressing inflation.
These developments have consequences on the pricing of all asset classes, especially equities and suggest that investors need to rethink their criteria for identifying growth and should value cash yield more acutely.
A starting observation
The latest core US PCE deflator (Fig. 1) which measures personal consumption expenditures, registered an unexpected fall in March. This was well below consensus forecasts and contradicted the common view that inflation and the cost of living was rising in the US.
The sharp lift in long-term bond yields that followed the election of President Donald Trump was halted and reversed. The flow of PCE data suggests that the cost of living for US households was not rising and continued to be below the reported core CPI rate. Concurrently wages growth had moved above both actual and expected inflation.
Figure 1. US various inflation data since 2007
Source. NAB, Macrobond
Other charts for consideration
The above demands a deeper analysis of what is occurring in the US. Bond yields in the US remain at historic low levels and their persistent behaviour (yields barely matching inflation across the yield curve) suggests that longer-term inflation expectations are not rising.
The following charts review the longer-term data on US unemployment rates, participation rates and wage growth. Ultimately it is employment that drives wages and thus household income. There are long-term trends in employment that are becoming significant. In May, unemployment in the US hit 4.3%, a low not seen for 16 years, and a figure which has been falling steadily since 2009. Inflation meanwhile has remained below 2% and shows few signs of picking up in any meaningful fashion. And thirdly, the participation rate has seen a big decline in the proportion of working age Americans in the labour force.
Figure 2. US Unemployment rate since 2000
Source. US Bureau of Labor Statistics
Figure 3. US employment participation rates since 1980.
Figure 4. US wages growth since 1985
Source. US Bureau of Labour Statistics, Bloomberg
Narrative and interpretation:
There is a financial narrative developed in the 3 charts above and it helps explains the fundamental changes we are witnessing in inflation, interest rate and bond yields, et al.
The implication for the equities market and investors is significant. They suggest the cycle we are in developed 15 years ago and the traditional concepts of growth and value is being challenged. Many Industries are being challenged by technology. Disruption is high and sustainable return on equity is subject to much conjecture.
What do the charts above suggest? Our interpretation is as follows:
- The recession of the early 1990s (in the US) was possibly the last of a traditional cyclical nature. We can observe the normal recovery of the US economy – exhibited by growth in employment and participation rates (Figs. 1 & 2) – that was followed by wage growth (Fig. 3). Twenty five years ago, a younger population (median age was 32.9) was relatively more energetic and eager to work as a whole. In the early 1990s, the impact of technology across industries that had been afflicted by recession was relatively insignificant and in its infancy. The internet and “fast computers” was solely in the domain of University research departments and specialised labs. The US recovery subsequently occurred just like previous recoveries had.
- During the 2000 “dot.com” recession, the unemployment rate did not lift to a normal recession level, peaking at 6% in 2002. However, underlying this was the beginning of the drift in the participation rate (Fig. 2) – the proportion of the economy’s labour force that is actively employed or actively looking for work. The impact from technological developments was beginning to be noticeable and while the real economy and households held up, the 2000 recession hit financial markets hard (especially tech stocks on the Nasdaq). The tragedy of 9/11 in 2001 resulted in the Fed being more accommodative than it probably should have been. The Fed excessively responded to Wall Street’s pain and interest rates were pushed lower for longer than was required (Fig. 5). The seeds were being sown for the dramatic growth in IT companies.
Figure 5. US Fed rates over time
- The sustained low interest rate period of 2001 to 2006 gave great support for technology companies. They were nurtured and began to thrive. Presented with low interest rate alternatives, investors were encouraged to have patience with loss-making tech heavy companies. Low interest rates ensured that equity risk capital became plentiful. IT companies were given time to improve, restructure and reinvent themselves. It was a low profit cycle that was the genesis for the development of the major IT multinationals of today. For a critical period, it was less speculative and more patient capital that flowed after the dotcom bubble.
- The adjustments to cash rates in 2005/06 were undertaken too late by the Fed and financial events overtook policy again from the middle of 2007. The major technology companies were flush with liquidity and the 2008 GFC had little effect on their financial viability. The significant short-term adjustment to cash rates from 2006 was sharply reversed in the GFC and that is where they have sat since. Technology companies were given more oxygen to survive and prosper.
- Post GFC, companies like Apple, Amazon, Tesla, Facebook, et al continued to either enjoy sustained strong capital support or had built sustainable cash generating businesses. Further, if required, they were afforded tremendous debt support as fixed interest investors aggressively bought their corporate bond issues.
- In the real economy (the “old” US economy) the GFC created massive job losses and unemployment spiked (Fig. 2). The participation rate dropped as discouraged workers pulled out of the job market. Since that point and unlike previous post-recession periods (prior to early 1990s), the recovery from the GFC in 2010 saw the participation rate continue to drop as employment recovered.
- Post 2010, the US has moved into the “baby boomer” retirement period. The median age in US has grown from 32.9 years in 1990 to about 37.5 years (2017) and is trending higher. It is likely to stabilise in 2040. Today about 10,000 people in the US turn 65 every day (3 million per year) and this will continue for the next 10 years. This demographic trend will act to further curtail consumption and inflation.
- The current and forward cycle has this demographic trend. Retired people tend to spend much less than the working population. Research confirms this in the US and also provides data on the rate of this fall. It suggests that for adults aged 60 years and over, there is an incremental drop in spending by about 20% per decade. This increases to 30% per decade for those aged over 80 as reported by United Income.
Longevity has increased steadily since the middle of the 20th Century.
Figure 6. Generational Life Expectancy
The current cycle
Today, we witness a unique period whose genesis was created from a period of sustained low interest rates – overlayed with demographic changes. We observe:
- Technology (and particularly the internet) has disrupted many traditional high employment industries. After a decade of sustained growth, new technology companies have gained sufficient traction to achieve “escape velocity” so that the drop in the participation rate in “traditional jobs” reflects jobs that are permanently lost. Workers have not re-entered the workforce as the skill base has moved;
- That whilst “traditional jobs” have been lost, there are many new jobs created in the “new economy” of technology-related sectors causing the unemployment rate to stabilise.
- Increasing part-time and the “casualisation” of the workforce. Increasing numbers of service jobs (rather than production). This trend of increasing casualisation and creation of part-time jobs will continue at the expense on full time jobs resulting in job data becoming a less reliable indicator of the strength of economy.
- An increase in single earners within a household. Similar to 1940 / 1950s era where the unemployment rate was in the ~4% mark with most households having one full time worker resulting in a lower participation rate overall. The single income family will continue to increase for the foreseeable future as the labour force is becoming more agile.
- The first phase of baby boomer retirement period (2010 to 2020). In this period, it is likely that the leading earner (usually the male worker in the older economy) will retire first. Data suggests that this trend where more males are retiring while the female counterpart is going back into the work in a part-time /casual capacity. As a result, the participation rate has continued to drop to a low of 62.9%, a figure not seen since the 1940s.
Based on all the trends described above, slower growth will endure for a sustained period. The US trends are being observed across the developed world.
This implies minimal wage pressure in the interim and in turn implies a low inflation rate. As the inflation rate is the biggest influence on interest rates, our view is that cash interest rates and longer duration bond yields will remain at historic low rates for the foreseeable future.
So where do we see growth?
To summarise, we perceive growth in the following areas:
- Large well established IT or branded companies. The positions of Facebook, Amazon and Google (for instance) appear sustainable;
- Servicing the needs of the ageing population – the western world is ageing fast.
- Servicing the wants of the top 1% to 3% of high net wealth households. Central Bank policies are making these people wealthier and they have massive discretionary expenditure.
- Offer services or products that also focus on the upper middle and middle classes of China and the rest of Asia. China and parts of Asia are creating affluent classes that are emerging quickly in contrast to the low or anaemic growth of the western world.
- The low and ultra-low interest rates of the past 10 to 15 years have sowed the seeds for the world we now are experiencing. The technological developments of the last 20 years have been stimulated much like World War 2 stimulated electronics, energy and productive capacity.
- The modulating of the PCE indicator in March 2017 (see Fig. 1), lower than expected US 1Q GDP growth and the long-term decline in the participation rate and the trend in wage growth over past decades, suggests mid to longer-term inflation will remain muted and interest rate will remain low. The longer-term bond curves reflect this.
- We expect that the 10-year US bond and the 10-year Australian Bond will struggle to move above 3% in the next few years. This means that required returns, driven from low risk-free returns, will remain at low levels for a while longer. PERs will remain elevated, but low growth companies will continue to be punished as markets demand higher yields.
One last thought: Technology improvement in renewable energy sector
In the mid-1990s, a solar battery retailed at approximately $6 /watt which was well above the $1/watt from the existing retail energy cost from the established grid. In 2010, the cost dropped to about $2.80/watt installed, which was still not competitive to fossil fuel. Despite that, the NSW Government at that time, came out with the $0.60 Kwh gross feed-in, thereby making the installing solar panels on a residential roof a no-brainer with a payback between within 3 and 4 years.
- The cost for installing solar panel dropped by roughly 5% pa between 1995 and 2010, a period to 15 years.
Following massive investment in solar manufacturing and development, the cost of these solar panels has dropped below $1 / watt installed (i.e. US$0.75/watt). With the anticipated increase in grid energy cost, it is now economically sensible to install solar panels on most roof tops. Indeed there are now numerous community solar panel installations which offer a return of more than 7% pa even without considering any environmental benefits. That is the reason why community solar projects in the order of $1 to 2 million or less were snapped up by investors within minutes of being offered.
- The cost for installing solar panel dropped by roughly 12.5% pa between 2010 and present, a period of ~6 years; more than double that of the ~15 years period between 1995 and 2010.
- The reduction in cost is accelerating at a faster rate and it is likely that the ultra-low interest has assisted in the investment in the renewable energy and the ultra-low interest rate post GFC has accelerated it even more.
Fossil fuel (coal, oil and gas sector)
The world consumes about 95m bpd of oil currently and forecast to go higher to more than 110m bdp by 2040 before it faces a decline. However, it is doubtful whether this forecast is correct given the cost of newer forms of energy are rapidly declining. The world currently has more than 2T barrel of probable and recoverable oil reserves at current oil prices. At 100m bdp consumption, it would take more than 55 years to consume all known oil reserve before further exploration. Current known and recoverable oil reserves (at current price and given available technology) could last 50 years.
But technology is likely to develop driverless and electric cars into full production and use by 2030. Currently 45% of the 95m bdp are consumed by all forms of transportation, and our guess is that cars equate to roughly half this amount.
Logic suggests that on the first order approximation you would not consider putting any investment capital in fossil fuel or companies that support and service this sector.
The above is simply a short narrative on just one small facet related to technology for energy. There are others spanning from bio-science to physical sciences or metallurgy. All of us should be excited about the future for investing for growth, but we need to carefully re-think how we look at the world before we allocate capital.
An important indicator the FED pays attention to closely when they set interest rate for the future.
 I should add that this is only an opinion and is not a theory in the context like physical sciences but I do feel confident that my narrative can probably explain 70% to 80% of what is occurring. However, we are speaking about finance and economics which is not really science.
 The mind set (ambitions, views and way of life) of this group of young people is probably different from the younger cohort in 2017 as well as the whole population is older now than 25 years ago. This important distinction should borne in mind.
I have deliberately placed inverted commas here because the fast computers in those days were actually very slow compared to today’s standard.
I was one of those who had access to the Internet in the early 1990s.
The US Fed was possibly overly worried about the fragility of the financial system coming so shortly after the dot.com bubble.
 Based on US Censors.
Matt Fellows, “Living Too Frugally? Economic Sentiment & Spending Among Older Americans”, p. 14 United Income. The outcome is based the University of Michigan Health and Retirement study.
 The PCE indicator is one of the most important inflation indicators the US FED uses to determine interest rates.
 Somewhere between $2.30 and $3.30 per watt on rough terms depending on where the solar cells and inverted come from and capacity.