You probably have been hearing the refrain that interest rates have nowhere to go but up for more than a few years now.Yet government bond yields are cratering not just in the United States but across the developed world. Germany and Japan now see negative yields through 7-8 years. Nonetheless, this is not a recent phenomenon since long-term nominal yields have been in secular decline for more than two decades now – the great bond bull market. Source : Brookings.eduOf course, tumbling yields are in no small part thanks to their respective central bank policies – Denmark, Sweden, Switzerland, the ECB and Japan have instituted negative short-term rates. Japanese 10-year yields fell to a record low below 0.10 percent after the Bank of Japan’s surprise announcement last Friday pushing short-term rates below zero after years of keeping them at the lower end of the positive range. In finance, the ‘ widowmaker ‘ trade – shorting Japanese bonds because of Japan’s gigantic national debt and a seeming top in their bond market, only to inevitably lose money – shows no sign of abating.We delve into some of the possible reasons behind the secular decline in long-term yields, beyond simply attributing them to central bank policy.
Underestimating the decline in long-term ratesEver since the great recession most prognosticators in the financial industry have been predicting higher interest rates, all the way back from 2009 , 2011 and more recently after the taper-tantrum in 2013 . The general meme is that the current regime of ultra-low interest rates just cannot last, and that the great bond bull market that began in the early 1980s will end any time now. In April 2014 , 100% of economists surveyed by Bloomberg expected the U.S. 10-year Treasury yield to rise in the following six months – rates fell instead. This inability to accurately forecast long term rates, even across a short horizon, is not new. As a recent report from the White House Council of Economic Advisors discusses, the Congressional Budget Office (CBO), financial markets and professional forecasters have consistently failed to predict the secular fall in nominal interest rates, focusing too much on cyclical factors. Source: White House Council of Economic AdvisorsAs we mentioned at the top, one can make a case that recent central bank policy is responsible for plunging yields across the developed world, but it still does not explain the secular decline over the past couple of decades. At the same time the argument is not quite as straightforward when you consider the United States, where central bank policy is rapidly diverging from other central banks around the developed world. The Federal Reserve is clearly confident that the economy has picked up enough steam so as to withstand the end of quantitative easing (October 2014) and liftoff from zero interest rates (December 2015). However, the yield curve has flattened significantly over the past two years. Source: Treasury.gov It stands to reason that if the Federal Reserve continues to tighten policy, as they did in December 2015, rates at the short end of the yield curve should go up (which we see above). The 2-year yield should essentially reflect what the market thinks 1-3-month rates will be two years from now. The 2-year yield ended 2015 at 1.06 percent, almost 40 basis points higher than where it began the year. The 1-year yield also finished 2015 about 40 basis points higher than where it started, rising to 0.65 percent.At the same time, the 10-year yield rose only 10 basis points in 2015, to 2.27 percent – at the end of 2013, the 10-year yield was 3.04 percent.Note that a flattening of the yield curve by itself is not indicative of future recession (unlike an inverted yield curve). Throughout the late ’90s, when the economy was expanding at a rapid pace, the yield curve was much flatter than what it is today. Source: Treasury.gov So what gives? Shouldn’t interest rates further out along the yield curve also move higher as the Fed raises rates along the short end (which they directly control)?
Lower expected inflationAs Ben Bernanke, the former chair of the Federal Reserve, explained last year, the yield on a long-term bond can be broken into three components:
Slowing ProductivityWhen you have strong economic growth, there is typically a lot of competition for investment dollars – for example, to build more factories, roads, houses and other tools to help business grow – and interest rates rise. Yet if economic growth is expected to slow (or reverse), there is not a lot of incentive to invest since the existing inventory should be able to serve current needs. Under that scenario, parking money in less risky assets like treasuries makes more sense and interest rates fall.Productivity growth across the developed world has been slowing recently and this is expected to continue according to OECD and IMF forecasts. Just in the U.S., productivity growth has been falling since the beginning of the last decade and is currently growing at less than 1% annually. Back in the 1990s, there was more capital spending and we got a lot of technological tools to improve productivity, but spending has fallen considerably since then.Robert J. Gordon, a distinguished macro-economist and economic historian at Northwestern University, has argued that recent developments in I.T. do not measure up to past achievements that powered economic growth between 1870 and 1970, namely: electricity, urban sanitation, chemicals and pharmaceuticals, the internal combustion engine and modern communication. In his new book “The Rise and Fall of American Growth”, Gordon points out that genuinely major innovations typically affect business practices in a major way. While the late-1990s and early-2000s saw some changes along these lines thanks to progress in I.T., there has not been much since.The direction and impact of future innovation remains unclear but interest rates will continue to remain at lower levels if productivity growth continues along its recent trend.
Demographic shiftsAlthough it is difficult to ascertain future technological progress and predict economic growth, it is reasonably clear that a demographic shift is taking place, especially in the developed world – with aging populations and slower labor-force growth.Why is slower population growth a concern? As this article from vox.com points out, if a country’s population stops growing, there is little point in building out new houses, offices or other infrastructure. While there will still be a need for servicing existing facilities, the total amount of investment spending will fall, which in turn puts downward pressure on interest rates. Recent academic work has indicated that demographic shifts can indeed push interest rates lower.Perhaps the best example of this is Japan, which has seen its population decline and experienced low economic growth and falling interest rates for more than two decades now. The country’s low fertility rate of 1.43 (total number of live births per 1,000 women of reproductive age in a population per year, and a good indicator of population growth) is among the lowest in the world. This also means that the population is aging. Japan has a very low unemployment rate and higher labor-force participation than the U.S., but essentially, the problem is that the country is running out of people to help keep the economy growing.In the following chart, we plot fertility rates across thirty-one OECD countries against the yield on their respective 10-year bonds (as of Dec 31st, 2015). Israel and Greece were removed as outliers due to idiosyncratic reasons behind the former’s high birth-rate and the latter’s high interest rates. Source: World Bank and investing.com .The rising regression line shows that fertility rates and long-term interest rates are positively related – countries with higher fertility rates generally have higher interest rates, and vice versa. The R-squared term of 0.21 indicates that fertility rate explains 21% of the variation in 10-year bond yields across these countries.The U.S. is among the few nations that has a fertility rate high enough to sustain population growth, and economic growth over the long-run. In turn, the 10-year yield in the U.S. is higher than most of its developed world counterparts. Similar to Japan, Germany is another example of a nation that has a low fertility rate (~ 1.41) that is insufficient to support population growth, and accompanied by depressed ten-year yields (sitting below 0.30 percent as of this writing).
Risk-off eventsAnyone in the financial industry, not to mention those who happened to glance across newspaper front pages during the last decade, cannot have failed to see the recurrence of global upheavals that have investors rushing toward safe, liquid assets.Tail-risk events – including events surrounding the financial crisis, the Greece debt crisis, Russia-Ukraine, political upheavals in the Middle-East and more recently China’s economic struggles – seem to be happening with more regularity. As Bernanke notes , treasury bonds offer an unparalleled combination of liquidity and safety during these episodes. They are also the asset of choice for foreign governments that want to hold large quantities of foreign exchange reserves. Consequently the yield on these bonds, and the term premium, is pushed down further.
What is a ‘normal’ interest rate?As we discussed at the top, commentators in the finance industry have been predicting higher rates ever since the end of the great recession and the customary thinking is that rates should snap back to their ‘normal’ pre-recession levels (when U.S. 10-year yields exceeded 4.50). At the same time, as the economist Neil Irwin writes : “… if you look at the longer arc of history, a much different possibility emerges. Investors have often talked about the global economy since the crisis as reflecting a “new normal” of slow growth and low inflation. But, just maybe, we have really returned to the old normal. Very low rates have often persisted for decades upon decades, pretty much whenever inflation is quiescent, as it is now. The interest rate on a 10-year Treasury note was below 4 percent every year from 1876 to 1919, then again from 1924 to 1958. The record is even clearer in Britain, where long-term rates were under 4 percent for nearly a century straight, from 1820 until the onset of World War I.
The real aberration looks like the 7.3 percent average experienced in the United States from 1970 to 2007.”The following chart shows interest rates between 1871 and 2015. The average over the entire period is 4.6 percent but the average between 1871 and 1970 is only 3.65 percent. Source : Dr. Robert Shiller’s online dataOur analysis here is not to say that interest rates will continue to decline over the next week, month or year but that there are broad, long-term forces acting to push them lower than we have seen in recent decades. We also see from the above data that low rates rates have persisted for long periods in the past.A lot of commentary on interest rates tend to focus on transitory factors like macro-economic policy, while ignoring longer-term structural factors like slowing productivity and demographic changes as well as the impact of recurring risk-off events. While the Federal Reserve and other central banks clearly have the ability to control short-term interest rates, this is not quite the case as far as long-term rates are concerned. Long-term interest rates are perhaps most dependent on the long-term inflation outlook, and the factors that normally boost inflationary pressure over the long-term seem to have all but dissipated.Akin to free-body diagrams in physics, which show the direction and magnitude of forces acting on an object in a certain situation, the following picture summarizes the forces currently acting to push interest rates up or down.