Interest rates have been falling for 40 years for reasons that are unlikely to reverse anytime soon. This has already started to have profound effects on financial assets. Bonds, stocks and land prices are at record highs. Leading the way are long maturity bonds such as UK gilts which are up 52% YTD! What happens when supposedly “risk free” assets turn into some of the riskiest assets in the market?
Interest rates are the cost of money. If our free money world is here to stay than you can perhaps justify negative yields on over $13 Trillion of the world’s sovereign bonds with (presumably) no default risk, price-to-earnings ratios of over 100, and flattening yield curves. Lower for longer interest rates will put further pressure on the coming PensionCrisis and fundamentally change the risks and correlations of financial assets for many years to come. If, however, our free money world is just a temporary illusion then we may be in for a rude awakening.
Theories on why interest rates have fallen focus primarily on the neutral interest rate, R*, which is the rate consistent with stable economic growth that does not cause inflation to overheat. It should come as no surprise then that Janet Yellen, Ben Bernanke, Mervyn King, Larry Summers, Robert Kaplan, John Williams, Robert Gordon, Ray Dalio, Tylor Cowen, Robert Shiller and other leading minds disagree on the value of R* and its causal factors. However, nearly all the factors they point to seem unlikely to reverse in the near future such as global trade imbalances, aging population, income inequality, the pension crisis, global instability, high debt levels, and a 65 year trend of falling productivity.
By taking all of these leading theories and factors into account, Lukasz Rachel and Thomas Smith, from the Bank of England, construct a decomposition of the contribution of each causal factor. They conclude that R* is unlikely to rise more than 0.5% through 2030. However, near term changes in central bank policy, inflation expectations or a financial shock could still move observable rates substantially. Long term trends did not cause long term UK gilts to gain 52% this year, or drive corporate bond yields to record lows. However, before we can say anything about whether we are experiencing a bond bubble, or a rational adjustment to lower discount rates and fewer high yielding assets, we need first understand why interest rates fell, and that is the goal of this paper.
Falling interest rates, a history
Long term interest rates have been falling in the United States since 1981. This is true for treasury bonds as well as non-government debt such as fixed rate residential mortgages and corporate bonds. The longer history on 10-year treasury bonds and Aaa corporate bonds reveals that this trend is a reversal from the prior period following WWII characterized by rising rates. [Chart Link]
This trend of falling rates also applies to the rest of the industrialized world. The countries below are cited as evidence of this because they (presumably) have no default risk. This means any remaining yield can only be attributed to other risk factors such as inflation and the term premium.
High inflation can only partially explain the all-time high yield of over 15% on 10-year treasury bonds. A spike in oil prices in 1979 resulted in higher gasoline prices which in turn drove up prices for just about everything else. Paul Volcker is largely credited for bringing this elevated level of inflation under control (albeit at the expense of a deep recession) by increasing the federal funds target from around 10% to near 20%. The subsequent four decade drop in nominal interest rates is in part due to a steady fall in inflation. However, even after removing inflation from the 10 year treasury rate there is still a clear downward trend (green line = real rate). [Chart Link]
The key to understanding falling interest rates (and monetary policy for that matter) lies in understanding the “natural interest rate” (R*) which is the short term rate consistent with stable economic growth that does not cause inflation to overheat. This is illustrated in figure 1 below taken from Thomas Laubach and John C. Williams at the Federal Reserve. They explain that when observable interest rates are equal to R* (the natural rate of interest), monetary policy is neither spurring nor restraining economic growth. This is the point where the investment/savings (IS) curve intersects potential GDP.
Central bankers set policy based on where they think natural rates are; which is difficult because they are not directly observable. However, R* can be approximated by examining how economic growth responds to observable rates. Persistently high growth indicates that observable rates are below R* and vice-versa. This is the basic idea behind the Janet Yellen’s December 2, 2015 estimates below.
Here we see again that this is true across the industrialized world. Estimates below come from John Williams on August 15, 2016.
If we believe natural rates are around zero, as just about every model seems to indicate, then we need to rethink all this talk about getting interest rates “back to normal”. In the seven years since the Financial Crisis, natural rates have come down close to observable yields. The short term (3 month) Treasury bond is yielding 0.27%. Core PCE inflation in July was 1.6% so in real terms the short term bond rate is negative 1.33%; (-1.33% = 0.27% -1.6%). Therefore, rates would need to rise about 1.33% in order to reach today’s natural rate. Traditionally, a “normal” interest rate is one that is neither tight nor loose. From this perspective, a “normal” rate can be achieved by raising the Fed Funds rate up about 1%. This is in line with what Janet Yellen presented at Jackson Hole as the most likely trajectory.
Assuming the Fed meets its consensus projection, this would put us nowhere near what we intuitively think of as normal interest rates (ex. 4-5%). Why would the Fed be unlikely to ever move back up to a 4-5% range? In order to justify this kind of move the Fed would need to see one of two things: 1) natural rates would need to move up to this level and in so doing defy a 40 year trend downward, or 2) the economy would need to appear to be overheating, an implausible interpretation of the data given low RGDP, low inflation, and a strong dollar.
Not everyone agrees that R* is near zero. One notable exception is Esther George who thinks current rates were quite simulative, implying that R* is higher than the Fed’s models suggest. She argues that a healthy labor market seven years into the recovery and forecasts of inflation around the traditional 2% target are reason enough to raise rates. However, neither of these are direct measures of economic growth (i.e. GDP and by extension productivity) which may be why she appears to be in the minority.
Ultimately, very long term interest rates are driven by the supply and demand for money. Therefore, the long term drop must be due to shifts in supply and demand forces. Let’s review the leading theories:
Mervyn King argues in his new book, “The End of Alchemy“, that the long term fall in real interest rates is largely due to trade imbalances between “saving” countries (ex. Germany & China) and “debtor” countries (ex. US & UK). Supply of money from savings countries has essentially overwhelmed demand from debtor countries resulting in lower natural rates. He argues that R* will increase when savings and debtor countries implement reforms to bring down their trade surpluses and deficits. This will likely require more coordination between central bank interest rate policies rather than competition aimed at increasing export demand.
Larry Summers agues in his The Age of Secular Stagnation hypothesis that falling rates are due to an increase in domestic household savings (more supply) and a decrease in business investment (less demand) within industrialized countries. He attributes the increase in household savings to higher risk aversion in response to the Financial Crisis, and uncertainly in the face of the coming Pension Crisis. He attributes lower demand for business investment to our new economy, which he characterize as having more disruptive new technologies making it less likely long term investments will be profitable (ex. Platform Revolution, AI, robotics, cloud computing). Summers argues that R* will increase if industrialized countries like the United States stimulate growth through expanded fiscal policy such as investments in infrastructure. This demand for money increases R* directly by soaking up the existing supply of money, but also increases productivity which makes private business investment more profitable.
Robert Kaplan argues in his speech on Key Secular Trends and Implications for Monetary Policy that the aging demographics in advanced economies, slower productivity growth and the continued emergence of the U.S. as a source of safe assets have all contributed to the decline in the neutral rate (R*). An aging population lowers R* both directly (more demand for safer investments like Treasury bonds) and indirectly by reducing economic growth which reduces return on investment. Slower productivity is a long term trend as shown by Robert Gordon shortly. The scarcity of safe assets lowers R* for the same reason that Quantitative Easing reduces interest rates; it reduces the supply relative to demand causing the prices on these assets to get bid up as their yields get bid down.
Ben Bernanke argues in his Savings Glut hypothesis that global oversupply of savings has been driving down R* since at least the 1997 Asian Debt Crisis. This crisis drove up the demand from emerging market economies (ex. China, Saudi Arabia) and their central banks for bonds backed by strong and stable economies (ex. USA, Japan, Germany, UK). Bernanke argues that R* will increase if global savings drop. At the margin he believes this drop is more likely to occur if (1) China continues to move away from export dependence toward greater reliance on domestic demand, (2) the buildup of foreign reserves among emerging markets, especially in Asia, continues to slow, and (3) oil prices remain low, resulting in a decline in excess savings from emerging markets and oil producers.
Ray Dalio argues in his theory on “How the Economic Machine Works” that savings are high and investment is low because the industrialized world is in too much debt (see chart below). Economic growth can only come from two sources, cash or credit, but households, businesses and governments across the industrialized world have effectively maxed out their credit cards. The four decade long trend of debt fueled growth is over. As a result, households are saving more and spending less which makes businesses investment less profitable. The result is more demand for safe assets like Treasury bonds to invest the savings from both households and businesses. Government, however, are reluctant to issue more debt because they are already heavily indebted, restricting the supply of Treasury bonds which reduces R*. Dalio sees many parallels between today and the 1930s when governments and central banks were eventually forced to print money to pay for more fiscal spending as advocated by Summers.
Source: BIS 85th Annual Report 2014-15
Robert Gordon is widely recognized as an expert (if not the expert) on the nature and causes of long term productivity. He points out that productivity has been falling since about 1950. Falling productivity reduces return on investment and so it is only logical that the cost of money to make investments has fallen over time. Therefore, R* will rise once productivity increases.
Forecasting Long Run Interest Rates
Leading minds clearly disagree on the nature and causes of neutral interest rates. However, a closer look at their theories suggests that interest rates are likely to be lower for longer. This is the same conclusion that John Williams from the FOMC published earlier this month in his “Low R-star World“:
The underlying determinants for these declines are related to the global supply and demand for funds, including shifting demographics, slower trend productivity and economic growth, emerging markets seeking large reserves of safe assets, and a more general global savings glut. The key takeaway from these global trends is that interest rates are going to stay lower than we’ve come to expect in the past.
In other words, just about every theory explaining how R* has fallen since 1980 point to persistent factors that are unlikely to reverse in the near term. So until we start to see global coordination to reduce trade imbalances, higher birth rates, less income inequality, an end to the pension crisis, more stable economies and safe places in the world to invest, lower household and government debt levels, or new technologies that reverse falling productivity…real interest rates are likely to remain low.
Here are two forecasts that are consistent with my conclusion that R* is likely to be lower for longer. The first is for the United States. The second is for the world.
R* Forecast for United States
John Williams estimates in the same paper that over the next few years R* is likely toguide us towards a new normal of 3 to 3½%—or even lower. Importantly, this future low level of interest rates is not due to easy monetary policy; instead, it is the rate expected to prevail when the economy is at full strength and the stance of monetary policy is neutral. While current estimates of R* are around zero the trend has been rising. This is in part because the economy is not yet back to full strength.
In my view, the economy will be at full strength when the “Want a Full Time Job Rate” hits 6%. This is a metric I constructed to account for slack in the labor market not captured in the unemployment rate which you can find here and shown below.
You should take note that Williams projection of 3 to 3½%—or even lower R* is close to Janet Yellen’s FOMC projections of the nominal 4.5% or lower Fed Funds rate from Jackson Hole shown earlier. If inflation remains in the 1-1.5% range these estimates are right on top of each other. Therefore, I think it is prudent to emphasize the “or even lower” part of Williams projections. A 2.375% Fed Funds rate by the end of 2018 would be most consistent with the median view of FOMC members attempting to reach R* within two years. However, market participants have good reason to be skeptical (see chart below).
R* Forecast for the World
Lukasz Rachel and Thomas Smith from the Bank of England provides an intuitive and comprehensive decomposition and forecast of world R* which touches on every factor we have discussed including and some we have not like the singularity. You may disagree with some of their assumptions and conclusions (as do I), but their approach provides a useful baseline. Here is the abstract with my added emphasis in bold.
“Long-term real interest rates across the world have fallen by about 450 basis points over the past 30 years. The co-movement in rates across both advanced and emerging economies suggests a common driver: the global neutral real rate may have fallen. In this paper we attempt to identify which secular trends could have driven such a fall. Although there is huge uncertainty, under plausible assumptions we think we can account for around 400 basis points of the 450 basis points fall. Our quantitative analysis highlights slowing global growth as one force that may have pushed down on real rates recently, but shifts in saving and investment preferences appear more important in explaining the long-term decline. We think the global saving schedule has shifted out in recent decades due to demographic forces, higher inequality and to a lesser extent the glut of precautionary saving by emerging markets. Meanwhile, desired levels of investment have fallen as a result of the falling relative price of capital, lower public investment, and due to an increase in the spread between risk-free and actual interest rates. Moreover, most of these forces look set to persist and some may even build further. This suggests that the global neutral rate may remain low and perhaps settle at (or slightly below) 1% in the medium to long run. If true, this will have widespread implications for policymakers — not least in how to manage the business cycle if monetary policy is frequently constrained by the zero lower bound.”
Leading theories on why real interest rates have been falling since the 1980s point to many different causal factors. However, a closer look at these theories suggests that interest rates are likely to be lower for longer. Global trade imbalances, the aging population, income inequality, the pension crisis, instability in countries outside the G10, record high debt levels across the industrialized world, and a 65 year trend of falling productivity have all contributed to a rise in savings and a fall in investment. This has led to a rise in supply and fall in demand for money. The result is lower for longer real interest rates.
This suggests we should get used to seeing new highs in the stock market and other elevated prices for financial assets. Interest rates underpin the discount rates used to price all financial assets. Therefore, longer term financial assets such as bonds and stocks that pay a reliable dividend will have higher and more volatile prices. Historical pricing models and risk management tools may need rethinking. Low discount factors fundamentally change the risk/reward profiles of various financial assets. Low interest rates also reduces the availability of high yielding assets. This has already forced investors to accept lower returns and risk premiums as indicated in the bond yields vs default rates and earnings reversions vs equity price changes charts below and elevated Cyclically Adjusted Price to Earnings ratios (CAPE).
Inflation is likely to have a more prominent role in determining interest rate risk. A 1% rise in inflation expectations had little effect on bond prices back when bonds typically paid out a 10% coupon. Inflation rates across G10 countries have been low relative to historical levels. However, this would likely change if central banks adopt higher inflation targets such as the 4% target proposed by John Williams. For example, a Japanese 30 year bond with a coupon of 0.25% would lose 56% of its value if the discount rate used to price the bond jumped from 1% to 4%. Unlike the 50% stock declines during the Financial Crisis which recovered within a few years, bond valuations adjusting to a new regime of higher inflation may never recover because changes in inflation tend to be persistent.
Lower for longer does not mean that rates we observe today cannot rise precipitously in the near term. In today’s global economy, any large central bank can influence both inflation expectations and observed yields in all major currencies. The 800 pound gorilla right now is the pace of global quantitative easing which is currently consuming all new issuance of sovereign debt (see below). The impact of all these asset purchases on observed sovereign bond yields is obviously difficult to measure, but might be between 1-3%. The 1% comes from a study by Michael Joyce at the BOE which looked specifically at the UK’s QE from 2009-2010 (see chart B1.D). The 3% comes from forecasts Ben Bernanke developed to illustrate how interest rates may evolve as the Federal Reserve ends their QE program. His timing was way off. After his projection in March of 2013 the BOJ and ECB QE programs increased enough to completely offset the US tapper 2014-2015. However, if his forecast is accurate than flat projections of R* will be dwarfed by the rise in yields if/when the ECB and BOJ end their QE programs (assuming no one fills the void).
Sovereign Bonds traditionally viewed as “risk free” are now high risk assets if they have long terms. High duration and low credit risk premiums make 30 year bonds from the US, UK, Japan and Germany highly sensitive to small changes in yields. This includes the long duration UK gilts which are up 52% YTD and $13 Trillion in sovereign bonds (30% of G10 debt) currently at negative yields. If they can go up 30-50% in 8 months they can go down at least twice as fast. Lower for longer real interest rates may be the trend, but changes in quantitative easing or inflation expectations can have dramatic impacts on yields around these trends.
Stock buybacks will remain high if the cost of debt remains cheap relative to the cost of equity. Share repurchases (i.e. buybacks) are already near record highs. Many to criticized this strategy “stock price manipulation”, “an addiction to corporate cocaine”, “self-cannibalization”, and “an overwhelming conflict of interest”. These arguments hinge critically on the assumption that stock prices are too high to justify the move (because the same amount of money buys fewer shares when prices are high). Lower for longer interest rates turn this argument on its head for two reasons. First, whether stocks are cheap depends on your discount rate. Second, given slow global economic growth, buybacks still represent the best use of capital for many businesses. Thus, buybacks are a rational response to an environment of low interest rates slow growth, and we should not be surprised when corporate strategists act rationally and in the best interest of their shareholders.
Estimates of the coming Pension Crisis and related shortfalls in the Social Security trust fund are worse than expected. Pension funds typically estimate their annual returns to come in between 7-8%. Bill Gross and many other analysts view these forecasts as highly unrealistic. The Social Security trust fund is invested in assets pegged to the interest rates earned by medium- and long-term Treasury bonds. As households come to realize this, they will save more resulting in even lower interest rates. This will further reduce yields on bonds which will then further deepen the pension crisis and social security shortfall. This scenario does not bode well for bonds or stocks. With limited options for earning a reasonable yield I think it is likely that pension funds, sovereign wealth funds, and individuals will increasingly rely on leverage to squeeze returns from bonds as evidenced by the near doubling of treasury bond open interest since 2009.
Investors are also likely to seek refuge in the stock market and other risky investments like emerging market debt despite elevated prices and/or questionable fundamentals. The logic here is that low interest rates mean low discount factor on future cash flows allowing for increasingly high price-to-earnings ratios which Warren Buffet has said could be justified at over 100 in a world of zero interest rates. This problem with this argument is that it fails to account for the relationship between interest rates and corporate profits. The interest rate is the cost of money. When the cost of money is low it implies that there are fewer profitable investment opportunities for business. If businesses had profitable opportunities they would not be borrowing money to pay dividends and buy back their own stock. The chart below indicates that buybacks currently account for over 70% of net income. Direct payments to stockholders in the form of dividends account for about 60% of net income. How long can companies continue to pay out 130% of their net income to stockholders?
The answer is, of course, how long the major central banks can continue to keep interest rates artificially low. The problem is they can’t reduce them far below zero without further increasing the risk of asset bubbles. The Bank of Japan already owns about 2% of the Nikkei 225 and in so doing became the largest shareholder of several major companies. The ECB is likely to run out of eligible bonds to buy before March 2017. Some European corporations are now getting paid to borrow money.
Lower for longer may be the new normal, but that doesn’t mean asset prices can defy the laws of finance indefinitely. This time is not different. Interest rates will not stay below zero first time in the 5000 years of recorded debt market history. Corporate bonds cannot continue to ignore rising corporate default rates. Stocks cannot continue to increase dividends and buybacks while their profits shrink in a world saturated in debt.
Feel free to share your own thoughts on why interest rates fell. There is no better compliment you can give us than your thoughtful criticism. You can reach us at firstname.lastname@example.org, or follow us on Twitter @intuitecon
Disclaimer: These are our personal views. This article is not, and should not be regarded as investment advice or as a recommendation regarding any particular security or course of action. Our hope is that these observations will merely help you to more critically examine your own beliefs about finance and stimulate dialogue.