The US 30-year bond yield touched a new low last week below 2% while the 10-year yield is down to 1.5%. Yet US inflation is still solid in the 1.5-2% range, wage growth has been mildly accelerating and the economy has been growing at over 2% (Y/Y 2.6% and Q/Q2.1% to Q2). So what is going on?
For the 30-year yield to be as low as 2% with inflation running at least 1.5% suggests something is seriously out of whack. It seems to me there could be three possible explanations for the 30-year yield to be so low.
First is that the markets foresee much lower inflation with the risk of deflation for the next 10 years or more. But that explanation isn’t supported by inflation expectations – see below. Secondly, there may be a structural excess demand for long-term low risk assets which is driving yields lower. The implication is that investors are buying bonds blindly either because they don’t care about the yield or perhaps because of regulatory requirements. There is some evidence for this but proponents of this view often ignore the significant expansion of risk-free assets in the last 10 years as well as the decline in the current account surpluses in China and OPEC. Thirdly, perhaps investors foresee very low real interest rates for at least the next 10 years. As we shall see, this seems to be the main driver.
Let’s look in more detail at the three explanations. The 30-year bond yield has been declining for almost 40 years but was still in the 4-5% range before the GFC with inflation running at about 2%. Even after the crisis it traded at over 4% for a period but in recent years the yield has been between 2-3% and now is at 2%. So, we need to explain why the 30-year yield is about 2.5ppts below its pre-2008 level.
We can derive data from bond yields to show five-year forward inflation expectations. Inflation expectations have fallen from about the 2.5% level during 2004-14 (leaving aside 2008-11 during the worst of the crisis) to about 2% in recent years. Since October 2018 inflation expectations have fallen from about 2.25% to 1.9%. So, this seems to account for only about 0.5%, or a fifth of the fall in bond yields.
What about the second explanation – that there is a structural excess demand for long-term safe assets. This point has been made eloquently by Caballero and others (see here). Caballero argues that there has been a decades-long trend for the demand for safe assets to rise faster than the supply. Demand is rising in line with global GDP (including rapid growth in China) while the supply of safe assets is determined by the growth of a few developed countries, principally the United States. A related but slightly different argument is that excess demand for safe assets is primarily driven by regulatory requirements for banks, pension funds and insurers to hold more liquidity, naturally in the form of safe assets.
There is support for Caballero’s view that the supply of long-term safe assets did not keep up with demand in the late 1990s and early 2000s. China and other Asian countries ran huge current account surpluses following the Asian Crisis, as did OPEC countries after the oil price surged in the early 2000s. At the same time the US budget deficit closed completely under President Clinton and only expanded gradually in the early 2000s, which meant less Treasury issuance. The shortage of safe assets was partially filled by privately-engineered ‘safe assets’ like mortgage-backed securities. But during the financial crisis all forms of property-related instruments became suddenly viewed as highly unsafe, creating an acute shortage of safe assets. For a while, not just privately engineered CDOs but also agency backed securities were shunned.
Since the crisis however, government borrowing has surged and, despite quantitative easing, the supply of US Treasuries with the public has increased substantially. Meanwhile agency-backed mortgage securities are again seen as safe. Supply has therefore recovered. Further, while China did build up large US Treasury holdings in the early 2000s the total peaked in 2014 and is now back to 2011 levels while the OPEC surplus has vanished.On the other hand the demand for safe assets is likely higher than pre-2008. Lingering caution is one factor, but regulatory requirements on banks and insurers are surely important too.
There is no fool-proof way to measure the structural demand for safe assets (as opposed to that based on market expectations) but I find it unlikely that it is the major factor in the decline in long yields since 2007 though it might account for some of the move over the medium term. But 30 year yields today are about 1ppt lower than the average for 2017-18 while inflation expectations are little changed. There is no reason to think that structural excess demand has changed much in that period so at least 1ppt (of the 2.5ppts we are looking for) must be accounted for by changes in expectations and I suspect rather more.
So what about the third point, expectations of ongoing low real short term rates? If we take the US 30-year Treasury yield and subtract inflation expectations 5 year forward as a proxy for long-term inflation expectations then the implied real yield has plunged to only 0.2% today. We can also look at the real yield given by 30-year tips which is at 0.3%, much the same.
So, how to explain market expectations that real yields will remain very low for years? First, there is the very real risk of an economic downturn before long. Some see it as imminent, with recent data wobbly, including consumer confidence, often an early warning signal. The thinking goes like this. A new recession in the next 2-3 years will push rates back to zero. Then, since the scope for new fiscal stimulus may be limited by high debt it is expected that central banks will keep rates low for years afterwards (as they did after 2008). Central banks will likely also re-start QE and potentially resort to further measures such as helicopter money and its variants, which will all hold yields down. Some bears add that the next recession will likely bring a major fall in stocks and property, which would also constrain economic recovery (as it did in 2009-12) helping to keep rates low.
The second way to explain low real rates may be complementary as well as alternative. Over the last 9 months the Fed has moved to cut real rates to nearly zero despite the economy still growing. This just shows, so the argument goes, how difficult it is to sustain interest rates above the inflation rate for long. One could respond that the Fed is doing this because of the economic weakness caused by President Trump’s trade war, which may be temporary if an agreement is reached. But the counter-response is that the trade war won’t go away and, in general, domestic and global politics may continue to throw up damaging economic policies (for example Brexit and Trump’s tariffs).
My view on these two arguments is as follows. On the first, that a recession is looming, I disagree. While the economy is slowing I still think a recession in the near term is unlikely. The economic data recently have been mixed, but the leading indicators index has actually picked up. More on this in a coming post.
On the second, that the Fed’s actions prove that a higher real interest rate cannot be sustained, I think the Fed’s reversal since last December reflects two shocks. One, a realisation that the trade war is hurting worse than the Fed originally calculated. But, for that reason, I don’t think President Trump will persist with it to the point of (self-) destruction. Two, China’s slowdown this year has been greater than expected. I, along with many others, thought that the authorities would pull all possible levers to ensure the economy accelerated. In fact they have not, as I wrote in a prior blog. In particular, they have proved very reluctant to give any stimulus to the housing sector, which has long been a key driver of growth.
But these are both shocks, which means temporary effects. My guess is that Trump will settle the trade dispute somehow, as election year approaches. China could continue to slow but, if the trade war is settled it will get a boost. All in all, I think the US economy will keep growing and the bond markets will have to reassess. This does not necessarily mean yields returning to mid-2018 levels (just over 3% for both 10s and 30s) but some of the way at least.
Comments