US 10-year government bonds reached an all-time closing yield low on 8th July 2016 of 1.37%. That just surpassed the prior record low of 1.43% in July 2012 during the depths of the Euro crisis. Prior to that, US 10-year yields had only been meaningfully below 2% in one year since the since 1880. That was in 1940 during WWII when yields briefly dipped below 2% - although only just. From the July low through to the Presidential election, yields backed up at a reasonably rapid pace. Since the election, the speed of the back-up has accelerated – such that yields are now trading just shy of 2.40%. Determining the direction of bond yields from here is critical.
The US 10-year government bond yield is the global risk-free rate. Most other assets are in some way priced relative to that yield. Recent years of investing have also been dominated by the ‘hunt for yield’ theme. If this is a genuine ‘generational’ turn in bond yields, that theme may now be finished. If correct, then considerable adjustment of portfolio positioning will be required over coming months and years. As such, this is a critical question.
While it’s clear from technical indicators that bonds are now deeply oversold, there are a number of key reasons why it’s likely that generational lows are in place for US 10-year yields and that the trend in yields will now be upwards for many years, even plausibly for a generation. Those four key reasons are laid out below.
1. The velocity of money is accelerating: There’s growing evidence that the US and UK banking systems are starting to operate normally once again. A number of pieces of data support that hypothesis: In the UK money supply growth has accelerated to levels not seen since pre-GFC (fig 1a); the UK aggregate commercial banks’ balance sheet is expanding for the first time since the crisis: UK mortgage equity withdrawal, having been negative and range-bound since the GFC, has accelerated from late 2014.
In the US, the commercial banks' aggregate balance sheet has been accelerating in 2016 despite shrinking commercial banks reserves at the Fed (fig 1b). In prior instances in this economic cycle, when reserves at the Fed have shrunk, the growth of commercial banks’ balance sheet has also slowed, and as such, the Fed has needed to restart QE. This autonomous expansion of the commercial banks’ balance sheet, without Fed QE support, points to healing and increased risk appetite at the commercial banks. That suggests the beginning of a normalisation of the money creation system and an acceleration of money velocity. As figs 1 & 1c show, there’s (rightly) a clear correlation between money velocity (approximated in these charts by the monetary base as a % of GDP) and the level of US interest rate, whether short or long dated. As money velocity accelerates, growth & inflation pressures will pick up, and the Fed will need to respond by raising rates. Since the GFC, the opposite has occurred. That is, money velocity, up until recently, has been depressed/subdued.
Consequently, the Fed and other global central banks have kept rates at emergency low levels to support growth as much as possible. Low money velocity reflects deleveraging and build-up of increased capital at banks. With UK & US banks now re-leveraging, money velocity and with it, money supply, is picking up and rates, therefore, need to rise, albeit gradually.
2. Post crisis normalisation: In their extensive analysis of financial and banking crises across time and across multiple economies, Rogoff and Reinhart found that economic activity post banking crises is typically markedly subpar for the first 6 to 7 years of the recovery. Since the US & UK GFC recessions ended in 2009, their recoveries are both now seven years old. Rogoff & Reinhart’s analysis is, therefore, consistent with the recent reacceleration of money velocity.
3. Switch from ‘Monetary’ policy support to ‘Monetary & Fiscal’ policy support for the economy: While hard to quantify, the switch to fiscal stimulus to complement monetary stimulus has accelerated markedly in the past few months. Governments and central bankers across the globe have concluded that monetary policy has largely reached its limits. As such there has been a shift in emphasis to utilising fiscal policy to support monetary policy stimulus. Japan, China, the UK & the US have all either increased actual fiscal stimulus or indicated that it will happen. Fiscal stimulus, in contrast to QE, is more likely to generate consumer price inflation, whereas QE has primarily generated asset price inflation.
4. (prior) Bubble-like valuations: There’s a strong argument that bonds by the middle of 2016 had reached bubble-like valuations. Negative government bonds yields for medium and long duration bonds point to extreme valuations. Paying a government (i.e. a negative yield) to lend them money is irrational (albeit there are ways of justifying/rationalising the decision). Equilibrium bond valuation techniques also point to overvalued government bonds. One method for valuing bonds is to calculate the equilibrium 10-year nominal yields using the aggregation of an economy’s trend growth in supply side factors (i.e. trend inflation plus trend productivity growth plus trend ‘hours worked’ growth). Aggregating those factors for Western economies leads to a fair value long term 10-year yield of between zero to 3%. Trend productivity growth for Western economies is arguably between zero to 1.0%; Trend inflation is between 0.5% to 1.5% (on a conservative forecast); while trend labour force growth is between -0.5% to +0.5% p.a. A number of Eurozone bonds sit at the low end of that range, while the US (and to a degree UK) sit towards the higher/top end of that range. Of note, trends in productivity growth are a key unknown in that forecast. If the UK & US economies do start to re-leverage, then growth (and with that trend productivity as well as inflation) will also probably re-accelerate.
Conclusion: With the beginning of the normalisation of the money creation system in the US and the UK, history would suggest that UK & US yields have made their generational lows. That suggestion is supported by ‘rough and ready’ valuation analysis which points to July lows as bubble-like valuations. While those lows are likely passed, the oversold nature of bonds highlights the likelihood of a rally in the near term. Price action during that rally will be instructive. Beyond that oversold bounce, though, we would expect bond yields to resume their move to higher levels in early 2017.
Written by Chris Watling, CEO & Chief Market Strategist at Longview Economics: (VIEW LINK)