When looking at variations in the short and long-end of Treasury curves in Europe versus the US, Goldman's Francesco Garzarelli made a remarkable observation: whereas market expectations of the trajectory for monetary policy in the US and Euro area continue to diverge, manifesting in a growing delta in short-end yields, the correlation of returns on long-dated bonds on the two sides of the Atlantic remains very high. The explanation for these cross-country dynamics, according to Goldman, can be traced back to the behaviour of the term premium, which is technically defined as the excess yield compensation investors usually require for holding long-dated bonds, but in practice is the umbrella "fudge factor" term used to "explain" unquantifiable central bank impacts on longer-dated bond prices and moves.
These two dynamics are shown in the charts below: while short-rate expectations continue to diverge between the US and Europe (left), the term premia in the US and the Euro area bond markets have tracked each other increasingly closely, especially since 2014 (right).
Whereas several years ago there was a broad disageement over what is the primary driver behind the depressed term premium, gradually the analyst community was forced to admit reality, and accepted that the term premium is merely another way of calling central bank intervention on bond prices. Indeed, Garzarelli admits as much, saying that "low term premia reflect both the present macroeconomic environment and central banks’ actions."
As Goldman futhers explains, while "historically, the premium tends to decline in economic expansions, when investors are less wary of bond price fluctuations. Low consumer price inflation, particularly when associated with greater confidence that it will remain contained, also acts to depress the term premium. But there is more to the decline in term premia during this cycle than can be ascribed to the growth and inflation outlook."
The chart below from Goldman shows a growing undershooting of the aggregate term premium on 10-year bonds in the major economies from where historical relationships with macro factors would have it. "This departure from historical norms coincides with the introduction of negative rates and sizeable purchases of long bonds by the ECB and the BoJ." In other words, the nearly 1% delta can be attributed to the actions of one or more central banks.
The above observations suggest that fixed income investors expect policy rates in the major blocs to continue heading in different directions, at the same time thanks to central bank yield suppression, they continue to search for yield and remain on the look-out for the highest term premia across countries. Here's what that looks like mechanistically:
When a central bank bids up long-dated bonds through a combination of negative rates and QE, the ensuing compression in the term premium on domestic bonds spills over onto other major fixed income markets where the term premium is the highest. After the start of the ECB’s QE in 2015, for example, the Euro area experienced substantial net outflows of long-dated bonds, which started to partially reverse this year as the term premium on US Treasuries picked up in relative terms.
The issue, as Goldman explains, is that such capital flows pose challenges for central banks. Confirming something we have repeatedly shown, namely that despite 3 rate hikes, financial conditions in the US have reacted as if the Fed has cut rates three times, as long as foreign QE contributes to keeping the term premium on US Treasuries low, the Fed may need to lean more than usual on short-term rates in order to tighten domestic monetary conditions. This means that if the Fed is indeed concerned about asset bubbles, it will be forced to tighten substantially more than the market expected.
By contrast, in order to preserve monetary accommodation, the ECB will be required to keep a firm grip on the front end of the EUR yield curve should long-dated Euro area bond yields be pulled up by a rebuild of global term premium resulting from the Fed’s ‘quantitative tightening’. This logic has been behind Goldman's expectation for a flatter term structure of US rates, and a steeper one in the Euro area. Beyond fixed income strategy, we see three further implications for macro investors stemming from the unusual dynamics in global term premium.
What are the implications of this divergence for traders? There are three.
First, as Goldman explains, international bond spillover effects are worryingly large. Indeed, "the international co-movement in term premia has now reached levels that warrant attention, and in particular the uncertainty over the inflation outlook is gradually increasing (smaller output gaps, higher commodity prices). Reflecting an unusually high degree of cross-country term premium spillovers, the diversification benefit of international bond portfolios has declined considerably. Moreover, empirical evidence shows that the higher price of long-dated bonds has made them in greater demand by institutional investors such as insurance and pension funds. When both the demand and supply of bonds are an increasing function of the price level, multiple equilibria can arise and elicit unanticipated large changes in yields.
Needless to say, when (rising) price becomes the only variable behind purchasing decisions, any reversals could have dire consequences. And, just to make sure the warning is heard, Goldman cautions that "in the current market environment, these yield shocks may propagate more quickly across the advanced economies than was the case during the ‘taper tantrum’ episode in 2015." This means that once the selling in the long-end begins, the consequence could be far more severe than the sharp selloff observed in one or more previous "taper tantrum" episodes.
* * *
The second implication is especially relevant for currency traders, because FX has become more correlated with relative term premia: As noted by ECB Executive Board member Coeure in a speech last Friday, in recent months the Euro-US Dollar FX cross has moved more closely with the differential in the Europe-US term premium than the differential in interest rate expectations.
Global fixed income investors were attracted to US bonds after a rebuild of term premium in the wake of US President Trump’s election (rate expectations rose as well, but to a lesser extent). Higher demand for US fixed income progressively eroded the US term premium, at a time when expectations of ECB tapering were starting to build. This swung the term premium differential further in the Euro area’s favour.
Then, more recently, the start of quantitative tightening in the US and the extension of quantitative easing in the Euro area rebalanced the term premium differential towards the US again, strengthening the Dollar. Deprived of (expected) yield, global investors are then forced to chase the highest premium that central banks afford them, affecting currencies in the process, in Goldman's view.
* * *
But it is the third, and most important implication, that is of particular note as it goes to the very core of the "circular reflexive" conundrum that has been plaguing the Fed, which according to Janet Yellen has been unable to explain the "mystery" of low inflation 10 years into this so-called recovery. The reality, is that there is nothing mysterious about suppressed inflation: in fact, it is the all Fed's own doing, and as a result of trillions in liquidity, "the information content in long bonds (and inflation) is low" in Goldman's words. The bank explains:
By amplifying the compression in the term premium on nominal rates, QE may have also created distortions in the breakeven inflation market. To be sure, inflation expectations derived from both surveys (consumers, professional forecasters) and market measures (forwards and options) are low, reflecting a protracted period of low realized inflation prints in spite of expansionary monetary policy. The uncertainty around inflation forecasts has also declined, all arguing for a lower inflation ‘premium’. Terminal real rates (R*) in advanced economies have been on a downward trajectory over the past decade, and have been affected by common factors, but they have been relatively stable in recent years, reflecting a lower frequency of real-side shocks. Against this backdrop, QE (and the price amplification dynamics it has set in motion by upsetting the demand-supply balance for long-dated bonds) has tightened the positive relationship between the term premium and inflation forwards.
The punchline can be taken right out of any Soros book on market reflexivity between cause and effect, to wit:
"this circularity – QE contributes to depress longer-dated inflation forwards, which in turn encourages central banks to deliver more QE – has lowered the information content that can usually be found in long-dated fixed income instruments."
That is Goldman's polite way to say the Fed's QE has broken not only the bond market (i.e., its "information contnent" is non-existant as it is entirely at the mercy of central bank liquidity), but also forward inflation measures, which is terrifying as it is these very measures that the Fed gauges in determining whether to do more QE. The perverse circularity is the daily bizarro world market participants have become all too familiar with, and boils down to the following: the more QE the Fed does, the lower inflation breakevens slide, the lower yields drop, the more QE the Fed believes it has to do, and so on in a self-reinforcing feedback loop, one which has now continued for 9 years because the "smartest people academics in the room" have been unable to figure out just how they broke the market.
And the final implication: since bond yields are artificially low - whether due to chasing term premium, or because of the Fed's perverse circularity - it means that the only justification for 20x P/E multiples, i.e., low rates (as per the Fed model) can be thrown out of the window. Of course, if that happens, both the bond and stock markets would crash... which is also why nobody at the Fed will ever be willing to openly admit what Goldman just said.