In recent weeks I’ve been fortunate to interact with dozens of financial advisors from around the US who have a unique perspective on both investing, and on the real economy. I found that these professionals more and more agree with the notion that the Fed will struggle to raise policy rates materially, but at the same time there is also still a seemingly innate distrust of the Treasury market at current levels of yields. This week I want to take a look at the linkages between US Treasury yields and future monetary policy to try to reconcile this dichotomy. 10-year Treasurys at 2.05% are low-yielding, but so too are the policy rates from which the Treasury yields ought to be derived.
The 10-year US Treasury yield, in theory, should reflect the market’s expectation of the average overnight Fed policy rate for the next 10 years, plus a risk premium. In the past I’ve described that risk premium (called “term premium” in the interest rate market) as a catch-all fudge factor, and initially here I’m going to assume it to be fixed at zero, though we’ll come back to it. For purposes of calculating the average future policy rate and hence a theoretical value for the 10-year yield, we make assumptions for when the Fed will begin to raise policy rates (liftoff), the pace at which rate hikes will occur, and how far they will go – the so-called “terminal rate.” Typically, the analysis assumes that once the Fed reaches the terminal rate, that the rate will be held constant there for the remainder of the period – in this case 10 years, depicted by Pattern A on Chart 1. However, I want to make a distinction between the “terminal rate” and the “neutral rate,” terms which are at times used interchangeably and probably should not be. If the Fed embarks upon a hiking cycle, and stops hiking at the terminal rate, that rate is by definition tight policy, and higher than the neutral rate if neutral is a reflection of a longer-run equilibrium policy rate which is neither stimulative nor anti-inflationary. I don’t think it’s realistic to assume that rates will be held at the terminal rate forever. I think Pattern B is incrementally closer to reality, with rates stabilizing around an equilibrium neutral rate that may be lower than the terminal rate.
Given the backdrop of global deflationary pressure which the US actively needs to avoid absorbing through a strong dollar, and credit destruction which is occurring with capital outflows from EM economies, the debate continues to rage on about whether the Fed can or should embark upon a tightening path. If the ability of the Fed to hike is still up for debate, while global growth is slowing and US growth is failing to impress (Q3 2015 is now tracking around 1%), ask yourself just how high the Fed will be able to hike rates on this policy cycle. Would you be all that surprised if the Fed was unable to push rates much above 1% before bona fide US recession risk became the dominant concern? I find that in my conversations with financial professionals immersed in the real economy, expectations are shifting lower, with 1 to 1.5% as a terminal rate, and that’s if the Fed even starts to hike.
If 1 to 1.5% is an expected terminal rate, and the neutral rate may be lower, 10-year yields implied by these policy pathways are low. The table below depicts implied 10-year yields at each combination of terminal and neutral rate, with a March 2016 lift-off date, 75bps per year pace, and zero term premium (“risk-neutral” calcs):
Now it gets trickier since we don’t live in a risk-neutral world. If the term premium were zero, given the choice between rolling Fed Funds overnight or buying the 10-year note at the ex ante expected average Fed Funds rate, risk averse investors would always choose to roll overnight. In theory, the term premium needs to be positive to entice the long-term investment. In practice, we believe term premium has been negative during periods of heavy demand for longer-term Treasurys, and in particular when Treasurys’ anti-correlation to risk assets is especially valuable for portfolio diversification. The NY Fed’s model-driven estimation of the 10-year term premium is shown in Chart 2, with those periods of negativity, but also a quite wide range of potential values. Because the term premium can vary so greatly, it’s not impossible for 10-year yields to go higher, even if expectations for the policy pathway remain anchored and frankly pessimistic. I would say though, that if you share the view that the Fed will stop hiking at 1 to 1.5%, then your implicit view of the term premium is pretty wide now – much wider than the NY Fed estimates– and should provide significant impetus to remain invested in Treasurys, even at these optically low yields.