The failure of long term yield to break their almost secular downward trend raises, once again, the issue of their drivers. The traditional decomposition of long term yields offers some clues: expected inflation, expected short term real yields and the term premium (see below). This highlights the drivers of long term interest yields but decomposition is not explanation.
It may yet highlight the topics around which the debate is centered: should we care about the low level of expected real short term interest rates, a sign that future growth prospects are very dim ? Should we, on the contrary be wary of the disappearance of the term premium and hence the factors that drove it into negative territory (frictions, regulation, central banks purchases)?
Econ 101 defines the interest rate of the economy as the marginal productivity of capital. More precisely, the real interest rates r (to which we should add the depreciation/amortization of capital) is equal to the marginal productivity of capital. In that sense, low real yields would suggest that the accumulation of capital has been excessive over the previous decades. The capital overhang and the associated low marginal productivity could explain the stubbornly low level of yields.
Yet, the marginal productivity of capital should be seen as the return of capital. The real interest rate is, on the contrary, the difference between the nominal rate i set by the central bank and expected inflation r =i-E(infl). The return of capital is hard to assess in real time but it is reasonable to assume that it is stable in the medium run. There is thus a non-observable "natural" real rate r* that has to be assessed by the central bank (through the trend in productivity gains and population for instance) in order to conduct monetary policy. The challenge for the Fed is to match an unobservable "natural" real interest rate (r*) to tame economic fluctuation, in order to avoid booms (r<r*) and busts (r>r*).
This framework is yet very useful to explain the run up to the Great Recession and the subsequent decline in yields. If monetary policy remains too lax for too long, that is r<r* (assuming there is no threat of inflation and big difference in observed and expected inflation), excess capital accumulation will be fueled by excess indebtedness. This will lead to :
1. Excess capital accumulation, which means, afterwards, a long lasting process of liquidation and low return of capital for long. It can hinder investment even if interest rates are lowered (if r* is very low and r stays positive because of the 0% bound on the central bank interest rates while inflation stays low).
2. Excess indebtedness that has to be dealt with an increase in the saving rate. This leads us to the most used definition of the secular stagnation: a situation in which the real yield cannot fall enough for saving and investment to equate. Not only real yields cannot fall when the central bank sets interest rates at 0% and inflation remains muted. But when caught between poor collateral and huge debt, economic agents might be less sensitive to the interest rate in their saving behavior (forced saving).
3. This situation can last for a long time, as is often put forward by the BIS: "the pace of deleveraging remains modest or non-existent in many sectors around the global economy, implying that the debt overhang may still be a significant impediment, even if debt crisis risks have receded for the moment." The chart below shows that low rates have spurred an increase in private leverage to a point where significant increases in interest rate (monetary policy normalization) are precluded by the risk that they have on growth.
This framework suggests thus that yields are low because there is too much capital (low return and weak investment), too much forced saving (deleveraging) and too much debt that preclude any normalization of monetary policy. This situation might last…
This explanation of the current level of interest rates is static, that is based on the current situation, the inheritance of past excesses. Several studies try to explain the low level of long term yields by the dearth of positive growth factors in the future. The economist Robert Gordon is the most famous of them with his list of structural headwinds: demographics, the illusion of the new wave of technical progress, structural public deficit… I don't dwell too much on them but they could explain why the potential for expected real short rates to rise in the future is limited (and why a 3.75% average "dot" for long run / equilibrium Fed Funds rates might seem too high!).
If capital accumulation gives an economic rationale for long term yields, asset pricing and, more generally, inter-temporal choices, are also an important framework to understand the drivers of long term yields. They mostly depend on time and uncertainty.
Within this backdrop, the real risk-free rate is:
1. Positively dependent on time preference ρ : higher when people are impatient;
2. Positively dependent on the dynamism of future/expected consumption E(∆C) ;
Time preference (ρ) is generally stable in the medium run and also depends on demographics (higher when ageing). For instance, in a recent NBER, Galor and Ozak show that "descendants of individuals who resided in geographical regions in which crop yield was historically higher are characterized by higher long term orientations", proof that time preference has a huge ancestral component. If we want to take into account the ageing of advanced economies, it is worth noting that it could have led to a rise in time preference and thus a rise in yields. I thus leave this factor aside.
The dynamics of consumption E(∆C), that is the change in consumption from now to the next period, rests on many factors but the question here is more about the causality. Does the expected future growth adjust to the low level of interest rates, or does the causality stand the other way? We just know that interest rates are low when the expected change in consumption growth is low.
I leave aside the factor γ of the equation above, which is a parameter that characterizes the willingness of households to accept (or not) jumps in their consumption profile ("intertemporal elasticity of substitution"). It matters but its value has proven impossible to measure.
We can thus link the dynamics of consumption to that of real interest rates. Households discount future incomes using their own discount factor or time preference: when impatience is high, their discount rate is high and conversely. In addition, they base their consumption decision on their wealth, that is their current financial wealth and their expected incomes, which are themselves dependent on education (human capital) and future taxation.
Households will thus have a path of expected consumption that depends on the difference between the real interest rate and their time preference. It is called the Euler / Keynes-Ramsey equation in economics: future consumption relative to today's consumption will decline if the difference between the real interest rate and the time preference dwindles. Said differently, the fall in the real yield suggests that households have brought forward consumption at the expense of future consumption: E(∆C)= (1/γ)*(r-ρ). The chart below shows that current consumption has remained significantly robust over the 2011/2014 period and confirm this view.
The advantage of this framework is that it can explain several features of the post 2008 recovery:
1. The stability of consumption among wage stagnation can be, beyond factors such as the rise in public transfers, lower interest payments, cheaper oil and natural gas, attributable to the building up of asset prices bubbles. The fall in the labor share seen below is consistent with an artificial increase in capital return in a world of excess capital. This is a very Marxist view and it can be refuted by the fact that a large part of the fall in the wage share of GDP is linked to the huge fall in capital good prices. Yet, it clearly helped maintaining a high level of profitability to businesses in spite of a dearth of investment opportunities.
2. The lower expectation of future income is, here again, the result of the long lasting weakness of the job market, but it can also be seen as a sign that households don't expect consumer spending to accelerate in the future. As can be seen, there is a link between income expectations of household and the labor share of GDP, whose fall has been compelling over the last decade.
If this framework is right, the Fed may face some dangers ahead:
1. A rise in long term interest rates would increase future consumption relative to present consumption (read: today's consumption would be adjusted on the downside)
2. It would also impact asset prices ("bubbles" for some) and seriously impact the growth rate of the economy.
This Euler-Based approach does not really provide the fundamental cause for the low level of real yields but has the advantage of providing a consistent framework. It also shows to what extent the future path of consumption might be dependent on long term yields and thus the dangers of a sharp rise in long term yields for today's consumption growth.
As I have mentioned above, the reference interest rate in the asset pricing approach is the risk free rate. The economic approach would rather focus on the net return of capital (marginal productivity of capital minus depreciation), even if it is artificially and unsustainability inflated (skewed distribution of income). The current level of risk premia (equity or credit, even in housing for some local markets), suggests that expected returns have diverged from the risk free rate in a manner that has to be explained. One major explanation is liquidity.
There is a growing scarcity of safe assets for reasons that are not so much linked to the reduction of public deficits but to the surge of capital/regulation-based demand. A huge part of the fall in the "term premium" should be attributed to the combination of liquidity created (central banks' balance sheets expansion) and the nature of investors involved (home bias, financial repression, liquidity and capital ratio, Basel, Solvency…).
Bottom Line: there is no consensus on the reasons behind the stubbornly low level of long term yields. They reflect a combination of monetary policy actions, new regulations, investors' behavior, past excesses, saving behavior, doubts on future growth and so on. In this post, I show yet, that there exist several frameworks that enable us to understand the logic behind and the "soft spots", that is where explanations might seem the most vulnerable to a change in the Fed policy.
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.