This article provides an overview of the factors that determine “risk free” government bond yields. A combination of three factors influence bond yields:
- The current short-term interest rate
- The expected future interest rates
- And a term premium
For purposes of this article, sovereign credit risk is ignored, though it is certainly a factor in many countries that are not certain to at least pay their debts in nominal terms.
The short-term interest rate and its expected path in the near term are heavily controlled by central banks. However, interest rates and term premia over the long term are driven by factors outside policy determinations.
Monetary policy, which is comprised of three common tools:
- traditional interest rate adjustments
- quantitative easing (QE), and
- forward guidance…
…reacts to economic variables and influence economic conditions from shorter term maturities to yields of longer duration.
But theoretically and empirically, factors outside monetary policy influence yields significantly, especially for bonds of longer maturity.
10-year Treasuries move especially reliably with changes in growth and inflation expectations out over that timeframe. Bond yields follow declines in nominal growth expectations.
While central bank buying influences bond yields, economic fundamentals are in large part driving bond markets.
This has implications for bond traders and fixed income investors. It’s commonly believed that with bond yields so historically low that the risk in the bond market is highly asymmetric.
For example, if bond yields are 0-2 percent, they can’t get much lower than that because all the nominal return is gone and the real return is less than zero. (Real yields are nominal yield minus the inflation rate.)
So it almost seems pointless to put money in bonds when the fundamental purpose of trading and investing is to get back more than you put in.
But if fundamentals drive bond markets and yields simply reflect the economy outperforming or underperforming expectations, then there’s still more of a two-way market than the low yields might imply.
We’ll go through the various factors throughout the sections below.
Intro
The idea of this article is to provide a fairly thorough but non-technical outline of the factors that influence long-duration government bond yields.
To anyone who trades these markets, knowing how these asset moves is key to trading it well, whether that be bond futures (e.g., /ZB, /ZN), ETFs (e.g., TLT, IEF), or the underlying cash bonds.
You need to know how the levels of bond yields are determined and what drivers cause them to vary over time.
Throughout financial history, there’s never been an instance of long-term yields going to zero or below zero. It doesn’t make much economic sense to lend out money and get nothing or negative returns (before even accounting for inflation).
So there are lots of questions about the bond market as to why bond yields are so low and where they may go next. Accordingly, we need to identify and understand the key drivers of them.
A simple decomposition of bond yields
The yield on a long-term bond is a simple, mechanical function of the expected short-term interest rate over the duration of the bond plus a “term premium”.
A term premium is the compensation that investors require to hold a long-term bond versus a short-term bond. They’re wrapping their money up longer and long-maturity bonds are more volatile than shorter term bonds, so market participants expect to be compensated for this.
So therefore bond yields depend on i) the current short-term interest rate, ii) the expected series of future interest rates, and iii) term premia.
When these variables change (the inputs), the bond yield (the output) changes.
The current short-term interest rate is heavily controlled by the central bank. What goes into the decision of where to peg the short-term rate is largely determined by nominal growth rates (real growth plus inflation).
The “Taylor Rule” has been used to describe how central banks go about determining the short-term interest rate.
The main two inputs into the Taylor Rule relate to the unemployment rate and inflation.
Expected future short-term interest rates are influenced by monetary policy. Forward guidance is a tool that central banks use to communicate to the public over the expected future path of monetary policy (as largely dictated by short-term interest rates), which has an influence on longer-duration bond yields.
Long-term inflation expectations and the so-called natural rate of interest are closely related to the nominal growth expectations of the economy (often termed trend growth).
Term premia are a focus of a later part of this article.
Two factors are key determinants of term premia:
a) Levels of risk aversion – i.e., desire for a hedge against risk assets – and
b) changes in the level of inflation certainty
Term premia also can depend on the supply and demand of government debt by both domestic and foreign entities.
Government debt demand is led by numerous factors, such as the amount wanted by:
- central banks/reserve managers
- pension funds
- corporations
- charities
- endowments
- private money managers
- commercial banks for reasons related to balancing portfolios, having liquid reserves, regulatory reasons, and so on
Monetary policy has an influence on this term premia.
When central banks purchase bonds and other longer-duration assets such as mortgage-backed securities, corporate credit, and sometimes equities, they can cause term premia to fall. This, in turn, causes bond yields to fall.
This policy of buying longer-maturity assets starting in response to the 2008 financial crisis and the debt crisis that came out of it. This is commonly called quantitative easing, or QE.
The monetary policy toolkit of short-term interest rate adjustment, QE, and forward guidance are all similar.
They work to influence the economy by changing yields on longer-maturity bonds. This helps to bring down longer term interest rates throughout the economy. That, in turn, helps boost lending (credit is spending power). It also helps boost asset prices, so there’s a wealth effect through that channel.
Other monetary policy functions and operations such as average inflation targeting (AIT), operation twist, tapering, yield curve control (YCC), nominal GDP targeting (among others) can be considered a form of forward guidance and QE.
Monetary policy is a big determinant of bond yields. But it’s not the only source.
Macroeconomic forces, such as long-term inflation expectations impact bond yields. Domestic investors want to know what inflation is going to approximate so they can factor that into the yield.
If an investor wanted to simply at least preserve wealth – lose nothing in real terms over time – he or she would want the bond yield to be at least the rate of inflation.
For example, 10-year inflation expectations – called the “breakeven” inflation rate – are borne out of taking the difference between nominal bond yields and inflation-protected bonds of equal maturities in the countries where they exist. (In the US, inflation-protected bonds are called TIPS. The UK has gilt equivalents and other countries issue them as well.)
(Source: Federal Reserve Bank of St. Louis)
If inflation is three percent, for example, then a domestic investor will want the bond to yield at least 3 percent in nominal terms to preserve wealth.
(On the other hand, foreign investors will a different incentive. They will care more about the currency effects rather than what the inflation rate is in a different country.)
Factors that influence short-term interest rates and term premia – and therefore exert material influence on bond yields, especially at longer maturities – include:
- Trend growth
- Long-run inflation expectations
- Inflation uncertainty
- External demand influences
- Regulatory reasons
- Risk aversion
With where bond yields are currently at very low yields throughout the real world, it piques interest about whether “the rules have changed” or “the bond market is broken”.
But fundamental forces are still a key force even with heavy-handed central bank intervention.
Long-term inflation expectations still remain low despite extreme monetary policy measures since 2008 and 2020. Even if central bank influence was taken out of the equation, yields would likely be low due to low nominal economic growth.
Even if short-term interest rates remain anchored low, both monetary policy and non-monetary policy factors can drive long-term interest rates and term premia higher or lower.
QE and forward guidance can also drive yields, as can macroeconomic forces that central banks ultimately have little direct ability to control (e.g., what other countries do, exogenous shocks, fiscal policies, productivity changes).
If long-term inflation or growth expectations fall this could put further downward pressure on yields by causing future short-term rates to fall.
There are also upside risks. Higher than normal inflation could boost yields.
But the main takeaway is that bond risks are two-sided.
For the equities-heavy investor that’s looking for a natural risk off hedge, bonds are no longer as reliable as they were from the 1981-2020 period.
If there are higher long-term inflation expectations, that’s expected to hit both stocks and bonds because the higher rates reduce their present values through the discounting effect.
If long-term growth expectations decline, this will hurt stocks but help bonds, providing a natural hedge.
Bonds can still provide some diversification effect, but likely not as much when their yields are so low.
On that point, there’s also the element of the income generation capacity of bonds if their yields are little to nothing. That makes them less attractive from the standpoint, though for purposes of this article we’re focusing on what causes bond yields to change as opposed to simply their level.
Breaking down bond yields
A bond yield is the average expected interest rate over its life plus a term premium.
An interest rate is commonly taken as a short-term cash rate. The three-month yield is commonly used for this purpose. “Yield” is taken as the yield on a long-maturity bond (naturally, of a longer duration than the cash rate).
Bond yields are therefore dependent on the current interest rate, expected future interest rates, and the term premium.
Both the bond yield and interest rates are well-known factors. Moreover, expected future interest rates is also fairly straightforward to understand.
However, “term premium” may be a more nebulous concept.
Term premium involves everything impacting the yield that doesn’t involve the current interest rate and future interest rates.
It’s the difference between the yield on a bond and the market’s expectations of future interest rates.
Perhaps the most useful way to think of term premium is the extra compensation from holding a long-term bond to maturity as opposed to rolling a series of shorter ones.
For traders, the bond decomposition is important to help determine where markets could potentially go.
For instance, in today’s age of super-low rates there’s talk that bond markets are more or less capped out because central banks have set interest rates at their lower bound.
But when taking into account the path of future interest rates and especially term premia, we know that bond yields could fall further and provide additional upside for government bond prices or what can be earned when marked to market.
This could be due to the policy actions of central banks or from fundamental macroeconomic forces.
The yield curve
The yield curve provides cues on the expected future path of interest rates. The curve simply shows the current yields on bonds of different maturities.
Accordingly, it provides information on the current discounted path of forward interest rates and term premia.
Yield curves are normally upward sloping, which reflects the general idea that investors need to be compensated more for holding longer-duration bonds versus those of shorter maturity.
This can reflect either the expectation that interest rates will rise, positive term premia, or a combination of the two.
Term premia are the more dominant factor over time.
We know that over a long enough time horizon the expected changes in interest rates should be fairly close to zero.
If we were to look at the level of interest rates from 100 years ago, they’d likely be quite similar relative to what they were today.
We can even take that back hundreds or even thousands of years.
Clearly, expected interest rates aren’t going to increase by something like one percent per year. That would mean a delta of 100 percent in a century, which doesn’t make much sense. Generally, the variations are only a few percent, even over very long time horizons.
But we do know that the yield curve has been upward sloping most of the time. This means term premia are also generally positive and typically increase with duration/maturity.
In markets with steeper yield curves, excess returns can expect to be more positive due to the term premia effect.
The Expectations Hypothesis
Theory is used to help understand the policy and non-policy drivers of bond yields in more depth.
The expectations hypothesis (EH) is a basic theory of the yield curve. The EH asserts that term premia hold constant over time, but may vary based on maturity.
Based on the EH, the current interest rate and expected future interest rates were therefore the determinants of bond yields.
EH is a versatile explanation for bond yield dynamics. It works under a variety of yield curve shapes. When yield curves are upward sloping it works by asserting that term premia are positive and get larger with maturity.
The EH also explains, accurately, that yield curves tend to be steeper when short-term interest rates are lower and flatter when short-term interest rates are higher.
Moreover, it explains that the yield curve will be steeper when interest rates are expected to increase and flatter when rates are expected to go down.
This tends to be accurate because when the interest rate is low markets generally expect it to increase over time (going back to what we said earlier about interest rates not straying far from their long-run average over time). So, the yield curve will tend to be steeper than normal.
An inverted yield curve is often considered a recession forecast tool.
The EH provides a framework for explaining why – an inverted yield curve means investors anticipate interest rates falling in the future, which usually occurs during an economic contraction.
Yield curve dynamics can be explained through the EH. When central banks hike short-term interest rates, long-duration bond yields tend to increase as well, but generally by not as much. In other words, the curve tends to flatten.
Why is this? When interest rates rise, that means expected future interest rates out 1-5+ years in the future might go up as well.
But in terms of long-run expectations, interest rates tend to be fairly well anchored and don’t change by as much.
The yield of a bond is the average of the expected interest rates over its lifetime. So as interest rates on the front end rise, long-duration bond yields will tend to rise as well, but by not as much. This causes the yield curve to flatten.
So, interest rate changes have an impact on both the level of the curve and its slope, as the EH explains.
But it’s true that term premia change over time. The EH is a good simplified explanation but doesn’t account for this. All models tend to encompass the fact that short-term interest rates and expected interest rates are drivers of bond yields.
Now that we know this, let’s cover the determinants of the interest rate.
Interest rates
We know what variables influence bond yields, so we need to know what drives changes in them, starting with interest rates.
In developed markets, central banks largely control the short-term interest rate (either an “overnight” interest rate or a three-month rate). Adjustment of this rate is also usually the primary way that monetary policy is managed.
Accordingly, we need to know how central banks carry out their monetary policy operations.
Central banks have a statutory mandate of:
1) Having low and stable inflation (sometimes inflation alone is targeted)
2) Low employment
For example, the ECB targets just inflation. The Fed targets both and has an additional “shadow mandate” of financial stability, which can mean, e.g., preventing bubbles when necessary.
How do they target these to influence economic outcomes?
Every economy has a certain amount of demand and a certain amount of supply. And of different things.
Roughly speaking, there’s a real economy (supply and demand for goods and services). And a financial economy (supply and demand for money and credit).
The components are often termed aggregate demand (AD) and aggregate supply (AS).
The financial economy facilitates (and leads) the real economy.
Central banks have what’s called a reaction function, where economic and financial inputs lead to policy making decisions.
The reaction function is often modeled based on the “Taylor Rule” (TR) to show how central banks set the interest rate.
Then we have a framework like the expectations hypothesis (EH), which works to link bond yield to current and expected interest rates.
Models involving AD, AS, TR, and EH are the basics of what central banks and academic economists use to understand how economic events influence output and inflation and the prospective impact on monetary policy actions.
Aggregate Demand
In economic, central banking, and macro trading circles there’s often reference to what’s called an output gap.
This refers to the difference between:
a) the potential output of the economy if it’s at “full employment” versus
b) where it currently is at its present level of employment.
This is why many policymakers and traders pay close attention to the unemployment rate and jobs reports in an economy. It impacts policy decisions. They typically compare the current unemployment rate versus the “natural rate of unemployment,” or “NAIRU.”
The level of output, and the level of output that would be consistent with potential output, is linked with the real yield on long-maturity bonds.
When we talk about the real yield, we mean the nominal yield minus inflation expectations matched by maturity. (Inflation expectations vary over time horizon. They might be unusually high or low over short time horizons, but generally stay fairly well anchored over longer time horizons.)
Real yield = Nominal yield – inflation expectations
Central banks set the nominal yield. When we talk about something like “zero rates” we mean that the nominal interest rate is set around zero. The real yield can be something else.
But since central bankers are always going to target an inflation rate of at least zero to avoid deflation, if the nominal yield is zero, then the real yield is likely negative.
The real yield is the important part of the aggregate demand relationship and not nominal yield.
The output gap is the major determinant of what central banks are going to look at when determining policy. Policy can help influence aggregate demand over the short run.
For example, if the output gap is high (e.g., after 2008, after 2020), then policymakers are likely to be easier with monetary policy – i.e., lowering interest rates and doing what they can to get the economy going again.
If the output gap is low and inflation is starting to pick up and demand starts running into supply constraints then policymakers are likely to tighten policy by raising interest rates and/or tapering any supplementary assistance (e.g., bond buying).
Potential output, however, is not something that central bankers have much control over. It is determined by real factors like long-term productivity growth.
Overall, the aggregate demand equation shows a negative relationship between the output gap and the real yield in an economy. When the output gap is lower, the real yield is higher; when the output gap is higher, the real yield is lower.
The real yield on government bonds feeds through into a variety od borrowing and savings rates throughout the economy.
It also flows into other types of assets, including stocks and houses and private sector wealth. Therefore, it can also impact aggregate spending.
The way interest rates influence asset values is often termed the “wealth effect”.
When the real yield is low, borrowing is less expensive. Savings rates are also lower, so there is more incentive to take on more risk exposure. It helps stimulate demand for goods and services, which helps the real economy and labor market, and the output gap tightens.
When the real yield is high, borrowing is more expensive. Savings earns more in these circumstances. Demand for goods and services falls, aggregate spending drops, and the output gap widens when the real yield is too high for the level of economic activity.
Aggregate Supply
Aggregate supply helps tie the level of inflation with expected inflation and the output gap.
Inflation depends on expected inflation. There’s a common term called inflation psychology where if people expect prices to rise it tends to be self-fulfilling.
Consumers are more likely to spend on things that are likely to increase in price, which causes prices to rise. For instance, if there’s believed to be a gas shortage, then people are likely to fill up. It tends to occur when there’s a run on anything.
Moreover, firms often set their prices and wages based on expectations of what the future will be like. Many companies raise their employees’ wages by around 2-3 percent per year because that’s what they expect inflation – i.e., the cost of living impact on an employee – to be.
But, in general, workers will try to negotiate higher wages if they expect higher inflation.
Expected future inflation, therefore, influences inflation down the road.
Inflation also varies positively with the output gap. When the economy is doing well and the output gap is positive (i.e., current output is above potential output), firms will see pressure in their marginal costs and raise prices.
When the economy is contracting and the output gap is negative (i.e., current output is below potential output), marginal costs decline and firms may need to lower prices to stimulate demand.
Monetary policy
The AD-AS-TR-EH model incorporates both standard and non-traditional monetary policies. The adjustment of short-term interest rates is the most standard, so that’s what we’ll focus on.
Central banks have some control over short-run inflation and the output gap.
How do they do this? How does changing the short-term interest rate impact the macroeconomy?
The sequence of how this is done is called the monetary transmission mechanism.
Under the AD-AS-TR-EH framework, it goes as follows:
- The central bank sets the nominal short-term interest rate. In the AD-AS-TR-EH model, the short-term interest rate doesn’t impact macro variables directly. Instead, it impacts the nominal yield in long-term bonds.
- Long-run inflation expectations tend to be anchored, so changes in long-term bonds are pretty effective in translating to changes in real yields.
- Through aggregate demand (AD), changes in real yields impact the output gap. Though aggregate supply (AS), changes in the output gap impact inflation.
- Accordingly, by changing the nominal interest rate, central banks have influence over the variables they’re mandated to cover – inflation and output (i.e., the output gap).
How is the interest rate set?
Central banks take in data on inflation and output.
Central banks have an inflation target. When inflation is below the central bank’s target, they generally lower the interest rate and/or keep it low for an extended period.
When inflation is above the central bank’s target, they will generally raise the interest rate to help bring output and inflation down.
Some central banks also have a mandate to track output.
When the output gap is negative (current output is below potential), the central bank will want to lower the interest rate or keep it low to boost output.
When the output gap is positive (current output is above potential), monetary policymakers will want to raise the interest rate to avoid risking undesirably high inflation levels.
The Monetary Transmission Mechanism
The Taylor Rule
In a 1993 paper, economist John Taylor wrote up a basic framework for determining the appropriate interest rate for an economy.
Many economists, policymakers, and traders today use some version of the Taylor Rule to check where interest rates “should be” as it tends to fit well with economic data.
The main variable inputs include an inflation and output gap, with the interest rate being the output:
i = r* + π + bπ (π – π*) + bY (Y – Y*)
π – π* is called the “inflation gap”. This is the difference between the current inflation rate, π, and the central bank’s inflation target, π*.
bπ is a positive number, so the Taylor Rule does a good job of showing that the central bank should set a higher interest rate when inflation exceeds its target and a lower interest rate when inflation is below its target.
Y – Y* is the output gap. This is the difference between output Y, and the full employment level of output Y*.
bY is also greater than zero, so the Taylor Rule asserts that the interest rate should be higher when the output gap is positive (an expansion) and lower when the output gap is negative (a contraction).
What about r*+π, the first two terms in the Taylor Rule formula?
r* (pronounced “r star”) is the equilibrium real interest rate and π is standard economics notation for inflation. In other words, r-star plus inflation is the nominal interest rate.
To understand the Taylor Rule as a real interest rate you can subtract inflation from both sides of the equation:
Real Interest Rate = i – π = r* + bπ (π – π*) + bY(Y – Y*)
(To be more technically accurate, inflation expectations over X years matched to the maturity of the interest rate should be subtracted – e.g., 10-year inflation expectations subtracted from the 10-year bond yield – but realized inflation is a quality stand-in measure.)
The Taylor Rule would prescribe a real interest rate above r* when the inflation gap or output gap is positive (i.e., overheating economy), and a real interest rate below r* when the inflation gap or output gap is negative (i.e., sluggish economy).
When both are zero the Taylor Rule would prescribe a real interest rate equal to r*.
Accordingly, r* describes the “natural rate of interest” or the rate at which an economy would have neutral monetary policy when there’s no inflation or output gap.
In other words, the real interest rate is consistent with output equal to potential output (i.e., full employment) and stable inflation.
The real interest rate, in practice, depends on the output gap and inflation gap via the Taylor Rule.
Similarly, short-run expected future real interest rates depend on forecasts of the output gap and inflation gap.
What about long-run expectations of the real interest rate?
As time horizons lengthen, cyclical forces become less important.
As we’ve explained in other articles, productivity trends are the biggest determinant of economic growth and performance over the long-term. In the short-run, the credit cycles that monetary policy helps control matter more.
Cyclical forces matter less as time goes on and monetary policy is neutral on net.
Accordingly, the expected future real interest rate, eventually comes to r*.
So, long-term expectations of the real interest rate are anchored by r*. Likewise, long-term expectations of the nominal interest rate are anchored by r* + πLT, where πLT denotes long-term inflation expectations.
Because long-duration bond yields are heavily determined by expected future interest rates, their yields and valuations should be more sensitive to changes r* + πLT relative to shorter-term bonds.
In general, all long-duration assets (including long-duration bonds and stocks) and more sensitive to fluctuations in r* + πLT.
In a later section of this article, we show that long-maturity yields tend to move in lockstep with changes in the natural rate of interest and long-term inflation expectations.
Finding r-star after the 2008 financial crisis
Because of the 2008 financial crisis and the super-low interest rates that came out of the debt crisis, estimating r-star has been a very important variable. It’s critical for not only monetary policy/central banking purposes, but for understanding long-term bond yields.
With yields at low levels, traders need to consider whether owning bonds is still worth it. This impacts strategies such as those in balanced portfolios/risk parity, barbell approaches, the 60/40 portfolio, and others.
Yet r-star is not an observable variable. This is true even ex post. It can only be estimated.
Estimating r-star is not straightforward and simple, but it can be done in a reasonable way.
Economic theory relates r-star to trend growth in output. Output itself is closely linked to productivity and growth in the labor force.
The image below measures real GDP growth against real interest rates over the 30-year period 1990-2020. Over that time horizon, average real GDP growth should provide a pretty good estimate of trend growth and the average real interest rate should provide a quality estimate of r-star. In other words, the output gap and inflation gap should net out to about zero.
It is clear that countries that have experienced higher real growth have higher real interest rates and those with lower real growth have lower real interest rates.
Average Real Interest Rates vs. Average Real GDP Growth
Term premia
The term premium is the additional compensation that traders and bond investors require to hold to maturity a long-maturity bond versus rolling over short-maturity debt (e.g., three-month Treasury bills).
Term premia are positive over time and rise with maturity. Namely, investors will typically require extra yield to hold long-duration bonds over short-duration bonds.
Term premia at any given point cannot be found exactly. The yield curve finds average term premia at a point in time. Estimating exact term premia is done by some economists (e.g., Kim and Wright (2005), the ACM model kept by the New York Fed). But they have large standard errors.
The general idea here is to understand the drivers of term premia.
Term premia encompass all factors that influence a bond’s yield outside the currency interest rate and expectations of future interest rate.
The biggest drivers are:
- Changes in perceived risk
- Changes in supply and demand
Risk
Term premia will be higher when bonds are riskier and when market participants have a lower risk tolerance (are more risk-averse).
The biggest risk for fixed-rate government bonds (that are “risk-free”) is inflation. Bond yields take into account inflation expectations over the life of the bond.
But if inflation goes up unexpectedly, it can lower the real return of a fixed-rate bond.
The inflation risk is generally why term premia are positive on average. If there’s unexpected inflation, this makes long-term bonds riskier than simply rolling over short-term debt.
For example, during the US’s gold standard period of 1879-1970, yield curves were on average much flatter. A gold standard puts a limit on money creation, which limits inflation risk.
By extension, variations in inflation risk should be a driver of term premia variation over time.
When inflation uncertainty is high, market participants should expect greater compensation for holding long-duration bonds relative to shorter maturities. In the 1970s and early 1980s, inflation rates in the US were double digits and the volatility in inflation was high, causing longer-maturity yields to climb.
Moreover, there’s a countercyclicality to term premia in bond markets. Term premia will tend to be higher in recessions (when inflation volatility is much higher) than in expansions.
Supply and demand
Variations in net demand can arise from various factors, whether they be central banking actions (or policy actions more generally), macroeconomic, or geopolitical.
These factors can also be related to market participants’ tolerance for risk.
For instance, if governments need to issue a higher supply of bonds, investors, in aggregate, would need to take on more risk. If there isn’t enough demand for them, yields would rise. So, they would require higher expected excess returns to offset their lower risk tolerance.
Government bonds are appealing because of their liquidity, safety, ease of meeting regulatory capital and liability-hedging requirements. These benefits are highlighted during times when financial markets are stressed and there’s a “flight to safety / quality / liquidity”. Government bonds are seen as a safe haven, which can lower their term premia.
Changes in net demand can also affect term premia. When central banks hold a higher amount of bonds – either due to foreign demand or domestic policy monetary policy operations – this can influence term premia.
Quantitative easing and its offshoots (e.g., operation twist, yield curve control) involve the purchases of longer-maturity bonds (and other assets) by central banks. This reduces their net supply and is explicitly done to dampen risk premia.
In sum, term premia rises when inflation uncertainty and risk aversion increases and goes down when net demand increases.
The case of 2004-06
During the 2004-06 period, the Federal Reserve increased interest rates by 3.75 percent from 150bps to 525bps. Long bond yields held relatively steady. This was largely due to a fall in term premia.
The decline was due to various factors
- Low macroeconomic volatility
- Low risk aversion (ongoing economic expansion)
- Well-anchored inflation (low inflation uncertainty)
- More global savings (more demand for debt)
With global growth comes more savings. China was rising quickly in the world as were other countries, especially in emerging markets.
This led them to wanting to save in the dollar, the world’s top reserve currency. This led to more demand for US Treasuries by sovereign wealth funds and foreign central banks, helping demand for US government debt.
Monetary policy drivers of bond yields
Monetary policy is a key driver of bond yields. Various monetary policy tools are in actuality quite similar, as they react to what are essentially the same economic variables (near-term inflation and output) to influence the broader economy.
Central bankers use three main levers to influence monetary policy
- Short-term interest rate adjustments
- Forward guidance
- Quantitative easing (asset buying)
Interest rate level
Central banks influence long-duration nominal and real yields when they change interest rates. Real yields on long-maturity government bonds affect a variety of savings and borrowing rates and asset valuations across the economy – e.g., mortgage rates, corporate bond yields, stocks, houses, etc.
This, in turn, influences aggregate demand, which shows up in the output and inflation statistics. This is the classical monetary transmission mechanism as discussed earlier in the article.
The Taylor Rule helps provide a framework for how central banks set interest rates.
In simple terms, the central bank sets the interest rate low when inflation is less than their desired target and/or when output is less than the level expected at a period of “full employment”.
There’s also a lag inherent in monetary policy. Once interest rates are set, this doesn’t influence the real economy immediately. There’s a lag. (The financial economy – i.e., asset markets – will anticipate changes in advance and react instantly to “surprises”.)
The Taylor Rule also makes the assumption that variables like r-star and potential output are known, even though they can only be estimated.
It should also be noted that the Taylor Rule is not meant to convey an exact determination for the interest rate. It’s meant to be a qualitative framework for how interest rates are set.
The interest rate set by the central bank will deviate from the one prescribed by the Taylor Rule. But it describes the general reaction function of central bankers as they go about setting the appropriate policy rate.
The Taylor Rule informs us that changes in the near-term expectations for inflation and the output gap influence the interest rate. This, in turn, impacts long-maturity bond yields.
Moreover, if there’s an unexpected change to interest rates from the central bank that’s not explained by economic drivers (e.g., falling/rising unemployment, falling/rising inflation), that can also change bond yields.
When central banks have set their policy rates at, around, or even less than zero, they’re said to have approached a so-called lower bound in interest rates.
To get around that barrier to stimulate the economy they have other means. They can set interest rates negatively or more negatively, but a more reasonable choice is to engage in a different form of monetary policy.
Forward guidance
For a long-maturity bond, the present interest rate will likely hold true for only a small fraction of its life. That means the forward expectations for future rates are the more important factor.
Central banks realize that future interest rate influence bond yields, so this is where “forward guidance” comes into play.
Forward guidance entails communications to the public over the anticipated path of interest rates. It has been a popular policy tool since the beginning of the 21st century.
It is often stated that the two-year yield, which encompasses beliefs over where the interest rate will be over the next two years, provides the most accurate measure of the overall stance of monetary policy.
Forward guidance’s transmission mechanism is essentially the same as that for adjustments in interest rates.
When central banks look at the near-term outlook for inflation and the output gap, they provide guidance on future interest rates.
This influences the nominal yields of longer-maturity bonds, which impacts long-maturity real yields (since inflation expectations are fairly well-anchored). This affects the total amount of spending in an economy and therefore output and inflation.
Forward guidance has gone through many changes over the years.
Forward guidance based on outcome
Some forward guidance is structured based on outcome. Namely, interest rates will change as a result of economic conditions.
For example, forward guidance during Covid-19 was heavily outcome-based, saying they would support the economy until conditions materially improved.
Average inflation targeting (AIT) is another form of guidance based on outcome. This means that the Federal Reserve is willing to tolerate inflation above two percent for a period equal to the time it spends below two percent and will maintain an accommodative monetary policy until it gets there.
Targeting nominal GDP and the price level are similar variations of this policy.
Forward guidance based on time
Forward guidance that’s time-based expresses how interest rates are expected to change over the upcoming months, quarters, or years.
So, broadly speaking, changes in inflation and the output gap will influence communications to the public about the path of future interest rates, which will in turn influence long-duration bond yields. There’s also the chance of an unexpected change in the communicated path of future interest rates that can influence long-maturity yields.
This means a combination of these three forward guidance factors will lead to higher yields:
- Rising inflation
- Improving economic conditions
- Above-expectation (positive) communications for the forward path for interest rates
And vice versa for the combination that will lead to lower yields:
- Lower inflation
- Worsening economic conditions
- Below-expectation (negative) communications for the forward path for interest rates
Even if interest rates are at their lower bound, forward guidance can still help to ease policy if necessary as a way to lower longer-maturity yields in response to a lower outlook for inflation and/or output.
Sometimes this is referred to as jawboning. Instead of adjusting policy outright in a mechanical way, forward guidance is a way to talk (or jawbone) the markets in one way or another.
Quantitative easing or asset buying
Quantitative easing (QE) refers to central bank purchases of financial assets.
They mostly and typically buy their own government debt and other forms of government-based guarantees (e.g., mortgage-backed securities). But they can also venture into riskier forms of securities (e.g., corporate debt, ETFs, stocks) if necessary to inject more liquidity into the system and bring down spreads to help private sector credit creation.
QE is the latest adoption by central banks in response to the 2008 financial crisis and global recession that followed.
The BOJ’s intro of QE in 2001
It was originally introduced by the Bank of Japan in 2001 as a way to increase the supply of bank reserves to help target a higher growth rate for M0 (the monetary base). Increases in reserve should nonetheless have no influence in a low interest rate environment.
When the interest rate is zero or near zero and close to the rate paid on excess reserves, there is no opportunity cost associated with holding reserves and the demand for reserve becomes infinitely elastic.
So the additional expansion of the supply of reserve has no impact on interest rates or other monetary aggregates, spending, or prices. The BOJ accordingly abandoned this policy in 2006.
QE, as it’s been done since the financial crisis, has been different from Japan’s original version. QE now emphasizes the buying of longer-duration securities. This helps reduce bond yields and lending rates across the economy, which helps credit creation and asset valuations.
The Fed took a similar approach in July 1932 during the Great Depression by buying Treasury bonds to lower long-term yields. It was done at a much smaller scale even adjusting for inflation.
But it was sufficient enough to get enough liquidity into the economy to get the bottom in financial assets. US stocks declined 89 percent from October 1929 until then, though there were many bull markets (i.e., rallies of 20 percent or more) in between.
Even though the term QE is now common, “asset purchases”, “targeted asset purchases”, or even “debt monetization” (when it is such) would be more appropriate.
However, QE is a more neutral and less controversial term than either of those, so it has stuck.
QE basics
In the most basic form, QE involves the central bank purchasing long-maturity government bonds. It then holds them on its balance sheet.
These purchases lower term premia and reduce yields by reducing the net supply of longer-duration assets.
Lower nominal yields helps influence real yields. This impacts credit costs across the economy more broadly. In turn, it eventually impacts spending, output, and inflation.
Importantly, QE’s transmission mechanism is the same as through interest rate adjustments and forward guidance. Each policy works to influence the yields on long-maturity bonds.
Interest rate adjustments and forward guidance work through current and expected future interest rates, respectively. QE works by impacting term premia.
As mentioned, though QE traditionally involves the purchase of government bond securities, it may also involve securitized assets (e.g., MBS), corporate credit, and even equities.
These purchases may be financed either by new reserve creation or by selling shorter-maturity bonds.
They are different methods involving the same policy.
- When QE is financed via reserve creation, the central bank creates new bank reserves and purchase bonds in the open market from major financial institutions. Compared to QE financed by selling shorter-maturity debt, QE financed by reserve creation increases the amount of reserves held by banks.
- However, as mentioned in the aside about Japan’s original QE program, in a low interest rate environment the amount of reserves has little consequence for bank lending and, therefore, little impact on the broader economy.
So, by either:
a) purchasing government bonds and suppressing their term premia, or
b) by purchasing other long-duration assets directly…
…central banks work to reduce borrowing costs and increase asset valuations broadly. This works to achieve the goal of influencing output and inflation.
There are various cases of central banks working to influence long-maturity bond yields. For example:
- QE1
- QE2
- Operation Twist
- Yield curve control (YCC)
- Yield targeting
These all aim to impact long-maturity bond yields. It’s essentially no different from standard interest rate policy and forward guidance.
QE is often referred to as “money printing”. This is common among both the financial media and traders. But it’s important not to take it literally. QE is not about simply increasing the money supply, however it’s measured.
QE is simply about long-maturity bond yields and the valuations of longer duration assets. It is not a big shift from traditional interest rate policy and forward guidance. It’s just another tool that central bankers use to meet their mandate of stable inflation and full employment.
QE reduces term premia, but it also helps reinforce forward guidance. QE helps signal policymakers’ intentions to keep interest rates low for a longer period of time.
Generally, but not always, QE is taken away before interest rates are increased as QE is viewed as a form of secondary policy that’s used after rates are exhausted due to the lower bound.
The 2013 “taper tantrum” involved long-maturity bond yields going up and was a consequence of the Fed slowing its pace of asset purchases. It revised the forward expectations of interest rate increases, causing the re-rating of bond valuations.
Also, in this sense, forward guidance sits in the center of modern central banking and is effectively the primary monetary policy tool.
QE as a yield driver is about the size and overall mix of QE policies – i.e., how much and of what kinds of assets. These impact expected future interest rates and the term premia on long-maturity bonds.
Moreover, unexpected changes in QE policy – unanticipated changes in the amount, timing, or mix of QE policies – will influence yields.
Improving economic conditions, rising inflation, or an adverse change in QE’s forward guidance will lead to higher bond yields, and vice versa.
If central banks don’t view additional interest rate cuts as a viable way to stimulate the economy, QE (like forward guidance) is a popular tool to reduce the yields on longer-duration assets to stimulate the economy.
Summary
Central banks operate by:
- influencing the current interest rate
- the expected path of future interest rates, and
- term premia…
…through the tools of:
- traditional interest rate policy
- forward guidance, and
- QE…
…which all impact long-maturity bond yields.
Central banks have a mandate of maintaining low and stable inflation and (for many) full employment. So, monetary policymakers’ reaction function pertains to changes in the outlook for output and inflation.
Policymakers react to improving economic conditions and/or increasing inflation by tightening monetary policy, which typically results in higher yields.
Likewise, they react to declining economic conditions or falling inflation by adopting a more accommodative stance, which typically lowers yields.
Changes in the monetary policy stance of the central bank, which can include:
- an interest rate surprise
- A different take on the path for future interest rates, and/or
- unexpected changes to size and overall mix of the central bank’s balance sheet…
…will also affect longer-maturity yields.
The implications of this extend to all financial asset classes.
For example, the best environment for equities is not a big, booming economy but rather one that the central bank is trying to get restarted by lowering rates and providing ample liquidity toward.
When inflation is getting higher and the output gap is about closed (or growth is even above-trend), they want to start slowing things down, which will hit financial assets before it hits the real economy.
The image below takes a look at the drivers discussed in this section.
How Monetary Policy Drives Bond Yields
That short-term interest rates are close to zero (or lower) in most developed economies shouldn’t set “new rules for the game”.
Several central banks – most notably the European Central Bank and the Bank of Japan – have been willing to go to negative interest rates.
Others, such as the Federal Reserve and Bank of England, have not been willing to go to negative interest rates. They largely believe it’s less effective than alternative policies and can induce adverse consequences such as bank disintermediation (lack of profitability) and very low rates on savings.
Even if interest rates stay at zero or thereabouts, forward guidance and QE/asset buying both remain viable policy levers to further ease policy through a reduction in bond yields.
Bond yields will continue to react to economic news consistent with their monetary policy drivers.
The image below represents a time series plot of the percentage of variation in G6 bond yields (Australia, Canada, Germany, Japan, UK, US) as described by changes in near-term forecasts of inflation and unemployment.
Despite policy rates being at super-low levels, and despite this going hand-in-hand with bond market volatility, bond yields are still reacting to the major macroeconomic variables in a manner very similar to other periods. The cause-effect drivers of bond yields are not different.
Inflation and Employment (Output) Still Drive Changes in Bond Yields
(Source: Bloomberg, Consensus Economics. G6 markets sample. Data from January 1990 through December 2020. Chart displays R-squared statistics in 3-year rolling regressions of changes in 10-year bond yields on revisions to unemployment and inflation forecasts (left side), and average 10-year yields (right side).)
Non-monetary policy drivers of bond yields
Bond traders have spent lots of time since the 2008 financial crisis (and also after the 2020 Covid-19 crisis) pondering whether central banks will embrace negative interest rates or choose to expand their balance sheets through QE instead.
But while central banks are important and have the biggest lever over global liquidity, they are not omnipotent.
Monetary policy is not the only variable that influences long-maturity bond yields. In fact, sometimes the ability of central banks to exert control is overstated.
There are various factors outside policy consideration that affect the level and overall shape of the yield curve.
Other factors that exert meaningful influence over bond yields include:
- changes in trend growth and long-run inflation expectations
- variations in inflation volatility
- shorter-term changes in the business cycle, and
- changing demand for liquid, safe-haven assets
The current interest rate is under the control of the central bank. Their job is to approximate it as best as possible.
Accordingly, non-policy factors have a limited influence on short-run yields. But expectations of the future path of interest rates, especially at longer time horizons, and term premia are heavily driven by macroeconomic factors that stand outside the current monetary policy stance.
Central banks have a large impact on the expectations of future interest rates in the near term.
Nonetheless, long-maturity bond yields depend on expected interest rates well into the future.
Based on the AD-AS-TR-EH model, expected future interest rates should eventually come to the sum of r-star and long-term inflation expectations (r*+ πLT).
This means that r*+ πLT is essentially an anchor for long-maturity nominal bond yields. As a result, the long end of the yield curve should be quite sensitive to changes in the discounted expectations of r-star or long-run inflation expectations.
This is what’s seen in the data.
The image below shows the yields of 2-year, 5-year, and 10-year bond yields on long-run inflation expectations, long-run real growth expectations (a stand-in for r-star, which is not directly observable), and the current short-term interest rate.
Long-Maturity Yields Depend on Inflation and Growth Expectations
(Source: Consensus Economics, Bloomberg. Sample is G6 countries (Australia, Canada, Germany, Japan, UK, US), taken quarterly. Timeframe runs from December 1990 through December 2020. Long-run inflation and growth measurements are median forecasts of average inflation and real growth rates between 6-10 years ahead based on estimates from Consensus Economics. Yields and interest rate (T-bill) come from Bloomberg.)
10-year yields move more or less in lockstep with long-run inflation and growth expectations despite the actions of central banks.
At shorter bond maturities, the impact of long-run inflation and growth expectations declines. They nonetheless remain economically and statistically significant. But the influence of the short-term interest rate is stronger as maturity declines.
Long-maturity bond yields depend on r-star and long-run inflation expectations. And both of these variables will vary materially over time. R-star estimates have declined from the 1990s in the US and developed Europe because of slowing labor force growth and slowing productivity.
The expected average inflation rate can be inferred from the 10-year TIPS rate minus the 10-year nominal rate (or any maturity as long as they’re matched).
Over its history, it’s generally varied between 1.5 percent to just over 2.5 percent outside of recessions (when it declines).
US Inflation Rate Expectations
(Source: Federal Reserve Bank of St. Louis)
The Fed now has an average inflation targeting framework to allow inflation to run above its official 2 percent target to offset past shortfalls.
Because of an easier policy stance and more economic management falling on the shoulders of fiscal spending – which has a more direct feed-through into the prices of goods and services in the real economy – inflation in the US is likely to be structurally higher.
The Fed is also likely to tolerate it with its new policy framework and the inflationary pressures that come with easier policy.
Moreover, financial assets are more dependent on lower rates and there are lots of liabilities coming due in the US (related to pensions, healthcare, and other unfunded liabilities) that require a lot of money to fund them (which means easier policy for longer is likely.)
This may drive up longer-dated US bond yields as traders require higher compensation for inflation. There’s also the possibility the Fed could buy more of the bonds to keep yields repressed (in which case would weaken the dollar instead of bonds).
In 2000, r-star was estimated at around 3.4 percent. It is now estimated at around zero. Long-term inflation expectations have dipped from around 3.2 percent to around 2.3 percent. That’s a decline in r*+ πLT from 6.6 percent 2.3 percent.
Over the same period, ten-year US Treasury yields declined from 6.5 percent to 1.4 percent. The natural rate of interest and trend growth can explain around 85 percent of the decline in US Treasury bond yields over the past two-plus decades. Declining term premia explain the remaining portion.
R-star and Long-Run Inflation Expectations
(Source: New York Fed, Cleveland Fed. Sample runs from January 1982 through April 2020 for r-star and January 1982 through December 2020 for long-run inflation expectations. R-star estimates follow Holston, Laubach, and Williams (2017), which are maintained by the New York Fed (https://www.newyorkfed.org/ research/policy/rstar). Long-run inflation estimates follow Haubrich, Pennacchi, and Ritchken (2012), which are maintained by the Cleveland Fed (https://www.clevelandfed.org/our-research/indicators-and-data/inflation-expectations.aspx).)
QE impacts term premia on nominal-rate bonds, but there are also other factors that influence term premia that are not directly controlled by central banks.
As discussed in a previous section, these include:
- inflation uncertainty
- risk aversion, and
- changes in net demand for government bonds
The latter may be due to variations in the demand for safe, liquid government bonds. Other factors, such as foreign demand for safe-haven government debt from emerging markets, China and other rising Asian economies, and oil producers are all savers in foreign safe government bonds.
Each of these factors has been important in driving term premia throughout the decades. And each is likely to continue to have a vital role in the future.
Inflation volatility and uncertainty has decreased since the 1970s and early 1980s. But with massive monetary and fiscal support provided to developed economies, will central banks still be successful managing inflation and inflation expectations? (And without bringing down asset markets, if they do choose to kill inflation by reeling back policy support.)
Inflation uncertainty and risk aversion are generally higher in recessions than in economic expansions, but greater demand for safe-haven securities can suppress term permia in the short-run.
In summary, non-policy factors play an important role in driving bond yields.
While shorter-maturity instruments are more sensitive to policy factors, longer-maturity instruments react more to trend growth (through the natural rate of interest) and inflation expectations.
Growth and inflation anchor where future interest rates and likely to be and therefore the yields on long bond securities.
Inflation uncertainty and investor risk aversion drive changes in term premia. And fluctuations in the net demand for government bonds impact yields on government debt. This can be through secular changes in the demand for liquid, safe assets or from safe-haven flow.
Changes in any of these factors can drive bond yields higher or lower.
Bond yields are near all-time lows in many developed markets. And they remain very low relative to history.
Those involved in the bond market debate how low they can go. Minus-1 percent or thereabouts has been seen in some markets, such as Switzerland and Germany. In other words, the lender (i.e., the owner of the bond) pays the borrower, which seems counterintuitive.
Many ponder if they can go even lower, while others consider whether the combination of easy monetary policy and aggressive fiscal policy might cause low yields to reverse and a normalization over time.
Any tactical asset allocation decisions should be made carefully, whether that’s long or short bond positions (or something more complex) – relative to a desired strategic mix.
This article won’t cover predictions, but it will cover the relevant questions.
Why are yields so low?
Even if, hypothetically, Covid-19 was never an event, and unprecedented levels of monetary and fiscal policy support were never necessary, long-maturity bond yields would still be low relative to history.
Economic growth is a mechanical function of productivity growth and growth in the labor force.
Slowing inflation has to do with several factors, including:
- High debt relative to income (i.e., if debt has to be paid it diverts away from spending in the real economy)
- Aging demographics (not enough workers, producing increasing obligations relative to revenue)
- Offshoring production of various forms to more cost-efficient places, a drag on domestic worker salaries in countries where workers are more expensive
- Technology helps increase economy-wide pricing transparency and reduces reliance on expensive labor
- Over time, in the US, there’s been a lower role for unions and organized labor
This causes equilibrium interest rates, both real and nominal, to come down across developed markets.
Amid all this, there has generally been:
- Strong confidence in central bankers’ ability to control inflation
- Strong demand for government debt as a source of storing savings
- Low levels of risk aversion
With the actions of central banks mixed in:
- Low interest rates
- Targeted asset purchases (QE) to reduce the net supply of long-maturity debt and pushing down term premia
Low bond yields can stay low and should be low because of the combination of both non-policy and policy drivers.
What can make bond yields go lower?
We’ve established that monetary policy influences bond yields through the current short-term interest rate, expectations of future interest rates through forward guidance, and adjusting term premia through asset purchases/QE (and through reinforced QE).
Some central banks have gone into negative territory on their interest rates. Others, like the Fed and Bank of England, have chosen not to go this route.
Going negative could suppress yields because:
- The current interest rate would be lower, and
- If yields decline it should cause the markets to discount a lower forward path for interest rates
But the Fed and others don’t need negative interest rates so low as they have forward guidance and QE as tools. (Not to mention fiscal policy as an additional important easing channel.)
If economic or financial conditions deteriorated, the base case is for the Fed and broader government to add stimulus to the economy. If inflation stays within normal bounds and they don’t face that acute trade-off between adding support and excess price pressures, that will enable them to remain aggressive in their approach.
Outside of policy drivers, r-star (the natural rate of interest) and long-run inflation expectations are most important. Both of these variables have been in decline for decades, leading to lower bond yields, as they work to drive long-term levels of interest rates.
If the two variables are declining on net, long-maturity yields will decline in conjunction. If risk aversion and inflation uncertainty are low, then big changes in term premia are unlikely.
A pulling back of stimulus could cause upward pressure on yields. Central banks will do this when economic activity is strong and looks sustainably good.
At the same time, from a risk/reward perspective, the risks of pulling back too aggressively outweigh the good. A tapering of bond purchases will be done slightly to gauge market reactions and the impact on yields.
Outside of policy considerations, an increase in either r-star or long-run inflation expectations would put upward pressure on long-maturity bond yields irrespective of what monetary policymakers do.
An increase in inflation uncertainty and/or falling demand for safe-haven debt, among other factors, can cause term premia to run higher, leading to higher long-maturity bond yields.
How low can government bond yields go?
We know that the floor on government bond yields is not zero. Many countries have gone below zero on their yields, including the US on shorter-duration maturities to account for the small possibility that the Fed could decide to go into negative rate territory.
The logic behind a not-much-below-zero lower bound is rooted in theoretical alternatives.
At a point, a person could stack bank notes yielding zero and that would yield a better return than a financial security yielding below zero.
But there are also other factors at play.
a) Bonds are viewed as a low-risk store of wealth.
b) There is diversification potential in putting money in bonds as a risk-off hedge against stocks.
c) Bonds often serve as regulatory capital for certain financial institutions. So there are reasons why private sector entities might want to buy them despite their poor income generation potential.
d) Other financial asset returns are also low. When bond yields go lower, it tends to reduce the yields on other financial assets as well because investors buy them up as they look comparatively more attractive.
The diagram below illustrates:
Falling short-term interest rates reduce the yields of other asset classes
So, bond yields at 0-2 percent don’t look quite so unusual when the forward “yield” on stocks is 3-5 percent (the inverse of that is the P/E ratio, or 20x-33x).
Stocks are also perpetual duration instruments because their cash flows theoretically go on forever. This is different from bonds, which typically have a set maturity date.
That means the volatility of stocks is structurally higher than most forms of bonds, especially those considered safe, liquid, and of shorter to mid duration.
The typical risk premium between the US 10-year Treasury and stocks is 3-4 percent.
So, if stocks are yielding only 3-5 percent in forward returns – based on their earnings relative to their prices – then US 10-year yields of 0-2 percent don’t seem abnormal in relative terms.
Even if bond yields were to go negative, they could still be a source of positive portfolio returns and even real yield in a deflationary scenario.
Central banks will always target an inflation rate of at least zero since it’s difficult to get normal expansive economic activity in deflationary environments.
So, deflation is generally unlikely, though it can transpire for shorter periods and bonds could help offset losses in the equities portion of a portfolio during a risk-off shock.
The lower-bound on bond yields isn’t zero or minus-75bps (both have been surpassed in various countries). It is likely to be somewhat lower.
So even though it seems going below zero is nonsensical – i.e., lender paying the borrower, opposite of the typical arrangement – bond risk might not be as asymmetric as some market participants think.
Namely, while the “short bonds” argument seems to make more sense (and is more common) because rates can only go so low but interest rates can theoretically go up a lot, the bond market may be more of a two-way bet than some believe.
Is the game different?
The general thinking on the bond market since the 2008 financial crisis and reaction to it by central banks usually revolves around a couple main claims/themes:
a) QE/asset purchases represent a shift in the game. Central banks are now materially involved in the market and therefore,
b) the bond market is not as responsive to economic fundamentals as it used to be because it’s no longer a “free market”.
But QE is basically interest rate adjustment for long-maturity yields, so it’s not that fundamentally different from traditional policy aimed at the front end of the curve.
QE works through term premia whereas forward guidance works through the expected path of future interest rates.
Fundamentals are still at the heart of government bond markets. Yields are low with the help of central banks, but also make sense in the context of:
- Low long-run inflation expectations
- Slowing productivity growth
- Slower labor force growth
- Low levels of risk aversion
- Expectations of fairly well-anchored inflation
Central bank stimulus has certainly helped through interest rate adjustment, forward guidance, and QE – in addition to fiscal help – to support baseline levels of economic activity and to keep inflation at at least zero.
There is essentially nothing out of the ordinary going on in bond markets. Yields are reacting to fundamentals and central banks using the now-familiar policy toolkit in the ways we expect.
Reactions to good economic growth tend to lift bond yields and move real rates higher. Less favorable news tends to push them lower.
During the March-April 2020 period, when markets were sinking rapidly and extremely volatile due to Covid-19’s continuing impact on the world, safe bond yields declined.
This reflected a decline in term premia, in part because of economic fundamentals – lower discounted growth and inflation – and the impacts of central banks buying debt to support their economies.
There is no “mystery force” behind the bond markets. During these times it’s important to remember the cause-effect drivers of the market to help navigate the path forward, not toss it out because the game is supposedly different.
Conclusion
The fundamental drivers of bond yields can be understood by breaking them down into their component parts.
Long-maturity bond yields are a function of:
- The current interest rate
- Expected future interest rates, and
- Term premia
Monetary policy is influential and is focused on the near-term outlook for economic activity and inflation.
Bond yields can be influenced by:
- unanticipated monetary policy changes to the current interest rate
- the path of forward interest rates, and
- to changes in QE policy (size and mix of the central bank’s balance sheet)
Non-policy drivers are also important. These include:
- The natural rate of interest, commonly known as r-star, and
- Inflation (long-run expectations)
Bond yields in developed markets have fallen in recent decades due to contributions from non-policy factors.
R-star has declined due to falling expectations for trend growth (due to falling productivity growth and labor force growth).
Inflation expectations have also declined.
These factors especially matter for longer-maturity bond yields, which are largely anchored by these variables – as opposed to central bank actions for shorter tenors.
This provides the backdrop for where we are today in the main three reserve currency areas – the US, developed Europe, and Japan.
Looking forward, bond market traders and investors should still view the government bond market as two-sided despite the super-low yields and ostensible limited upside in bond prices (i.e., further decline in yields).
Even with low yields and ample central bank intervention in the market, fundamentals still drive markets and a fundamental approach is still applicable when trading the bond markets.