Government bonds are a great deal in global financial markets, representing approximately 30% of total market capitalization (Doeswijk and Swinkels 2024). Somehow it echoes the mantra of a seasoned mentor I once had, “Bonds! Bonds, anyone must have…”
To crack the code of government bond returns, our framework employs a simple, but broad-based analytical toolkit. Breaking away from the short-term focus that dominates trading circles, we take a deliberate long-term stance on government bond returns. Let’s for a moment zoom out, and cut through the noise to identify those more enduring trends shaping the long-term prospects of 10-year U.S. Treasury!
Let’s briefly outline our approach.
Our analysis begins by examining the historical monthly returns of 10-Year U.S. Treasuries, identifying trends, patterns, and statistical measures like skewness and kurtosis. We then incorporate exogenous variables to explore how U.S. government bonds interact with broader market conditions. Our goal is to uncover hidden relationships and potential spillover effects using a structured, systematic macroeconomic approach.
Let’s take a closer look at a set of 11 predictor variables: macroeconomic, financial, and commodity-based indicators. The selection of these variables is guided by both empirical evidence and economic intuition. The current outcome is based on monthly time-series data updated up until the end of September.
What’s the output now of our InflectionPoint Model?
Based on the overall output of our Inflection Point Model, we believe that the market environment is currently constructive for 10-Year U.S. Treasuries. We are therefore inclined to have a strong long bias in our trading strategy.
A Short Insider’s Guide. How do we use this long-term model?
The model identifies market regimes that are more conducive to strong, risk-adjusted returns for U.S. Treasuries. Rather than generating short signals, the model creates a clear distinction between high-conviction and low-conviction regimes for 10Y Treasuries.
This approach is a valuable input for managing multi-asset portfolios and informing trading strategies, allowing for a flexible stance based on market conditions. We combine systematic guidance with tactical considerations, encompassing macroeconomic factors and short-term indicators to achieve a comprehensive analysis.
Let’s a take a closer look at the various components of our model.
1) 10Y U.S. Rates Time Series
This channel uses three elements the underlying market trend and two realized higher moments from the distribution of 10Y Treasury returns based on the last 3 years of monthly data. Previous research, for example, Pedersen (2012) showed evidence of past bond returns predicting future returns (i.e., momentum) in the developed markets. It’s a rather well-known concept, but we believe momentum and persistence of returns should not be missing. Market momentum is complemented by realized higher moments. Several recent studies have analyzed the forecasting power of higher moments, i.e. skewness and kurtosis, for stocks and corporate bonds mainly from the U.S. markets (Dittmar, 2002, Xing et al., 2010, Amaya et al., 2015, Bai et al., 2016). Our model identifies the most favourable market regime for bonds when skewness is negative and kurtosis is low.
What about the current regime?
While recent October price action may suggest a potential market shift in bond momentum, overall this subcomponent of our model focused on higher moments continues to indicate a favorable market regime for 10-Year Treasuries.
Outcome = Bullish Bonds!
2) U.S. Rates Curve Channel
Another building block of our model looks at how 3-Month Treasury Bill rates and the short-end yield curve slope are moving. Past studies by Dyl and Joehnk (1981), Fama (1984), Campbell and Shiller (1991), Ilmanen (1995, 1997), Yamada (1999), Ilmanen and Sayood (2002), and Duyvesteyn and Martens (2014) have shown that a steeper curve usually means higher future bond returns.
While many studies have focused on longer-term yield spreads, we’re more interested in the short end of the curve. We think trends in 3-Month Treasury Bill Yield are also very important because they tell us what the Central Bank is doing with monetary policy.
We my seem naïve, but don’t fight the FED!
The term spread in our model is less constructive for the bond market regime, leading to a neutral outlook for bonds.
Outcome = Neutral Bonds!
3) The Equity Channel
To the best of our knowledge, Ilmanen (1995) was among the first to propose using past equity returns to predict international bond returns. He found that negative past equity returns often lead to positive future bond returns. We incorporate this factor into our model, along with two other elements that we believe offer valuable insights: the Robert Shiller CAPE Excess Yield and the correlation between bonds and equities.
Our equity channel is built on three key factors:
CAPE Excess Yield: We take a long position in bonds when the CAPE Excess Yield falls below 3%, indicating an overvalued equity market.
Stock-Equity Correlation: A positive 3-Year correlation between stocks and equities further supports a long position in bonds, as it often signals higher expected returns.
Equity Market Momentum: Similar to Ilmanen’s findings, we become more bullish on bonds when equity markets exhibit a downward 12-month momentum.
10y yields and equities: 3y rolling correlation
The 3-year rolling correlation has shifted into positive territory. When government bonds can reduce overall portfolio risk, it’s logical to expect lower or even negative returns on bonds. The economic intuition is that investors are willing to pay a premium for insurance against equity market downturns. In other words, the bond risk premium (brp) becomes smaller during periods of negative stock-bond correlation.
Conversely, during periods of positive stock-bond correlation, like the current environment, investors may demand higher risk premia to own bonds. This is because bonds have lost their traditional diversification and hedging properties. The higher risk premium associated with bonds during such periods can be a leading indicator of higher returns in the future.
CAPE Excess Yield
The highly compressed CAPE Excess Yield suggests that equity valuations are currently stretched. Within our framework, this metric serves as a strong leading indicator of potential outperformance for bonds.
Despite the strong equity momentum, the overall score of the equity channel remains robust. The positive cross-asset correlation and elevated equity valuations contribute to a favourable bond outlook.
Outcome = Bullish Bonds!
4) The Commodity Channel
The Commodity and Gold trends are defined as the past 12-month Commodity/Gold momentum. Ilmanen and Sayood (2002) already used the trend in commodity prices to predict German bond returns. The economic intuition is that falling commodity prices signal disinflationary pressures, which should boost bond returns both contemporaneously and in the near future, given the observed under-reaction effects in many asset markets.
Our analysis has also confirmed the added value of incorporating commodity and gold trends into the bond return prediction model.
While the commodity trend is more mixed overall, the strong bull trend in gold is bearish for bonds. However the broader commodity complex is currently trading sideways, it remains in a downtrend.
Outcome = Neutral Bonds!
5) The Macro Channel
Ludvigson and Ng (2009) find that “real” and “inflation” factors have important forecasting power for future excess returns on U.S. government bonds, above and beyond the predictive power contained in forward rates and yield spreads. In our model, the macro channel is focused on “real” factors. For simplicity, we use the OECD US Leading indicator as the commodity channel already captures the inflationary pressure for us in real-time.
Our model takes a long position in bonds whenever the Composite Leading Indicator (CLI) Normalized for the United States falls below 101. A weaker economic outlook is generally indicative of stronger bond returns in the future.
Outcome = Bullish Bonds!
Conclusion
In our previous post, on the 28th of September “Decoding the Fed’s Playbook: Simple Monetary Policy Rules”, we advocated for a tactical short position in rates:
“Market pricing seems to overstate rate cuts, our analysis suggests a slightly less steep trajectory. The current market pricing might offer opportunities to marginally reduce exposure to overly aggressive rate cut expectations, potentially targeting one or two fewer cuts than implied by the market. The key rationale for this view is relatively more sticky inflation hovering around these levels or just slightly above. The base case, however, barely offers an attractive trading opportunity.”
Given the recent market sell-off, which has pushed 10-year rates up by nearly 40 basis points in October,
we believe the recent sharp correction in 10Y rates driven by accelerating economic momentum, slightly higher inflation print than consensus and the pricing of a Red Sweep in the next US presidential election on Nov 5th is overdone and represents a good set-up and entry point to re-establish a long position in rates based on the fact that our long-term bond model remains structurally constructive on the asset class!
Happy trading and see you next week!
InflectionPoint
Disclaimer
All views expressed on this site are my own and do not represent the opinions of any entity with which I have been, am now, or will be affiliated. I assume no responsibility for any errors or omissions in the content of this site and there is no guarantee for completeness or accuracy. The content is food for thought and it is not meant to be a solicitation to trade or invest. Readers should perform their investment analysis and research and/or seek the advice of a licensed professional with direct knowledge of the reader’s specific risk profile characteristics.
Leave a Reply