Treasury Yield Curve Research Articles

Neural Networks, the Treasury Yield Curve, and Recession Forecasting

The authors use neural networks to examine the power of Treasury term spreads and other macro-financial variables to forecast US recessions and compare them with probit regression. They propose a novel three-step econometric method for cross-validating and conducting statistical inference on machine learning classifiers and explaining forecasts. They find that probit regression does not underperform a neural network classifier in the present application, which stands in contrast to a growing body of literature demonstrating that machine learning methods outperform alternative classification algorithms. That said, neural network classifiers do identify important features of the joint distribution of recession over term spreads and other macro-financial variables that probit regression cannot. The authors discuss some possible reasons for their results and use their procedure to study US recessions over the post-Volcker period, analyzing feature importance across business cycles. <b>TOPICS:</b>Fixed income and structured finance, big data/machine learning, financial crises and financial market history <b>Key Findings</b> ▪ It is difficult to make the case that neural network classifiers greatly outperform traditional econometric methods such as probit regression when forecasting US recessions if performance is measured on the basis of forecast accuracy alone. ▪ That said, neural network classifiers identify important features of the joint distribution of recession over term spreads and other macro-financial time series that probit regression and other traditional methods cannot. ▪ The authors propose a three-step econometric process for conducting statistical inference on machine learning classifiers, with the goal of better explaining recession forecasts and enabling model outputs to be linked back to the instruments of monetary policy.

Price Discovery in China’s Corporate and Treasury Yield Curves

As financial development progresses, the maturity structure of bond yields plays a rising role not only in the financial system but also as a key transmission channel of monetary policy. China is likely to be no exception. However, specific characteristics in China’s bond markets raise two major questions. First do China’s Treasury bonds offer the benchmark term structure of yields, or is this role fulfilled by the young but fast expanding corporate bond market? In other words, where does price discovery take place in the Chinese bond market in terms of the different components of the yield curve? We identify both dynamic and long-run relationships between each of the level, slope and curvature factors of the Treasury and corporate bond markets yield curve in China. We aim at determining which market plays a leading role in the discovery of each factor of the yield curve. We obtain three main results. First, we document for the first time the presence of a long-run relationship between the corporate and Treasury bond markets in China both for the level and the slope of their yield curve. Second, such a long-run relationship appears to be stable between the slopes over the full sample 2006-2017, but shows a break for the level factor in 2012. Third, the source market for price discovery varies with the parameters of the yield curve. While the corporate bond market is the source of price discovery for the level factor, this function is fulfilled by the government bond market for the slope parameter. The finding that the Treasury bond market is not fully dominant in level bond-pricing may not come as a surprise. Although China’s corporate bond market has developed rapidly in the past fifteen years, there were few default cases during that period. It is believed investors treat the default risk of corporate bonds as similar to that of Treasury bonds, and benefit from the high corporate spread. Our results for the slope parameter imply that market-oriented reform has progressed enough for the Treasury bond market to already provide a benchmark slope for the yield curve of corporate bonds. When the reform progresses further, we would expect corporate bonds to be priced according to their risk profile which should make the Treasury market lead in price discovery also for the level of the yield curve.

Macro-Factors of Credit Spreads in Corporate Bonds

This study emphasizes on analysis of the macro factors of the credit spreads (CS) of the fixed-rate corporate bonds of non-financial enterprises.Firstly, industry growth (IG) has a significant negative impact on CSs. IG reflects the prosperity of the real economy. The increase of investment demand promotes the rise of bond prices and narrows the CSs; Secondly, consumer price index (CPI) has a significant positive impact on CS. On the one hand, the increase of inflation will reduce bond investment demand, on the other hand, it will increase investors' expectations of future tightening monetary policy. Uncertainty increases while default risk rises and CS widen. Thirdly, spreads of treasuries have a significant negative impact on CS. When the slope of the treasury yield curve increases and the expectation of interest rate increases, macro-economy is in an upward stage, and CS narrow. Fourthly, the analysis of VAR, variance decomposition and impulse response function show that the impact of CPI, IG and interest rate difference of national debt on spreads has time lag. While money supply affects CS by affecting CPI. This study discusses the CS which is very important for the asset allocations of investors. The investors and market participants should consider the macro-factors while invest in corporate bonds.

Forecasting U.S. Yield Curve Using the Dynamic Nelson–Siegel Model with Random Level Shift Parameters

In this paper, we develop a new model based on the classical dynamic Nelson-Siegel model by introducing random level shift (RLS) parameters. The built-in RLS can capture cyclical fluctuations in interest rates and structural breaks induced by technological progress, financial crisis, major monetary policy interventions, etc. In addition, the model can be used to forecast future structural breaks. We apply the model to fit and forecast daily U.S. Treasury yield curves and the model outperforms other widely used models. The empirical results show that the model not only has a better in-sample fit with residuals exhibiting less persistence but also has superior out-of-sample performance. Moreover, the model performs very well especially for short-term and long-term bonds, and the performance improves as the forecasting horizon increases.

Echo over the great wall: Spillover effects of QE announcements on Chinese yield curve

This paper examines the spillover effects of announcements on quantitative easing (QE) conducted in developed economies on the Chinese Treasury yield curve. In spite of China’s firewall of cross-boarder capital control, we find that the US QE announcements move the Chinese yield curve immediately and significantly, through the channels of signaling and portfolio balancing. The results are robust across a variety of event analysis methods using reduced-form or structural models. Using the heteroskedasticity assumption for shock identification in an arbitrage-free term structure model, it was identified that the impact of US QE on Chinese yields was comparable, in size and significance, to China’s own monetary policy shocks.

Liquidity risk and corporate bond yield spread: Evidence from China

AbstractThis paper investigates the contribution of liquidity risk to Chinese corporate bond spreads. We calculate corporate bond spreads based on the full treasury yield curve and establish a set of liquidity measures of the Chinese corporate bonds. Our empirical study shows that liquidity premium accounts for a relatively smaller portion of corporate bond spread in China, although the market liquidity is low and corporate bond issuers are strictly pre‐screened. These findings are interesting, as the developed markets have better liquidity and less pre‐issuance restriction, and liquidity premium still explains a relatively larger portion of corporate bond spread. Besides, we also explore the determinants of Chinese corporate bond liquidity and default premiums.

Are Shadow Rate Models of the Treasury Yield Curve Structurally Stable?

We examine the structural stability of Gaussian shadow rate term structure models of Treasury yields over a period that includes the time during which the U.S. policy rate was at its effective lower bound. After a conceptual discussion of several potential sources of a structural break in the context of the shadow rate model, we document various pieces of evidence for structural instability based on predictive tests and Lagrange multiplier tests, as well as with separate estimations of the pre-ELB and post-ELB subsamples. In order to overcome the difficulties associated with the latent-factor nature of the model in testing for a structural break, we focus on objects that can be given intuitive interpretation, such as principal components, or that are constructed to be invariant to factor rotations.

Machine Learning, the Treasury Yield Curve and Recession Forecasting

We use machine learning methods to examine the power of Treasury term spreads and other financial market and macroeconomic variables to forecast US recessions, vis-à-vis probit regression. In particular we propose a novel strategy for conducting cross-validation on classifiers trained with macro/financial panel data of low frequency and compare the results to those obtained from standard k-folds cross-validation. Consistent with the existing literature we find that, in the time series setting, forecast accuracy estimates derived from k-folds are biased optimistically, and cross-validation strategies which eliminate data &amp;amp;quot;peeking&amp;amp;quot; produce lower, and perhaps more realistic, estimates of forecast accuracy. More strikingly, we also document rank reversal of probit, Random Forest, XGBoost, LightGBM, neural network and support-vector machine classifier forecast performance over the two cross-validation methodologies. That is, while a k-folds cross-validation indicates tha t the forecast accuracy of tree methods dominates that of neural networks, which in turn dominates that of probit regression, the more conservative cross-validation strategy we propose indicates the exact opposite, and that probit regression should be preferred over machine learning methods, at least in the context of the present problem. This latter result stands in contrast to a growing body of literature demonstrating that machine learning methods outperform many alternative classification algorithms and we discuss some possible reasons for our result. We also discuss techniques for conducting statistical inference on machine learning classifiers using Cochrane&amp;amp;#x27;s Q and McNemar&amp;amp;#x27;s tests; and use the SHapley Additive exPlanations (SHAP) framework to decompose US recession forecasts and analyze feature importance across business cycles.

Forecasting the Term Structure of Interest Rates with Dynamic Constrained Smoothing B-Splines

The Nelson–Siegel framework published by Diebold and Li created an important benchmark and originated several works in the literature of forecasting the term structure of interest rates. However, these frameworks were built on the top of a parametric curve model that may lead to poor fitting for sensible term structure shapes affecting forecast results. We propose DCOBS with no-arbitrage restrictions, a dynamic constrained smoothing B-splines yield curve model. Even though DCOBS may provide more volatile forward curves than parametric models, they are still more accurate than those from Nelson–Siegel frameworks. DCOBS has been evaluated for ten years of US Daily Treasury Yield Curve Rates, and it is consistent with stylized facts of yield curves. DCOBS has great predictability power, especially in short and middle-term forecast, and has shown greater stability and lower root mean square errors than an Arbitrage-Free Nelson–Siegel model.

Recession Prediction Markets and Macroeconomic Risk in Asset Prices

This paper analyzes recession prediction markets from Intrade and PredictIt where individuals bet on the binary outcome of a recession occurring by a certain date. Using a time series of such historical recession prediction market data to measure macroeconomic risk in a variety of asset classes, we find a 1% increase in ex-ante recession prediction market probability is associated with a -0.29% decline in U.S. equity markets and a -0.77 basis point decrease in the 10-Year U.S. Treasury yield. Recession prediction markets also explain a significant amount of the variation in credit spreads, the U.S. Treasury yield curve, and U.S. dollar foreign exchange rates but cannot explain the value and momentum anomalies in equity markets as well as the U.S. treasury term premium.

Systematic Financial Intermediation and Business Cycles

Theory suggests that financial intermediation (FI) spurs economic growth by reducing investment frictions. Prior literature focuses on bank lending as a driver of economic growth, with limited success for US GDP. We augment bank balance sheet data with that from shadow banks, mutual and pension funds, insurers, and brokers, finding that aggregate FI assets lead GDP growth and forecast recessions up to three quarters in advance, although the weight of each FI group in the aggregate fluctuates. Aggregated FIs expand their total assets when the slope of the Treasury yield curve is upward sloping and the effective Federal funds rate is higher, thereby revealing the importance of monetary policy in the financial intermediation process. We further show that aggregate FI contains leading information about investment and consumption and predicts industrial production and unemployment.

Managing the Treasury Yield Curve in the 1940s

This paper examines the efforts of the Federal Open Market Committee (FOMC) to first control, and later decontrol, the level and shape of the Treasury yield curve in the 1940s. The paper begins with a brief review of monetary policy in 1938 and a description of the period between September 1939 and December 1941, when the idea of maintaining a fixed yield curve first appeared. It then discusses the financing of U.S. participation in World War II and the experience with maintaining a fixed curve. The paper concludes with a discussion of how the FOMC regained control of monetary policy in the second half of the 1940s. The Committee’s efforts offer two lessons in yield curve management: (1) the shape of the curve cannot be fixed independently of the volatility of interest rates and debt management policies, and (2) large-scale open market operations may be required in the course of refixing, from time to time, the shape of the yield curve.

Time-varying response of treasury yields to monetary policy shocks

Purpose This paper aims to examine the Treasury bond yields response to monetary policy shocks in Tunisia under a heterogeneous economic environment. Design/methodology/approach Using a traditional fixed coefficient model, the impact of monetary policy changes on the term structure of interest rates for the whole period from January 2006 to December 2016 is estimated first. Then the stability of this relationship by distinguishing two sub-periods around the revolution of January 2011 is studies. To investigate how the relationship between the monetary policy and the Treasury yield curve evolves over time, a time-varying parameter model is estimated. Findings The results show that the impact of monetary policy is more pronounced at the short end of the yield curve relative to the longer end. Furthermore, this impact declines significantly across all maturities following the revolution and exhibits wide time variation. This evidence supports the negative influence of high levels of uncertainty on monetary policy effectiveness and highlights the desirability of more active monetary policy, especially in turbulent environment. Research limitations/implications The impact of uncertainty on the effectiveness of monetary policy shocks needs to be explored further in future research to understand the structural sources of uncertainty and their dynamic interactions with monetary policy and risk aversion in asset markets. Practical implications A more active role of the central bank to influence the yield curve mainly through Treasury bond purchases covering medium and long maturities may be warranted. Communication also needs to be reinforced to ensure predictability of the monetary policy stance. Originality/value This paper extends the empirical literature on the pass-through of monetary policy to interest rates for an emerging country in context of transition by estimating a state-space model to test the time-varying behavior and examine the influence of increased economic uncertainty on monetary policy effectiveness.

Measuring Bond Investors’ Risk Appetite Using the Interest Rate Term Structure

The distribution of term premia includes additional information beyond that contained in the traditional three-parameter decomposition of the Treasury yield curve. The author finds that the position and the size of the curvature hump are both important. A yield curve hump positioned at longer maturities appears to be consistent with lengthier time horizons of investors and with more risk-seeking behavior. The author finds that this new parameter Granger drives both future US economic activity and bond returns. TOPICS:Fixed income and structured finance, fixed-income portfolio management, statistical methods

Business Cycles, Optimal Interest Rate, and Recession Forecast From Yield Curve, Unemployment, GDP, and Payrolls

We developed a factor regression model, nicknamed “GUPTY”, to study the business cycles, and their relation to the monetary policy. It covers several major macro-economic quantities, including unemployment rate, GDP, and weekly payrolls in the U.S. after WWII. The model postulates that these macro-economic quantities are predominantly a nonlinear function of the 10Y and 3M interest rates in the Treasury yield curve, and their regressions share similar model specification. The concept of the business cycle is injected into the model through the yield spread between 10Y and 3M rates. The notion of the natural rate of unemployment and GDP growth comes out of the model naturally. An optimization framework is proposed to determine the optimal short-term interest rate, which is a central concern for policymakers – Once the model parameters are obtained from regressions, the model is inverted to calculate the optimal 3M rate via optimization of the squared errors. Its first-order analytic solution is derived. We find that FED’s interest rate policy has leaned heavily on the unemployment data. We also observe that, on the optimality, the model GDP growth rate matches the realized data almost perfectly. We discuss some policy challenges, and attempt to provide an explanation for the plunge of GDP growth after 2009 due to super-low 10Y rate, in which the “2% GS10 floor hypothesis” is suggested. Unemployment’s 1Y momentum and its 6M acceleration are used to construct a Hidden Markov Model for real-time recession probability. Furthermore, a Random Forest model is developed to forecast recession up to 24 months ahead. And we show recession forecast is closely associated with yield spread inversion.

Echo over the Great Wall: Spillover Effects of QE Announcements on Chinese Yield Curve

This paper examines the spillover effects of announcements on quantitative easing (QE) conducted in developed economies on the Chinese Treasury yield curve. In spite of China’s firewall of cross-boarder capital control, we find that the US QE announcements move the Chinese yield curve immediately and significantly, through the channels of signaling and portfolio balancing. The results are robust across a variety of event analysis methods using reduced-form or structural models. Using the heteroskedasticity assumption for shock identification in an arbitrage-free term structure model, it was identified that the impact of US QE on Chinese yields was comparable, in size and significance, to China’s own monetary policy shocks.

Co-jumping of Treasury Yield Curve Rates

We study the role of co-jumps in the interest rate futures markets. To disentangle continuous part of quadratic covariation from co-jumps, we localize the co-jumps precisely through wavelet coefficients and identify statistically significant ones. Using high frequency data about U.S. and European yield curves we quantify the effect of co-jumps on their correlation structure. Empirical findings reveal much stronger co-jumping behavior of the U.S. yield curves in comparison to the European one. Further, we connect co-jumping behavior to the monetary policy announcements, and study effect of 103 FOMC and 119 ECB announcements on the identified co-jumps during the period from January 2007 to December 2017.

Which leading indicators have done better at signaling past recessions?

In this article, I analyze a broad range of leading indicators?economic or financial data series that change in advance of the rest of the economy?to see which ones have done better at signaling past U.S. recessions.1 I also use these leading indicators to form a new index that outperforms existing leading indexes and the Treasury yield curve at signaling historical downturns.

Reconstructing the Yield Curve

The constant-maturity zero-coupon Treasury yield curve is one of the most studied datasets. We reconstruct the yield curve using a non-parametric kernel-smoothing method with a novel adaptive bandwidth specifically designed to fit the Treasury yield curve. Our curve is globally smooth while still capturing important local variation. Economically, we show that applying our data leads to different conclusions from using the leading alternative data of Gurkaynak et al. (2007) (GSW) when we repeat two popular studies of Cochrane and Piazzesi (2005) and Giglio and Kelly (2018). Statistically, we show our dataset preserves information in the raw data and has much smaller pricing errors than GSW. Our new yield curve is maintained and updated online, complemented by bandwidths that summarize information content in the raw data. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

What Can Markets Tell Us about Future Economic Growth? Historical Predictive Power of the Bond, Stock and Real Estate Markets

Do bond, stock and real estate markets provide useful signals for forecasting future GDP? Using market data since 1970, we examine several market signals and find that, in general, they have not been very helpful in predicting next year’s GDP, or the change in next year’s GDP versus last year. If we fit the market signals to the full data period, the signals explained 48% of the variation in next year’s GDP and 44% of the variation in the change of next year’s GDP. However, what matters more to investors is the predictive power of market signals. In terms of prediction, where only contemporaneous market signals are used to form forecasts, the predictive power of market signals has been poor. Relative to other market signals, the Treasury yield curve slope, stock market returns and the change in CAPE have exhibited better predictive power. However, even their absolute predictive power has been relatively low, and this power has fluctuated over time. To improve the predictive power of market signals, we explore combining many signals, and selecting them dynamically. Nevertheless, the average prediction error (RMSE) for next year’s annual GDP was 2.05 (in annual GDP percentage points), and 2.45 for the GDP change.