57 Spillover Effect on the Stock Market and Bond Prices in Relation with GARCH

Abstract

This study examines the spillover effect between bond and stock markets in the U.S. using GARCH. The finding of a unidirectional spillover flow from bonds to stocks in the U.S. is discussed in the light of new marketplace variables that have been introduced into the markets in the previous decade. These variables include the rise of HFT, algorithm-driven trading, and central banking interventionism via unconventional monetary policy. The effect on forecasting volatility, price and return of asset classes, studied through the lens of other commodity price movement and volatility—such as oil and gold markets—creates a compelling picture for why GARCH models may need to be reworked to incorporate new data regarding the new ways in which the 21st century marketplace is using technology and central bank interventionism to shape market movements and market outcomes.

Table of Contents



1 Introduction 4

1.1 Why this research is important 4

1.2 What this paper examines 6

1.3 The most important findings and contributions this study makes to existing literature 7

2 Literature Review 11

3 Hypothesis Development 19

3.1 Spillover as a Result of Unconventional Monetary Policy 19

3.2 Low-Vol Complacency 20

3.3 Spillover is Unidirectional 21

3.4 Algorithm-Driven Trading and Market Movement 22

4 Data and Methodology 23

5 Findings 27

6 Analysis 37

7 Conclusion 47

References 49





1 Introduction



Globalization has substantially altered the working dynamics of markets of both developed and emerging nations. Sakthivel, Bodke and Kamaiah (2012) note that as a result of globalization, the spillover effect has become far more commonplace than what once used to be the case, for “world financial markets and economics are increasingly integrated due to free flow capital and international trade” (p. 253). This phenomenon has been seen most recently in the spillover effect of the 2008 housing bubble crisis in the U.S., which was quickly felt across sectors both domestically and internationally and served as an indication of the interconnectedness of modern markets and economies in the 21st century. Understanding spillover among equities and bonds along with other commodities, such as oil and gold, can play a significant role in how to forecast volatility and better manage funds, such as those responsible for guaranteeing pensions for workers in the not-too-distant future.

1.1 Why this research is important

Research in the spillover effect between the stock and bond markets is important because each represents one of the major asset classes for investors. According to Dean et al. (2010), there a number of theories which may explain the relationship between equities and bonds—the asset substitution hypothesis, the financial contagion hypothesis, the news specificity hypothesis, the news decomposition hypothesis, and the asymmetric price adjustment hypothesis. In short, the two asset classes are more or less seen as two sides of a negatively correlating coin. However, with the introduction of QE (quantitative easing) in the U.S., the relationship has seemingly altered, especially with respect to volatility, as a rebalancing towards both US and non-US assets was triggered with QE1 through 3 (Fratzscher, Lo Duca, Straub, 2016). The CBOE Volatility Index US:VIX on a two-year chart shows an ever-dwindling sense of volatility or risk in the market.



Figure 1. Two-year chart for VIX.



Source: CBOE Volatility Index (2017)



Park and Um (2016) likewise highlight the spillover effect of unconventional monetary policy in the US on bond markets and note that a mere mention of “news” of unconventional monetary policy in the US is enough to trigger a short-term spillover in the bond market in Korea. Short-term hedges in gold during such swings have been found by Baur and Lucey (2010) to be effective in protecting a portfolio from headline-driven spillover between equities and bond markets. But for forecasting the spillover effects, today’s GARCH models may need to take into consideration a rapidly changing market culture—namely, one that is driven by algorithms which have been blamed both for flash crashes and for melt-ups in recent years, along with the “existence of pure contagion” (Jayech, 2016, p. 631; Kirilenko, Kyle, Samadi and Tuzun, 2017). The velocity of volatility, the absence of liquidity (save for that provided by central banks), counter party risk, sovereign debt, risk parity, and overall fundamentals (AMZN is trading currently at...

In short, both QE and the rise of machines in trading have introduced new dynamics into the market that traditional GARCH models may not effectively factor into their equations. The result is that while some models may serve as better predictors of market movement, there remains a lack of fundamental understanding of spillover in relation to institutional forces.
This raises the question: Has the duel-engine of algorithm-driven trading and the central bankers’ policy of QE (bond purchases) created an irrational spillover effect between the stock and bond markets that has, as one result, crushed volatility to all-time historic lows? Has marketplace complacency taken hold because of an 8-years-long impression that the central banks will indeed do “whatever it takes” to backstop markets—as Draghi stated in 2012 with regard to the ECB’s role in maintaining the integrity of the European marketplace (Draghi, 2012). The VIX used to rise and fall much more frequently than it does today, as the chart above shows.

1.2 What this paper examines

This paper will examine volatility in the light of spillover between equities and bonds to more deeply understand the present day markets and what can be inferred from them. To assess this relationship, 20-year charts spanning from 1997 to 2017 will be used.

Equities offer investors ownership in a company; in times when the economy is booming, ownership of stocks can be financially rewarding. Bonds offer investors the opportunity for fixed income through interest; when the economy is stagnating or slowing down, investors can turn to bonds as a kind of safe haven. Oil as a commodity is often seen as an indicator of overall market sentiment, while the price of gold is commonly viewed as a hedge or safe haven in times of economic uncertainty. As more investors seek one or the other, prices go up—the basis of a supply-and-demand market economy. In the case of the bond market, higher bond prices produce lower yield. When unconventional monetary policy (such as the preservation of low interest rates over a significant amount of time, which can substantially distort markets) becomes conventional the world over, an unknown mechanism has entered into the dynamic and old formulas must be considered anew.

1.3 The most important findings and contributions this study makes to existing literature

This paper contributes to the existing body of literature examining monetary policy and asset price volatility—particularly to the work by Bernanke and Gertler (2000), which found that “it is desirable for central banks to focus on underlying inflationary pressures” (p. 17). One of the Fed’s aims over recent years has been to achieve 2% inflation. The pursuit of this aim should be factored into modern economic theory as it has had and will go on to have a substantial effect on economic forces. As Benford et al. (2009) show, “the aim of quantitative easing is to inject money into the economy in order to revive nominal spending” (p. 91). However, after eight years of QE, the inflation target is still as of yet unachieved, just as it was during the economic crisis of thirty years ago (Llewellyn, Nieto, Huertas and Enoch, 1992). Heller (2017) describes the current situation as one of “monetary mischief” brought about by central bank tampering that has inordinately inflated asset prices, decoupled markets from fundamentals, and created an “everything bubble” that Dowd and Hutchinson (2017) describe as the result of the “biggest monetary experiment ever” and the path towards what could quite possibly turn out to be “the biggest ever collapse” (p. 306). With this in mind, a fund manager of state and local pensions may be forgiven for bringing in a 0.6% return in 2016 when the assumed rate is more than 5% (Aubry, Crawford and Munnell, 2017)—market forces are increasingly entering into unknown territory and few managers have the tools to respond to what is more and more resembling a “wild west” type of economy, driven by AI, central planning and headlines instead of fundamentals.

It is necessary to add to the existing literature because of the present climate in which many financial managers find themselves today. As the recent report on pension funds by Aubry et al. (2017) has shown, there is a high degree of difference between the assumed rate of return for many funds and the actual rate of return. Indeed, the figure below indicates how poorly pension funds have performed when the economy experiences a downturn (notably following the dotcom bubble and the housing bubble).



Should spillover effect…