Central Banks Response to Movement of Stock Prices
Disclaimer: This work has been submitted by a student. This is not an example of the work written by our professional academic writers. You can view samples of our professional work here.
Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
Published: Tue, 13 Mar 2018
This paper discusses a hotly debated topic from a new perspective. Since past papers focus on whether or not should central banks respond to the movement of stock prices, this dissertation turns to the problem if different central banks give same response to stock index. Starting from the general Taylor rule, the author uses OLS and GMM method as one strategy to do analysis, and utilizes ADF and Granger causality test as an alternative approach to confirm the conclusion obtained by former technique. By checking the result of data of the USA and China, the author concludes that different central banks may respond disparately to the movement of stock prices.
Chapter one. Introduction
In the contemporary world, financial asset is becoming increasingly important. The year 2007 witnessed the sub-prime mortgage loan crisis, occurring in the United States and then influencing the whole capital market as well as the real economy all over the world. The downward plunge of stock market trapped numerous investors and firms, which jeopardized the consumption, investment and output as individuals and companies had less fortune in this bear market. Out of question, the role of central banks is to implement monetary policy and control the money supply of its nation, but the dramatic movement of equity prices would undoubtedly deeply affect the money supply.
Therefore, in recent years, the movement of asset prices has attracted wide and serious social attention all over the world. Specifically, as stock market plays an increasing important role in one nation’s economy, many people believe that equity prices have exerted vital influence on the decision of monetary policy. It triggers a hot debate as whether or not should central banks respond to the movement of equity prices. In my opinion, due to the different situation in distinct nations, reaction of central banks might be different. To ascertain this question, this paper will carry out some analysis on this issue based on the data from two representative countries: America and China. One is the biggest economy in the world; the other is the largest developing country. By checking different situations, we will get an elementary relationship between the response of central banks and the movement of stock prices, to see if the same response is given by different central banks.
Actually, since the launch of the Taylor rule by Taylor (1993), the number of research on the issue of whether or not should central banks respond to stock prices has becoming more and more, and the views on this problem have divided into two parts. Bernanke and Gertler suggest (1999) that central banks only need to react to movement of asset prices when they tend to influence future GDP and output gap. However, with fast development of capitalization and marketization globally, some economists believe that asset prices undoubtedly are more important than before, and should be taken into consideration of monetary policy when they are made. Given this background, I believe it is necessary to conduct an analysis to make sure if the different opinions derived from different source of data they investigated.
The structure of this paper includes four chapters: the first part is introduction, giving an outline of the whole article, with the objective and motive to do this research. Literature review is listed as the second chapter, showing different opinions supporting by various arguments and methods. In order to better understand of the relationship between interest rate and the movement of stock market, the author has read numerous past academic papers and articles, so the research could be completed smoothly. Next chapter is the main body of the research. Firstly it gives details of the raw data we are investigating, which include quarterly figures of interest rate, CPI, GDP and stock index of the United States and China. Through a concise comparison, I believe it could be helpful to understand the conclusion we will get. And then it shows the detailed methods to deal with two groups of raw data, aiming to give an answer if central banks give same response. In addition, as Taylor rule method is traditional way to conduct this research, the author also starts with it, followed by an augmented Taylor rule model to cope with the main concern of this paper—if there is a relationship between interest rate and stock prices. During this process, OLS and GMM method are employed. Besides, subsample test is also given to show a reasonable analysis. What is more, in order to confirm the conclusion obtained by American data, ADF test and Granger causality test are also used to give an objective and precise answer. In addition, data are analyzed separately. Finally, the last part is the conclusion. As an overview, limitations and recommendations are also given in this part.
Chapter two. Literature Review
It is universally acknowledged that one target of macroeconomic policy is to maintain price stability. As a central bank, its primary mission is to utilize monetary policy to keep stable price indices. Stock market plays a vital role in modern economy, and the amount of transacting dollars in stock market occupies a large percentage of total money supply. It is believed that, more often than not, stock markets reflect the status of social economy. For instance, a boom in stock indices may suggest a strong economy, so families and firms are willing to increase their expenditure, which undoubtedly gives an inflation expectation. Especially, with the development of financial market in China, firms find that they have more alternatives to raise funds to invest a new product or research of a new technology. Besides, when the stock market is prosperous in the period from 2005 to 2007, individuals from all walks of life start to purchase private cars, which are thought to be a luxury in the past. All these phenomenon suggest that stock market directly or indirectly influences CPI and GDP to some extend. Therefore, an inflation-minded central bank may take actions to control this trend when real inflation occurs and pay more attention to the movement of stock index.
However, whether the stock index is a trigger for central bank to take measures is still a hot debate. In terms of this question, Fuhrer and Tootell (2008) contend that little proof is found to prove that FOMC (Federal Open Market Committee) reacts to stock market directly. In their research, they adopt two approaches. The first one is a regression model taking the form of Taylor rule, attempting to document the relationship between stock prices and federal funds rate. Unfortunately, this method fails to show the mentioned purpose above in light of the role of stock price appearing differently. Given the poor instruments and inefficient use of equity price in approach I, Fuhrer and Tootell use Greenbook forecast combined with Taylor rule as their new instrument, comprising up-to-minute information and other necessary elements as well. Specifically, they test inefficiency of forecast by virtue of Ordinary Least Squares estimation at the beginning, and then examine if the equity prices influence the decision of FOMC independently. In addition, data are separated into pre- and post- Greenspan eras to test the efficiency. Interestingly, the early part of the data illustrate that the forecasts are inefficient in terms of equity prices, but they does not comment in this paper. After examining the results of coefficient and p-values, they conclude that the FOMC did not respond to the stock indices directly. Nevertheless, this approach bases on the belief that the staff at the Federal Reserve Board could afford a more thorough opinion after deep research targeting at then situation before every FOMC meeting, missing the consideration that even specialists could suffer from some behavioral pitfalls such as representative and other affect heuristics.
Ben Bernanke, the successor of Greenspan, addressed this issue as well. Bernanke and gertler (1999) believe that central banks should not react to fluctuations of equity prices since they are at most signals of inflation, but central banks should keep a close eye on them. They utilize the BGG model, which is new Keynesian model per se, to mimic the reaction of monetary policies in the context of asset price bubbles, taking the United States and Japan as examples. In 2001, they restate their point of view, but give a complementary comment by virtue of a similar approach. Compared to their previous work, they consider shocks of equity prices and shocks driven by technology in latter article, providing a new policy evaluation approach. The result is as expected, responding to changes of stock indices afford no extra benefit. Unlike other economists holding this stubborn point of view, Bernanke attaches importance to the financial markets, arguing that they must be given intense and successive observation.
There are also some other economists believing that central banks should ignore the fluctuations of stock values. Bullard and Schaling (2002) conduct a research adopting the form of an equation from Rotemberg and Woodford’s(1998) work, taking arbitrage relationships and equilibrium conditions into consideration. And they find that if stock prices are contained in Taylor-type rule, the social economy will be damaged. Besides, it is useless to add stock prices to government’s reaction function whose purpose is to obtain better monetary policy.
However, economists believing that central banks should respond to asset prices also have compelling reasons. Interestingly, using the similar model as Bernanke and Gertler, Cecchetti et al. (2000) obtain a different conclusion. They allege that it is of great benefit for central banks if they take changes of stock prices into account when monetary policy is made. The reason why they get a divergent answer, according to Cecchetti et al, is because they utilize more possible policy responses than Bernanke and Gertler did. What is more, they also employ a model developed by Batini and Nelson (2000), to compare the effects of central banks reacting to different source of forecast. Needless to say, they get the desirable conclusion in favor of their argument. They state that the main reason affecting central banks’ decisions is that equity prices are volatile and their movements are hard to measure. But they believe it should not be a reason for not adopting stock price as an input when making monetary policy. In addition, they point out that the measurement of equity price misalignment is a fundamental process to estimate other variables influencing the amendment of monetary policy.
Goodhart and Hofmann (2000) believe that as long as asset prices influence output significantly, central banks would better to take this element seriously and incorporate it into decision of monetary policy. Besides, they also stress that their research is not to encourage a stubborn active monetary policy response to asset prices, but with a purpose to show that asset prices contain valuable information to estimate other variables of the economy.
Nowadays in China, there are also some scholars and researchers concerning about this problem. They once tried to test if Taylor rule is useful in China. Xi and Liu (2004) carry out a research on this issue based on Taylor rule. They use the augmented static and dynamic Taylor rule to examine if there is relationship between interest rate and stock bubbles. Due to the answer they obtained, they concluded that China’s monetary authority react negatively to the stock overvaluation, that is to say, central banks did not use interest policy to offset the unreasonable boom of stock market.
Different opinions seem to be supported by a variety of reasonable reasons. But to my perspective, different results may derive from the different data source they use. For example, developed countries with well-established financial system and related legislations may obtain a result that central bank respond to the movement of stock prices; but as to some emerging nations, with only a short history of financial system, which perhaps is not sound enough, are tend to secure a conclusion that central bank does not react to the fluctuations of equity prices.
Chapter three. Empirical analysis
Since the main concern in this dissertation is to examine whether different central banks respond same to the movement of equity prices, a well-known classical approach, known as Taylor rule, is firstly employed to test the simple relationship between interest rate, GDP and output gap. Then next step is to put focus on the significance of stock prices. To ascertain the role of stock prices in determining other economic parameters, OLS method is utilized to get an ordinary relationship between interest rate and stock prices. If the result shows stock prices is insignificant, we can complete our research. But if any connection if found between them, we can not say there is a direct relationship between them, we should take a further study, which is to use GMM approach to test if the equity prices influence interest rate directly or just a effect transmitted by other variable such as GDP or output. Up till now, the work almost is finished, but one more step needed to be done is to test if there are breaks in the sample, so the subsample test is of great importance. In addition, in order to reassure the conclusion secured by the data of the United States, another strategy is utilized to achieve this goal, which is unit root test and Granger causality test. One thing needed to mention is data of different countries are analyzed separately.
There are two groups of data used in this paper. The first group is economic variables of United States, which are the quarterly data of the federal funds rate, CPI, real GDP and the S&P500 index, collected from 1990 to 2009; totally, there are 80 observations. The second group of data is information about China’s economy, which is also quarterly data of interest rate, CPI, real GDP and Shanghai stock index. But due to certain reason, there are only 44 observations, from 1997 to 2007.
3.1 Data source
Now the author is going to present the raw data of quarterly Interest rate, CPI, GDP, and stock index of both countries. The data of the United States are collected from the first quarter of 1990 to the fourth quarter of 2008, but in order to keep integrity and continuity, the figures of China only from 1997 to 2007. By this comparison, it is easy for us to understand the future result we will get. The graph on the left is about information of the United States. On the other side, it is China’s situation.
Figure 3.1 the interest rate of the United States and China
The first pair of graphs describes the interest rate of United States and China. We take federal funds rate as the interest rate in the United States. As to China, we use inter-bank lending rate as interest rate here. From the pictures, we can see that the federal funds rate goes up and down all the time, leaving room for us to think America’s Federal Reserve adjust the interest rate frequently to meet certain requirement. When it comes to the situation in China, the interest rate drop dramatically from 1997 to 1999, and then it keeps stable.
Figure 3.2 the CPI of the United States and China
These two bar charts show Consumer Price Index(CPI) in both nations. It is manifest that the CPI of America keeps increasing, meanwhile China’s CPI nearly does not change before 2001, and booms gradually between 2002 to 2005 and then keeps decreasing till 2007.
Figure 3.3 the GDP of the United States and China
The third group of pictures illustrates the Gross Domestic Product (GDP) of the mentioned countries. One thing to mention is that they are calculated in different unit: the GDP of America is measured by dollar, while the GDP of China is calculated by Yuan, which is the basic unit of money in China. We can see from the two graphs that the GDP of both nations are on the increase gradually. Interestingly, it is manifest that the GPD of every fourth quarter in China is far greater than any other quarters.
Figure 3.4 the stock index of the United States and China
The last group of graphs depicts the stock prices in both countries. As is well known, there are several stock indices in the two countries, so we just choose the typical one as an example to illustrate the issue we concern. From the first picture, we can see that S&P 500 index experienced a stable increase before 1999. After that, it fluctuates drastically. As to the Shanghai stock index, after almost 10 years of fluctuation form 1997 to 2005, it surged to a new level.
3.2 Data processing
The analysis is done separately based on different source of data. Initially, we start with the situation in the United States.
3.2.1 Data of USA
As a result of a variety of reasons, the raw data are not suitable for analysis; essential steps need to be taken for further research. To begin with, we should get the inflation rate, thus CPI is used to get inflation by virtue of the following formula: Inflation=100*(CPI-CPI(-4))/CPI(-4)
Then, we take the logarithm form of real GDP. What is more, in order to get potential output, Hodrick-Prescott filter is employed, which is a detrending procedure that with a goal to obtain the trend from a time series. After doing this by eviews, we can get the following picture, which shows that gaps and cycle can be found in this time series. To get output gap (y_gap), we use this equation: y_gap=100*(y-y_star)
Figure 3.5 information of output gap
Completing above steps, we can arrived at a first impression of relationships
among federal funds rate, inflation and output gap.
3.6 relationship between interest rate, inflation and output gap
As we can see from this graph, it seems that output gap, inflation and federal funds rate roughly move together, but it is manifest that in the period from 1994 to 1998, and also from 2004 to 2008, federal funds rate rise dramatically above the other two.
Following that, we now can estimate the standard Taylor rule by Ordinary Least Squares (OLS) and we can get the following table, which means =0.389, =0.687. To test these figures, we use F-test. The null hypothesis is =o, =o. From the table, we get that the value of F-test is 21.511, which is far greater the critical value at 1% significance level, so we can reject the null hypothesis, meaning that inflation and output gap are useful in influencing the federal funds rate. In addition, when we looked at R-squared, we can get a result that the regression do not explain the federal funds rate very well, which is lower than 50%.
Table 3.1 result of OLS test without stock index
Based on the fundamental analysis and data processing, now, we turn to the key issue of this paper, whether or not should central banks respond to the movement of equity prices. Initially, we should take year-to-year change of S&P500 index. By plotting the federal funds rate and the year-to-year change of S&P500 index in one graph, we can not help guessing that there is a relationship between them.
Figure 3.7 relationship between interest rate and stock price
To obtain a formal conclusion of the impact of stock price, we need to use a augmented Taylor rule, the equation is like this:
As you can see, there are expected variables in the formula, so we need to estimate them. Fuhrer and Tootell (2008) provided an approach to solve this problem, they estimate the unknown variables by the following equations and save the residuals.
Then, we can get inflation forecast by subtracting the residual series from the actual inflation series. The output gap forecast can be secured in the same way. Finally, we can estimate the augmented Taylor rule by OLS, adding lags of stock prices to check how many should be involved.
Table 3.2 result of OLS test including stock index
Compared to the p-value, we can get from the table that only the value of one lagged S&P500 index is lower than 0.05, which means that one lagged value of S&P500 index is helpful in explaining federal funds rate.
Up till now, we have already got the conclusion that S&P500 index is helpful in affecting the federal funds rate, however, through influencing the GDP or CPI, stock prices also can have an effect on the federal funds rate. As a matter of fact, with the boom of equity prices, consumers as well as the firms are wealthier than before, at least for the time being, they tend to spend more and invest more, thus the GDP and the CPI will undoubtedly be influenced. Therefore, we would like to examine whether stock prices affect the federal funds rate directly or indirectly. To achieve this goal, we need to estimate the augmented Taylor rule by GMM(Generalized method of moments). By doing GMM test via eviews, we get the following table.
Table 3.3 result of GMM test
As we can see from the table, the p-value of S&P500 index now is 0.617, which is bigger than 0.05, we can conclude that the S&P500 index may influence the federal funds rate indirectly.
In terms of regression model such as Taylor rule, it is of great significance to take a further analysis to see if monetary policy looks stationary or whether there are changes in the sample we choose. As is known to all, in 2006, Bernanke succeeded Greenspan as the Chairman of Federal Reserve. It is believed that the monetary policy would go towards more of inflation targeting since the new chairman is an advocate of inflation targeting. As a consequence, we believe that there would be a change of monetary policy, that is to say, there is a breakpoint in our sample.
Thus, Chow’s breakpoint test is employed to test if there is a breakpoint, and we choose the year 2006 as the break point due to the appointment occurred in that year. This method divides the sample into two separate parts, to see if there is significant difference in the equation we estimated. Here it is Taylor rule in our case. Firstly, we get the sum of squared residuals of the entire sample, and then we should get the sum of squared residuals of the different subsamples, finally, we should compare their sums. In addition, that there is no structural change is the null hypothesis here.
Table 3.4 result of Chow breakpoint test
Given the p-value we get, it is obviously that we could reject the null hypothesis that there is no structural change.
Based on the OLS and GMM method, we have arrived at the conclusion that Federal Reserve responds to the movement of S&P 500 index. In order to confirm this result, we will analyze this problem by virtue of an alternative approach, Granger causality test. Despite its name, it does not suggest true causality. Actually, it only provides information that if one time serious is helpful in forecasting another. Before that, we should do the so-called ADF test (Augmented Dickey–Fuller), which is a unit root test to examine the stability in a time series sample.
Now we are going to employ stata instead of eviews as econometrics software. ADF test is conducted on the following variables: federal funds rate, inflation, output gap and S&P 500 index. The null hypothesis is there is no unit root. The result is as follows:
Table 3.5 result of federal funds rate
Table 3.6 result of inflation
Table 3.7 result of output gap
Table 3.8 result of stock index
The larger p-value implies that the null hypothesis of no unit root should be rejected, thus the time serious contains unit root, meaning it is I (1) integrated non-stationary serious, which are federal funds rate, output gap and S&P 500 index. Inflation is stationary serious.
Now, we continue to use Granger causality test to test if the variable is useful in explaining the federal funds rate.
Table 3.9 table of Granger test of inflation
The p-value is too big to accept the null hypothesis, so we conclude that inflation is the Granger cause to federal funds rate.
Table 3.10 table of Granger test of output gap
The result shows that the null hypothesis does not held, so we can say output gap is the Granger cause to federal funds rate.
Table 3.11 table of Granger test of stock index
The p-value is also very big, thus the null hypothesis should be rejected too. The conclusion is that the S&P 500 is the Granger cause to federal funds rate.
By employing ADF test and Granger causality test, we also get the conclusion that the S&P 500 index is helpful in explaining the federal funds rate, which coincides with the conclusion by former strategy.
3.2.2 Data of China
Now we will transfer our focus onto the relationship between Chinese monetary policy and stock prices of Chinese mainland stock market. With the development of Chinese economy, more and more people begin to invest in a wide range of stocks. Nowadays, Chinese stock market has a very large population of investors. Undoubtedly, it will exert a big influence on civilian’s daily life. For example, the period between the year of 2006 to the year of 2007, witnessed a larger population buying new cars than any other years, since there was a spectacular stock market boom just at that time, which would played a significant role in other variables such as CPI and GDP.
The data used in this paper includes quarterly figures of interest rate, CPI, GDP and stock index of Shanghai exchange from 1997 to 2007. In addition, inter-bank lending rate is taken as the interest rate here.
The first impression is given by the following graph; it is manifest that from 1997 to 1999, interest rate drops dramatically while equity prices fluctuate at that time. And we also can see that from the year of 2005 to the year of 2007, stock index keeps soaring while interest rate only changes slightly.
Figure 3.8 relationship between interest rate and Shanghai stock index
Now I am going to use OLS method to test the augmented Taylor rule. By doing the same steps as we did before, we get the following table.
Table 3.12 result of OLS test of Chinese data
As we can see from the table that the p-value of shanghai stock index is bigger than 5%, which means it is insignificant both at 1% and 5% significant level. Compared to the result of S&P 500, whose conclusion is that stock price is helpful in explaining the decision of monetary policy making, the result of Chinese market could be expected. Because Chinese stock market is still not mature, equity prices often go up and low unexpectedly, it is hard for policy makers to make monetary policies based on such a volatile variable.
As we obtain a conclusion that central bank does not respond to the movement of stock prices in China, we can complete the analysis on the data of China by now.
Chapter four. Conclusion
4.1 Limitations and recommendations:
When the author tries to answer the question that whether or not should central banks respond to the movement of equity prices, a simple Taylor rule is used. But Fuhrer and Tootell (2008) argue that this could be a poor instrument for central banks to forecast future information since it contains ex-post data in this model, so they suggest a more careful analysis which could reflect information in time. In my opinion, this is a good advice which can provide a more precise result.
As to the data we used in this paper, the author only takes USA and China as examples. Since the issue we concerned is whether central banks respond same to the movement of equity prices, so two examples are not enough to answer this question. What is more, when it comes to the data of China; the population of the sample is too small, which could lead to a bias answer. In order to achieve a more exact result, future researches should collect more data through a long period and from more countries as well.
As is known to all, both nations have several stock indices. But the author chooses S&P 500 and Shanghai stock index randomly from a group of stock indices without much deep research, it also could be a source of bias. A better conclusion should base on the analysis of each stock index, to see how important it is to the nation’s economy, and then select a more precise one.
Up till now, we can answer the question raised in introduction: Whether or not do central banks respond same to the movement of equity prices. The author raised the issue in introduction, and then read numerous articles and academic paper to master the main opinions nowadays, during this process, the author established a framework to carry out the analysis. Taylor rule method is chosen as the main instrument. Then the author continues a further analysis to study the relationship between interest rate and stock index based on the OLS and GMM method. And then subsample test is conducted. What is more, in order to reassure the result obtained by former strategy, ADF test and Granger causality test are also utilized .During this process, data of USA and China are analyzed separately. Based on the above analysis, we can say that different central banks may provide different response to stock prices. At least, it is true in terms of situation in this paper.
In this age of change, more and more elements begin to play a role in influencing the smooth run of the economy. Financial asset, especially stock prices are far more important than before. In addition, the real estate market and exchange rate also exert vital influence in determining the destiny of a nation’s economy. Therefore, even an inflation-minded central bank should keep a close eye on the variables which affecting the soundness of economy.
The nations we talked about in this paper are two main representatives of world economy. By checking the result, we have already known that the Federal Reserve seems to give a positive response to the movement of equity prices. When making monetary policies, they will consider the stock index. As to China, its central bank seems not to response actively. Actually, it is easy to understand the difference response. It is universally acknowledged that the United States has experienced a long history of financial market, with sound financial system and legislation. Even though, it suffered several financial crises, which deeply affected the real economy of the USA and even the world. China has developed fast since its reform and opening-up in 1978. But the history of its financial market is very short, and still need to learn form developed countries as well as broaden and open its financial market to the outside world. Actually, with the rapid development of financial market, the Chinese authority should pay more attention to these elements due to the fact that it is easy for bubbles to form in this kind of markets. As a consequence, it is of great significance for Chinese central banks focus on the stock index as well as real estate market. When there is necessity, Chinese central bank may also take the stock index into account at the time when they make monetary policy to keep a healthy economy.
Overall, based on the analysis of this paper, we can conclude that central banks may respond disparately to the movement of stock prices.
Batini, Nicoletta and Nelson, Edward (2000), “When the Bubble Bursts: Monetary Policy Rules and Foreign Exchange Market Behavior.” ,Working paper, Bank of England, 2000.
Bernanke, B. (2002), “Asset-Price `Bubbles’ and Monetary Policy.”, Remarks by Governor Ben S. Bernanke before the New York Chapter of the National Association for Business Economics, New York, Federal Reserve Board, 15 October 2002.
Bernanke, Ben and Mishkin, Frederic (1997), “Inflation Targeting: A New Framework for Monetary Policy?” ,Journal of Economic Perspectives, Spring 1997, 11(2), pp.97-116.
Bernanke, B.S. and M. Gertler (1999), “Monetary Policy and Asset Price Volatility.”, Federal Rese
Cite This Work
To export a reference to this article please select a referencing stye below: