1
The 2015 Chinese stock market bubble.
An analysis of the rise and fall.
By: Sjaak Hoogstraaten (10466886)
Graph 1, the Chinese markets combined stock index. 0 1000 2000 3000 4000 5000 6000 7000 8000 9000
2 Statement of Originality
This document is written by Student Sjaak Hoogstraaten who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Inhoud
1. Introduction ... 4
2. Literature review ... 5
2.1 Literature on price volume relation. ... 6
2.2. Interest rate, stock price relation. ... 8
3. Empirical tests and results ... 8
3.1 Interpretation of results. ... 11
4. The Chinese bubble ... 12
4.1. Volume of trading relation to stock prices ... 13
4.1 Political catalyzation. ... 14
4.2. Shadow banking system ... 16
5. Conclusions ... 18
4
1. Introduction
In the recent past, the Chinese economy has been growing fast. The economy moved from being mostly closed to becoming a major global player (Sornette et al., 2015.) Between 2000 and 2008 the average yearly growth was 13% (Sornette et al., 2015.) But in recent years growth has slowed down. Reported growth for 2015 was 6.8%.
With the growth of the economy, the financial system matured and the Chinese stock markets grew. This growth was not without highs and lows. The Chinese stock markets were affected by the 2008 global financial crisis and, more recently in 2015, the Chinese stock markets experienced another big crisis. In 2015 the stock markets reached record heights but came crashing down after.
The two indices composing the Chinese stock markets are the Shanghai composite index and the Shenzhen composite index. The combined value of the two indices grew from 1583 points in December 2004 to 5458 points in December 2015. An average yearly growth of 12.62%. The highest point of the combined stock market indices was 8307 points, reached June 2015. After this high the markets fell to 5391 points on July 8th. This peak gives reason to suspect a bubble. An empirical study by Sornette et al. (2015) shows that the Chinese markets were in a bubble. It is difficult however to determine the reasons for this bubble. With empirical tests and literature research this paper tends to do so.
So what exactly causes stock prices to move in the direction of a bubble. Volume of trading and interest rates are important elements that could determine the stock market prices. Granger causality tests (Granger, 1969) determine these relations. Political policy also influences stock market prices. The policy of the Chinese government has contributed to the rising stock markets. The People’s Bank of China (PBoC) made interest rate cuts while the market was growing in 2015. The period before the bubble, 2004 till 2015, trading volume and the number of investors grew. Trading volume has a positive relation with stock market returns, as is latter shown in this paper. According to present value pricing a decrease in interest rates also causes stock market prices to go up.
The rest of this thesis is structured as follows: Part 2 provides a literature review about bubbles and the relation between volume of trading and stock prices. Part 3 provides results of empirical tests. Part 4 reviews theoretical reasons for causality between volume of trading and the stock markets, and for causality between interest rates and the stock markets. Part 5 gives political reasons for the bubble to emerge. Part 6 summarizes and concludes.
5
2. Literature review
A stock market bubble is present when stock market prices are far above fundamental values (Rappoport & White, 1993)(Gürkaynak, 2008.) Mishkin (2013) defines two kinds of bubbles, a credit-driven bubble and bubble credit-driven by irrational exuberance. An event like an interest rate cut might cause the start of a credit boom. Mishkin describes how this can cause an asset price bubble: ‘easier credit can be used to purchase particular assets and thereby raise their prices. The rise in asset values, in turn, encourages further lending for these assets, either because it increases the value of collateral, making it easier to borrow, or because it raises the value of capital at financial institutions, which gives them more capacity to lend.’ This feedback loop can create a bubble, driving asset prices well above fundamental value. (Mishkin 2013; Shiller, 2000.) The second kind of bubble is caused by irrational exuberance; Mishkin: ‘High expectations about the future value raises stock prices above their fundamental value. Eventually when it becomes clear that the high expectations were wrong, the stock prices will go down fast.’ The first edition of the book Irrational Exuberance published in 2000 by Shiller appeared to be forward looking as soon after the release of the first edition the US stock market crashed. Irrational exuberance and high expectations were present in this dot-com bubble on the US markets prior to the crash.
A third kind of bubble can be found in the field of behavioral finance, a speculative bubble. When speculative prices go up, creating successes for some investors, this may attract public attention,
promote word-of-mouth enthusiasm, and heighten expectations for further price increases. This same feedback can also cause a negative bubble or the burst of the bubble. Traders have the tendency to attribute good results to their own ability and bad results to bad luck. This biased-self attribution can also aid in a speculative bubble. (Shiller, 2003.)
According to the efficient market hypothesis rational traders stabilize market value arounds its fundamental value (Fama, 1970.) Rational traders buy a stock when irrational pessimists sell, and they sell a stock when irrational optimist buy. However this ‘smart money’ doesn’t necessarily offset the impact of irrational investors (Shiller, 2003.) In fact rational traders can amplify the effects of irrational traders and be destabilizing (Long, Shleifer, & Summers, 1990.) Smart money can stop investing in a particular asset but cannot take action to stop others from bidding prices further up. (Shiller, 2003): ‘When rational speculators receive good news and trade on this news, they recognize that the initial price increase will stimulate buying by positive feedback traders tomorrow. In anticipation of these purchases, informed rational speculators buy more today, and so drive prices up today higher than fundamental
6
news warrants. Tomorrow, positive feedback traders buy in response to today's price increase and so keep prices above fundamentals even as rational speculators are selling out and stabilizing prices.’ The market model Long et al develop confirms that the strategy of rational traders can be destabilizing. If Investors expectations are rational they see through the impact of each shock on long-run equilibrium prices for stocks and commodities. This rationality of traders can be questioned. From financial history we have learned that real world expectations react different to shocks then would be expected with rational expectations. Sometimes expectations change slowly sometimes more rapidly (Kindleberger & Aliber, 2015.)
2.1 Literature on price volume relation.
I distinguish four main reasons for a causal relation between the volume of trading and stock prices. First and second the difference in availability and timing of relevant information to traders can cause the relation between trading volume and stock prices. Described as the mixture of distribution hypothesis (MDH) and asymmetric information hypothesis (AIH) (Liu, Liu, & Liang, 2015.) Third the way informed agents trade affects the relation and fourth ‘noise traders’ feedback strategy stimulates the relation.
The first reason for the importance of the availability of information to the price volume relation is supported with artificial market models research. In the real market it is impossible, or at least very difficult, to measure the difference between available information among traders. (Liu et al, 2015) For economists “sequential or simultaneous" information arrival is in part a semantic issue. Karpoff (1987): ‘Empirical research indicates that price adjustment to new information is "very quick," but "very quick" can be interpreted as nearly instantaneous or as supporting gradual information dissemination.’ There is no general consensus about sequential or simultaneous information arrival.
Agent-based artificial stock markets enjoy the advantage of producing artificial data that researchers require, such as the information-driven component of trading volume. Artificial data also allow researchers to test price–volume relationship easily (Liu et al., 2015.) Copeland (1976) constructed a market model to test the price volume relation. In the Copeland model information is disseminated to only one trader at a time. This sequential information model predicted a positive correlation between the absolute value of price changes and volume. That implies a positive correlation between Volume and the absolute value of change in price. In the artificial market model constructed by Liu et al., (2015) there are two types of agents. Agents with private information and agents without private information. As in
7
real markets agents with private information are the minority. The model has three conclusions. First information-driven trading produces the price-volume relationship, second the tendency of informed agents to take a higher price to buy and a lower price to sell affects the price-volume relationship and third the tendency of informed agents to trade more at one transaction affects the price volume relationship.
These conclusions can be taken to the real market but are difficult to prove, however are likely to hold. In the real market information also differs among investors. Investors will be more and less informed about important elements influencing the stock markets. Some investors will have private information. This asymmetric information contributes to the price volume relation (AIH.) When there is asymmetric information among traders, more trades will occur because opinions about the value of shares differ. Simulation tests indicate that Volume is highest when investors are all optimists or all pessimists. (Copeland, 1976.) therefore if expectations are more homogenous the volume, price relation will be higher. Generally, in a bubble expectations are optimistic for all traders, until the bubble ‘pops’ then all expectation will become pessimistic. In a bubble a strong price volume relation will be present.
Let’s use an example to clarify why sequential information arrival contributes to the price volume relation. We assume there are two kind of traders; large financial institutional, and private speculators. We also assume sequential information arrival. The large institutions receive positive information about a particular company, which they therefore think is underpriced on the stock markets. The institution trade upon the new knowledge and they start buying the company’s shares. Later the private speculators receive the same information and start trading on that information.
Another possible reason for a relation between volume and stock prices are noise traders. Noise traders do not trade on economic fundamental but based on previous stock price movements. ‘A positive causal relation from stock returns to volume is consistent with the positive-feedback trading strategies of noise traders, for which the decision to trade is conditioned on past stock price movements’ (Hiemstra & Jones, 1994.) When speculative prices go up, creating successes for some investors, this may attract public attention, promote word-of-mouth enthusiasm, and heighten expectations for further price increases. (Shiller 2003.)
8
2.2. Interest rate, stock price relation.
This paper also investigates the causality between the interbank lending rate and the stock market prices in China. The interbank lending rate is the cost of borrowing for Chinese banks. When there cost of borrowing goes up they will charge higher borrowing costs to their customers. When cost of borrowing goes up, profits go down. So a rise in the interest rate will lead to lower stock prices. The present value of future investment determines the value of a firm. Interest rates are an important part of the discount factor. Higher interest rate will result in a lower present value of future investments, this is a second reason for the negative correlation between stock prices and interest rates (Berk and DeMarzo, 2013.)
3. Empirical tests and results
Monthly data spanning from November 2004 till may 2016 from the Shanghai composite index, the Shenzhen composite index and their trading volume were retrieved from DATASTREAM/People Bank of China. Interbank lending rate is obtained for the same period. Interbank lending rate is the most appropriate measure for the cost of borrowing money. Monthly data is used because only monthly data of the interbank lending rates and the volume of trading were available. Granger test are performed in several ways. Hiemstra & Jones (1994), first determine the appropriate lag of the independent variables and then perform F-tests. Mok (1993) performed restricted and unrestricted regressions on the variables of interest with multiple lags of the independent variables. Because this method captures the causality of the most recent past values to the dependent variable I will follow their method.
On all variables the Unit Root Dickey Fuller test (1981) were performed with and without a trend variable. The Dickey-Fuller test without a trend is based on the following equation:
Yt= α + βYt-1 + ut. With a trend the equation is:
Yt= α + βYt-1 + β1T + ut.
Yt is the current value of the tested variable, Yt-1 is the lagged value of this variable. T is the trend variable. Both the DF test with and without a trend is to make sure no false conclusions are made with the results. When the variables are influenced by a trend, the trend can adjust the values and still capture the unit root. A variable has unit root when β = 1 in the test without a trend. With a trend the variable has unit root when β = 1 and β1=0. The null hypothesis is that the variable doesn’t have unit root and the alternative hypothesis is that the equation has unit root. A unit root variable is stationary and
9
has the tendency to return to a constant. A stationary variable can be used for a regression without causing a bias.
For the DF-test without the trend term, the null hypothesis can only be rejected for the
Interbank lending rate and the volume of trading on the Shanghai market. The DF test with a trend term gives completely different results. The null hypothesis can be rejected for all variables except for the Shenzhen composite index. The results of the Dickey-Fuller test are summarized in table 1.
Dickey-fuller test, without drift. Dickey-Fuller test, with drift term
series F-statistic P-value series F-joint test P-value
Chinese market -1.995 0.2889 Chinese market -1.995 0.0241
Shanghai market -2.081 0.2521 Shanghai market -2.081 0.0197
Shenzhen market -1.47 0.5481 Shenzhen market -1.47 0.0719
Interbank lending rate -3.538 0.0071 Interbank lending rate -3.538 0.0003
Volmarket -2.661 0.081 Volmarket -2.661 0.0044
Vol shanghai -3.029 0.0323 Vol shanghai -3.029 0.0015
Vol Shenzhen -2.093 0.2472 Vol Shenzhen -2.093 0.0191
Table 1
Based on both DF-tests I did Granger causality test. This means that for some of the test the differenced variables were created of the variables not rejecting the null hypothesis.
To investigate Granger causalities between two variables two regressions are performed. A restricted and an unrestricted regression. The unrestricted regression does on ordinary least squares (OLS) regression from a dependent variable with its own lagged values and the lagged values of a second independent variable. The restricted regression doesn’t include the lagged values of the second variable. The F-test determines if the unrestricted regression is a significant better predictor then the restricted variable regression. Both Granger causalities from the stock market to volume of trading, and from volume of trading to the stock market are determined. The Granger tests I conducted took the first to lags of two independent variables in the unrestricted regression, and the first two lags of only one variable in the restricted regression. The F-test determines whether the unrestricted regression adds significant explanatory power to the regression. In both the differenced and on-differenced case the volume of trading granger causes the value of the stock market for the entire Chinese market. The Shenzhen market is granger caused by the volume of trading of the Shenzhen index (only differenced
10
Shenzhen index is shown.) Granger cause of volume of trading in the Shanghai market on the Shanghai market index is insignificant at a 5% level, but significant at a 10% level. The differenced shanghai market is significantly Granger cause by the volume of trading in that market at a 5% level. On all markets the Granger causality of the market towards the volume of trading is clearly rejected. We can conclude the volume of trading granger causes the Chinese stock market indices. But bidirectional causality (or feedback) doesn’t exist.
The interbank lending rate has unit root in both the DF test with and without a trend. Differencing the rate is unnecessary. With the same restricted and unrestricted models as before Granger Causality between the Interbank lending rate and the stock markets determined. None of the regression show a Granger causality in any direction between the interbank lending rates and the stock markets. Table 2 till 5 summarize the results
Series Unresricted regression on Restricted regression on F-test P-value
Market Market, volmarket Market 4,9 0,0088
Shanhai market Shanghai market, volShanghai Shanghai market 2,97 0,0547
Volmarket Volmarket, market Volmarket 0,01 0,9915
VolShanghai VolShanghai, Shanghaimarket VolShanghai 0,03 0,9744
VolShenzhen VolShenzhen, Shenzhen market VolShenzhen 0,08 0,927
Table 2, Granger test of market volume on stock prices and of stock prices on market volume with on-differenced variables.
Differenced series
Market Market, volmarket Market 4,83 0,0094
Shanghai market Shanghai market, volShanghai Shanghai market 3,76 0,0258
Shenzhen market Shenzhen market, volShenzhen Shenzhen market 9,81 0,0001
Volmarket Volmarket, market Volmarket 1,47 0,2337
VolShanghai VolShanghai, Shanghaimarket VolShanghai 1,17 0,3148
VolShenzhen VolShenzhen, Shenzhen market VolShenzhen 1,01 0,3683
Table 3, Granger test of market volume on stock prices and of stock prices on market volume with differenced variables.
11
Table 4, Granger test of market prices on interbank lending rate and of interbank lending rate on stock prices with on-differenced variables
Table 5, Bidirectional Granger causality tests of stock market prices and interbank lending rates.
3.1 Interpretation of results.
The conducted tests show a significant Granger causality from trading volume to stock prices on all Chinese indices. Past values of trading volume help explain current stock prices. May 2014 (the
estimated start of the bubble) trading volume was 3.05 trillion shares and went up to 5.07 trillion during June 2015. The rising trading volume accompanied the cumulative phase of the bubble. A big part of the trading volume comes from small retail investors. In 2010 90% of trading accounts belonged to individual investors (Lee et al., 2010.) Tan et al (2008) remark that Chinese investors tend to be optimistic and confident of government support in rising markets. They conclude that under such circumstances investors show herding behavior. Herding behavior is the tendency for an investor to follow the actions of others. Lee et al (2010) mention that Individual investors have a short-term and speculative
investment perspective, and are more susceptible to the influence of psychological biases and attention-grabbing events. Herding and speculative behavior aid to a speculative bubble as is described by Mishkin (2013.) Interbank lending rate and stock prices don’t show significant granger causality in any direction. The frequency of data points likely plays a significant role in these results. Interest rates from one or two
Series F-test P-value
Market Market, Interbank Market 0,37 0,6893
Shanghai market Shanghai market, Interbank lending rate Shanghai market 0,09 0,9147 Shenzhen market Shenzhen market, Interbank lending rate Shenzhen market 1,4 0,2512 Interbank lending rate Interbank lending rate, market Interbank lending rate 0,33 0,7213
Interbank lending rate
Interbank lending rate, market
(differenced) Interbank lending rate 0,29 0,7523
Differenced series
Market Market, Interbank lending rate Market 0,35 0,7061
Shanghai market Shanghai market. Interbank lending rate Shanghai market 0,08 0,9193 Shenzhen market Shenzhen market, Interbank lending rate Shenzhen market 1,26 0,2865
12
months ago likely don’t affect the current stock markets and vice versa. But from economic theory it seems clear the on the short term interest rate changes will have an effect on the stock markets.
4. The Chinese bubble
In this section the case of the Chines stock market bubble will be discussed. The Chinese stock market, consisting of the Shenzhen composite index and the Shanghai composite index experienced a bubble in the period from mid-2014 till the end of 2015. The Financial Crisis Observatory of the university of Zurich observes and investigates financial markets across the world. They search for instabilities and bubbles in these market with the use of econometric modeling. Sornette et al. (2015) of this observatory found that the Chinese stock markets were in a bubble. The start is estimated at mid-2014. The bubble lasted till June 2015 when the market started to decrease. June 2014 the index value of the market was 3100 points. From there the market grew over 160% in one year. The highest point was reached June 6th 2015; 8307 points. After this record high the indices fell to 5391 points on July 8 of 2015. A decrease of 35% in just over a month. Since then the market has been going up and down. The value in June 2016 is 4807. Far from the high values of 2015. For a graphical representation of the rise and fall see graph 1.
Graph 1, the Chinese markets combined stock index.
The Chinese stock market bubble has elements of a credit driven bubble, a bubble cause by irrational exuberance and a speculative bubble. Credit became more liquid in the period before the
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
13
bubble, also speculation and overconfidence were present in the Chinese markets (Burdekin &
Weidenmier, 2015, Wang et al, 2015.) Kindleberger (1978) distinguished three stages of a bubble; mania, distress and panic. During the mania stage stock prices keep going up, and during the distress and panic stage the stock market can crash quickly. Sornette et al (2015) suggest that the mania stage of the bubble, in the 2015 Chinese stock markets, might in part be caused by interest rate cuts by the People Bank of China (PBoC). Lower interest rates consequently lead to easier credit which could be the root of a credit driven bubble.
4.1. Volume of trading relation to stock prices
Theoretical models show that information is crucial in for the relation between volume of trading and price changes. This is summarized in the mixture of distribution hypothesis (MDH) and asymmetric information hypothesis (AIH) (Liu, Liu, & Liang, 2015.) In the Chinese markets it is likely that information is not available at the same time for all investors (some information might not be available at all to some traders.) This leads to a price volume relation according to the mixture of distribution hypothesis (Liu et al, 2015.) Information also differs among investors. Investors will be more and less informed about important elements influencing the stock markets. This asymmetric information contributes to the price volume relation (AIH.) When there is asymmetric information among traders, more trades will occur because opinions about the value of shares differ. Simulation tests indicate that Volume is highest when investors are all optimists or all pessimists. In China there is a big difference in availability of data. The largest part of investors are retail investors (Lee et al., 2010.) The other part can be described as institutional investors. Generally institutional investors will be more informed then retail investors because of the resources they possess.
Before the bubble and during the mania phase of the bubble, the number of trading accounts and the volume of trading rose across the Chinese stock markets. December 2004 the combined trading volume was around 360 billion shares, and grew towards about 11 trillion shares in December 2015. An average growth of 36.70% per year. Graph 2 shown the volume of trading from 2004 till 2015. Before the estimated start of the bubble (May 2014) there were a total of 180 million accounts registered on the stock exchanges. May 2015 there were 240 million registered accounts, which are mostly retail investors. Both during the mania phase and the distress/panic phase trading volume shows peaks. This suits with herding behavior and feedback theory (Shilller 2003.) Retail investors are more receptive to herding behavior and feedback. They make decision based on what others did. High trading volume and a big
14
percentage of retail investors make a speculative bubble more likely. The granger tests result also show causality from volume of trading to stock prices.
More reasons for the relation between trading volume and stock prices are found in Chinese politics. In recent years the Chinese government has encouraged its population to start investing in the stock market, which raised volume of trading at a time when stock prices were also rising. It became political practice in China to reward China’s elite for their investments in the stock market. Thereby encouraging investment in the stock market.
Graph 2, Trading volume on the Chinese markets.
4.1 Political catalyzation.
China’s economy has been growing in the recent past. But the Chinese economy has experienced difficulties after the 2008 global financial crisis and with the recent contraction of the world wide
economy. The Chinese government aims for high economic growth. Booming stock markets help to accomplish that goal (Li, Hsu, & Qin, 2014.) Economic growth and the growth of the stock markets are stimulated in a number of ways. During the global financial crisis in 2008 the Chinese government announced a 4 trillion renminbi stimulus package (Diao, Zhang, & Chen, 2012) furthermore lending requirements were reduced and interest rate cut. Incentivizing political programs are in place stimulating investments in the economy and stocks by politicians. This section explains how these Chinese political measures aided to the 2015 stock market bubble. As Blanchard 1981 puts it: ‘The stock market is not the
0 5000 10000 15000 20000 25000 1- 11-200 4 1- 6-2005 1-20 06 1- 8-2006 1- 3-2007 1- 10-200 7 1- 5-2008 1- 12-200 8 1- 7-2009 1- 2-20 10 1- 9-2010 1- 4-2011 1- 11-201 1 1- 6-2012 1-2013 1- 8-2013 1- 3-2014 1- 10-201 4 1- 5-2015 1- 12-201 5
15
"cause" of the increase in output, no more than the increase in output is the cause of the initial stock market change. They are both the results of changes in policy.’
Before the 2008 crisis, the People's Bank of China including, as a final measure, enforced purchases of central bank bills by banks deemed guilty of excess lending. During the crisis these constraints on lending were lifted and lending soared (Burdekin & Weidenmier, 2015.) Not only credit constraints were lifted, interest rates were also continuously cut. At the start of 2008 interest rates were 7.47%. September 2008 interest rate was cut to 7.2% and eventually cut to 5.31% in 2009. The cuts in interest rates also lead to lower interbank lending rates which were close to 3% in 2008 and close to 1% in 2009. Without credit constraints and a lower lending rate credit to finance investments in the stock markets is easier and cheaper. This should have positive effects on the economy and stock markets (as was the goal after the 2008 crisis), but in the long run could lead to overinvesting with much leverage, aiding to a credit-driven bubble.
The 4 trillion renminbi stimulus program announced by the Chinese government in November 2008 (equivalent to $586 billion at the time) represented around 15% of the country's GDP(Burdekin & Weidenmier, 2015). Implementation of the Chinese stimulus package was further aided by soaring fixed asset investment by the central government's State-Owned Enterprises (SOE’s). Investment of SOE’s was already growing. Before the stimulus program the annual growth rate of investments was 11.59%. The second quarter of 2009 investments of SOE’s grew at a 45.3% annual rate (Wen&Wu, 2014.) Wen and Wu also notice that these investments were fueled by increased borrowing such that, even as private manufacturing firms lowered their leverage ratios in the face of the crisis, the SOE’s raised their average leverage ratio from 57.5% in the first quarter of 2008 to a peak of 61.4% in the second quarter of 2009. High investments by SOE’s stimulates the economy and the stock markets. SOE’s listed on the stock markets will have high values partly because of high leverage ratios. This high leverage also means that investing in these companies involves more risk. A realization of this risk might have started the distress phase of the 2015 stock market bubble.
The Chinese political incentives programs also contributed to the bubble. Both the high degree of political centralization and fiscal decentralization in China make political promotion an important
incentive for yardstick competition among local government officials (Jin et al., 2005; Wang et al., 2015.) Regularly the performance of Chinese politicians is evaluated. Anyone who has been ranked towards the top in these evaluations or has been awarded the title of “advanced leader,” is more likely to be
promoted as a “political bonus” (Xu, 2011). Higher promotion incentive is associated with more credit of medium and long term investment. Growth is stimulated by debt-financed programs which are mainly
16
financed via bank loans. The stronger the performance incentive the riskier the respective local bank exposure becomes. (Wang et al., 2015.)
Wang looks at the relation between political pressure and risk taking. The main findings are that local banks were forced to take on riskier loans, they were not able to avoid them, and after the 2008 financial crises the relation between political pressure and risk taking has become stronger. Almost 80% of government debts come from bank loans, indicating that the usage of bank money and its financing models will inevitably impact bank risk. Frequently politicians are promoted or placed at another position. This turnover of government officials encourages short term optimism. Wang et al: ‘If politicians can expect the current debt burden to be transferred to their successors after a successful political promotion, they will tend to be more careless about investing and debt-financing’. This leads to inefficient investment and increasing bank risk (Wang et al., 2015.)
The stock markets are an indicator for economic growth. High stock markets contribute to economic growth and consumer confidence. Thus the Chinese government aimed to keep economic growth high and succeeded (at least partly.) A general equilibrium model (Diao et al., 2012) estimates that growth could have dropped to 3.8% in 2009, after the 2008 crisis, instead growth was 7.8%. The taken measures however contributed to the 2015 stock market bubble. Lower interest rates and less credit constrained combined with a big fiscal stimulus packages and Chinese politicians engaging in risky investment all aided to the growth of the stock markets.
4.2. Shadow banking system
The shadow banking system plaid an important role in the rise and fall of the 2015 stock market. This banking system, less regulated by the government than the regular banking system, provides many investors with loans. There are different definitions for the Shadow banking system. A common
classification is to define shadow banking as lending activities outside the regular banking system. Using this hypothesis has two disadvantages, practices that are generally not considered shadow banking such as leasing are included, and it describes shadow banking as primarily acting outside of banks, but many shadow banking activities operate within banks. Alternatively Claessens and Ratnovski define shadow banking as “all financial activities, except traditional banking, which require a private or public backstop to operate”. Shadow banking firms are similar to regular banks, but because they are outside of the regular banking system shadow banks lack a safety net. For instance lender of last resort facilities of central banks are nog available (Elliott & Qiao, 2015.) A backstop is crucial for shadow banking firms
17
because unlike with regular banking activities residual risks and systemic risk cannot be diversified away (Claessens & Ratnovski, 2014.)
China's shadow financial system is comprised of non-bank financial products. The shadow financial system is dominated by commercial banks, insurance companies, and trusts. For smaller Chinese businesses and retail investors it is difficult to get finance form the regular banking system. Because these banks are restricted and heavily regulated by the government. The shadow banking system can provide this smaller bussinesses and retail investors with credit who cannot obtain regular banking credit (Lasak, 2015.) This provision credit helped the growth of the chinese economy and assisted in the move towards a more market based-economy (Elliott & Qiao, 2015; Lasak, 2015.)
The Chinese shadow banking system has grown rapidly in recent years (Claessens & Ratnovski, 2014; Li, Hsu, & Qin, 2014.) After the 2008 global financial crisis the governemnt did nog restrain shadow banking. The crissis actually led to a more developed credit intermediation behind the traditional
banking system. Furthermore the stimulus package (mentioned earlier) contributed to the growth of the shadow banking system (Lasak, 2015.) It is difficult to impossible to separate out banks' shadow banking transactions from regular transaction. Estimations are that currently about 60% of loans in China are issued via the shadow banking system (Li et al., 2014.)
Retail investors are the main buyers of non-traditional shadow banking loans. The rise in shadow banking activities coincided with the rise in the number of trading accounts and traders. Which were mostly retail investors. Shadow banking helped to create more margin for this retail traders so they could get higher returns. Data from the China Security Finance corporation show that margin purchases grew constantly from 2012 till January 2015. January 2012 the value of margin purchases was 27 billion renminbi for January 2015 margin purchases were 210 billion renminbi. China's Securities Regulatory Commission (CSRC), restricted margin trading. January 2015 three brokerage firms that were trading with too much leverage, they highest allowed leverage ratio was 2, were forbidden to open new margin accounts for three months. Investors, searching for more margin, found other ways to keep high
leverage ratios. They turned to the shadow banking system, which provided loans with leverage ratios up to 3.(Qian, 2016) This high margins contributed to growth of the stock markets, Chinese investors were able to invest more with the margin, but also to the crash that started on June 12 2015.
June 12 (CSRC), commission seeking to contain risks in the country's highly-leveraged stock market, announced draft rules that cap a brokerage's margin trading and short selling business and shadow banking activities (Reuters, 2015.) This announcement started the crash of the Chinese markets.
18
Investors will have realized the risk they were facing with shadow banks and margin trading. Retail investors subjective to herding behavior started opting out the market. Local governments could also expect problems, they had so much debt that they were left with two options: either repaying old debts by raising new ones or not paying off debts that are overdue (Wang et al., 2015.)
After the fall in stock market prices Chinese investors with the highest leverage ratios were forced to sell. Their forced selling triggered the liquidation of the lower leverage portfolio, eventually even the two timed leverage brokerage portfolio holders were forced to sell. This margin call explains the big crash of the Chinese markets from 8307 points on June 12 towards 5391 on July 8. At first the Chinese
government didn’t react strongly to the fall of the stock markets. The government might have seen the fall as a correction to the prior bull markets (Qian, 2016.) However on June 27 the PBoC cut interest rates again from 5.1% to 4.85% and thereafter in two steps to 4.35% on October 31. With the fall many
companies suspended trading of their shares. July 5th the CSRC announced they would stop planned IPO’s and punish ‘malicious short selling’ activities (Qian, 2016.) All this measures couldn’t undo the big crash the Chinese stock market had encountered but the markets stabilized at the end of 2015.
5. Conclusions
Proficient evident is provided by Sornette et al (2015) to determine that the 2015 Chinese stock markets were in a bubble. The Chinese markets had experienced an almost constant growth over the past 10 years and the growth accelerated from mid-2014 onwards. In one year the Chinese stock markets grew over 160%. In June 2015 the market crashed and during 2016 stabilized around the mid-2014 value.
With Granger causality test correlation between variables influencing the stock prices were determined. Causality between the interbank lending and stock prices could not be determined, likely caused by the limited availability of data. The Granger tests did show strong evidence for a positive correlation between trading volume and stock prices. Past values of trading volume granger cause stock prices. The difference in availability and timing of relevant information to traders can cause this causality (Liu et al, 2015.) Speculation and herding are other reasons for trading volume to influence stock prices (Shiller 2003.)
19
The Chinese political policy also played a role in the formation of the bubble. A booming stock markets would aid in economic growth (Qian, 2016), the bull markets were initially supported by the Chinese government. Furthermore, led the Chinese political promotion system to risky investment by politicians. Lending requirements were lifted after the 2008 financial crisis making credit more liquid which supports a credit driven bubble (Kindleberger & Aliber, 2015.) Interest rates were cut several times by the People bank of China during the accumulating of the bubble.
The shadow banking system played an important role in both the accumulation and the crash of the bubble. Retail investors, unable to obtain credit from regular banks, could invest on margin with credit from the shadow banking system. This led to more trading volume and a rise of the stock markets. However, the fact that much of the trading was on margin also made these investments riskier. When on June 12 the CSRC announced it would take measures to halt margin trading a shadow banking activities the stock market started to crash. Although the government made numerous efforts to stop the downfall they didn’t succeed. The same speculative and herding behavior that accompanied the soaring market also were present during the crash of the market. This lasted till the end of 2015 when the markets started to stabilize.
Further research can be done for other time periods in the Chinese market and with more frequent (daily) data. Also other reasons for the formation of a bubble can be investigated. False conclusion can be drawn when relevant variables aren’t added in a regression (Eichler 2011.) The more complete the formulation of reasons for a bubble the less power this criticism has. For example, the precise effects of the shadow banking system on the Chinese economy and stock markets can be examined.
20
6. References
Berk, J., & DeMarzo, P. (2013). Corporate Finance, Prentice Hall
Burdekin, R. C. K., & Weidenmier, M. D. (2015). Assessing the impact of the Chinese stimulus package at home and abroad: A damp squib? China Economic Review, 33, 137–162.
Blanchard, O.T. (1981). Output, the Stock Market, and Interest Rates, The American Economic Review, 71(1), 132-143
Claessens, S., & Ratnovski, L. (2014). What is Shadow Banking? IMF Working Papers, 14(1), 1-8 Copeland, T. E. (1976). A Model of Asset Trading Under the Assumption of Sequential Information
Arrival. American Finance Association, 50(1), 319–339.
Diao, X., Zhang, Y., & Chen, K. Z. (2012). The global recession and China’s stimulus package: A general equilibrium assessment of country level impacts. China Economic Review, 23(1), 1–17.
Dickey, D A and Fuller, W A, (1981), Likelihood ratio statistics for autoregressive time series with a unit root, Econometrica, Vol 49, 4, 1057-1072.
Elliott, D. J., & Qiao, Y. (2015). Reforming Shadow Banking in China, The Brookings Institution (May) Fama, E. F. (1970). Efficient Capital Markets: A review of Theory and Empirical Work. The Journal of
Finance, 25(2), 383–417.
Granger, C. W. J. (1969). Investigating Causal Relations by Econometric Models and Cross-spectral. Econometrica, 37(3), 424–438.
Granger, C. W. J. (1980). Testing for causality. A personal viewpoint. Journal of Economic Dynamics and Control, 2(C), 329–352.
Gürkaynak, R. S. (2008). Econometric tests of asset price bubbles: Taking stock. Journal of Economic Surveys, 22(1), 166–186.
Hiemstra, C., & Jones, J. D. (1994). Testing for Linear and Nonlinear Granger Causality in the Stock Price‐ Volume Relation. Journal of Finance, 49(5), 1639–1664.
Jin, H., Qian, Y., Weingast, B.R. (2005) Regional decentralization and fiscal incentives: Federalism, Chinese style. Journal of Public Economics 89, 1719–1742
Karpoff, J. M. (1987). The Relation Between Price Changes and Trading Volume: A Survey. Journal of Financial and Quantitative Analysis, 22(1), 109–126.
Kindleberger, C. P., & Aliber, R. Z. (2015). Manias, panics and crashes: a history of financial crises. Lasak, P. (2015). Regulatory responses to the Chinese shadow banking development. Jagiellonian Journal
21
Lee, B. S., Li, W., & Wang, S. S. (2010). The dynamics of individual and institutional trading on the Shanghai Stock Exchange. Pacific Basin Finance Journal, 18(1), 116–137.
Li, J., Hsu, S., & Qin, Y. (2014). Shadow banking in China: Institutional risks. China Economic Review, 31, 119–129.
Liu, X., Liu, X., & Liang, X. (2015). Information-driven trade and price-volume relationship in artificial stock markets. Physica A: Statistical Mechanics and Its Applications, 430, 73–80.
Long, J. B. D. E., Shleifer, A., & Summers, L. H. (1990). Positive Feedback Investment Strategies and Destabilizing Rational Speculation. The Journal of Finance, 45(2), 379–395.
Mishkin F., K. Matthews and M. Giuliodori (2013). The Economics of Money, Banking, and Financial Markets (European Edition)
Mok, H. M. (1993). Causality of interest rate, exchange rate and stock prices at stock market open and close in Hong Kong. Asia Pacific Journal of Management, 10(2), 123–143.
Qian, J. (2016). The 2015 Stock Panic of China: A Narrative 1, Antai College of Economics and Management, Shanghai Jiao Tong University.
Rappoport, P., & White, E. N. (1993). Was There a Bubble in the 1929 Stock Market? The Journal of Economic History, 53(3), 549–574
Reuters Internet article, (2015), China market regulator seeks to cap margin trading, short selling, www.reuters.com/article/china-margin-financing-idUSL3N0YY3H620150612
Shiller, R. J. (1980). The use of volatility measures in assessing market efficiency, The journal of Finance, 36(2), 291–304.
Shiller, R. J. (2000). Measuring Bubble Expectations and Investor Confidence. The Journal of Psychology and Financial Markets, 1(1), 49–60.
Shiller, R. J. (2000). Irrational Exuberance, Princeton University Press.
Shiller, R. J. (2003). From Efficient Markets Theory to Behavioral Finance. Journal of Economic Perspectives, 17(1), 83–104.
Sornette, D., Demos, G., Qun, Z., Cauwels, P., Filimonov, V., & Zhang, Q. (2015). Real-Time Prediction and Post-Mortem Analysis of the Shanghai 2015 Stock Market Bubble and Crash. SSRN Electronic Journal, (October), 15–32.
Tan, T., Chiang, T.C., Mason, J.R., Nelling, E. (2008) Herding behavior in Chinese stock markets: An examination of A and B shares. Pacific-Basin Finance Journal 16 (2008) 61–77
Wang, L., Menkhoff, L., & Schröder, M. (2015). Politicians ’ Promotion Incentives and Bank Risk Exposure in China, Discussion Paper, 15, 1-52
Wen, Y., & Wu, J. (2014). Understanding great recession like China. Federal Reserve Bank of St. Louis Working Paper Series
