Does China’s Financial System Amplify Risks in the Real Economy?

2.1 Risk Absorption Mechanism of the Financial System for the Real Economy

To be specific, the financial system can mitigate risks in the real economy through risk management, resource allocation and risk sharing.

First, the financial system has the function of diversifying, transferring and managing risks (Merton, 1995). This function is formed by three aspects: one, the financial system manages liquidity risks by smoothing the supply and demand of intertemporal funds (Allen and Santomero, 1998); two, the financial system can share risks by transferring risks between investors (Merton 1989); three, by acquiring and assessing the information of investors, the financial system cuts supervision costs, mitigates moral hazards, ensures the safety of savers’ funds and reduces real economic risks caused by losses in consumer property and fluctuations in consumer demand (Levine, 2004).

Second, the financial system promotes economic development through effective resource allocation based on financing functions, and thus indirectly mitigates real economic risks. The financial system has the function of resource allocation (Merton, 1995), and this function is used for allocating resources among different economic entities, different regions and different periods with high efficiency, which improves the efficiency of fund use and facilitates product sales, consumption scale and steady economic development (Zhang and Zhu, 2014). From this perspective, the financial system directly promotes economic growth and indirectly mitigates risks in the real economy through the resource allocation function.

Third, in normal periods, with small external impacts, the internal correlation within the financial system can diversify the risks and thus mitigate the risk spillover from the real economy. In theory, correlation among financial institutions includes direct creditor-debtor relationship and indirect holding of common assets. Highly correlated financial network can ensure that financial institutions complement each other’s liquidity to mitigate the pressure of insufficient liquidity and mitigate risks in the financial system. Moreover, in normal periods, impacts of the real economy are small and follow a “thin-tail” distribution. Thus, financial institutions can share risks with each other through inter-bank lending and repo market, so that impacts of the real economy on the financial system are mitigated (Elliott et al., 2014; Acemoglu et al., 2015; Cabrales et al., 2017). Therefore, when external impacts are small, the highly correlated network among financial systems can effectively absorb and resolve risks in the real economy.

2.2 Risk Amplification Mechanism of the Financial System for the Real Economy

To be specific, risks in financial system can spill over to the real economy through business linkage, inter-institutional infection, negative externalities and financial accelerators.

First, risks in the financial system directly spill over to the real economy through business linkage. In this paper, banking industry is taken as an example. External impacts trigger withdrawal requests from a large number of depositors, resulting in a sharp drop of banks’ lending funds, and thus corporate financing restraints will be tightened, investment will be hindered and the development of real economy will slow down. In addition, with a negative impact, the amplification effect of high leverage ratio will cause heavy losses to financial institution, and they are forced to undersell assets. Sell-off of assets further drags down the prices of assets, leading to sharp drop in the financing capacity of related entities and aggravating risks in the real economy. Therefore, risks in the financial system spill over to the real economy through lending business.

Second, risk spillover from the financial system to the real economy is amplified by internal correlation. The internal correlation of the financial system includes two types. On the one hand, financial institutions are directly correlated through borrowing and lending in the wholesale capital market (Shin and Shin, 2011). On the other hand, financial institutions often choose to hold similar asset portfolios, which lead to an increase in indirect correlation of holding common assets. In summary, due to excessive correlation, negative impacts spread within the financial system, causing heavier damage and larger influencing scope and amplifying the possibility and magnitude of infection.

Third, risks in the financial system will have serious externalities to the real economy. Once the financial system is negatively impacted, a large number of financial institutions will collapse, and the basic function of the financial system—capital allocation—will be affected. The failure of financial institutions in providing the real economy with sufficient fund will cause huge damage to the real economy (Laeven and Valencia, 2018).

Fourth, the impact of the financial system on the real economy is amplified through financial accelerator mechanism. Financial accelerator model is proposed by Bernanke et al. (1999), demonstrating the risk spillover from the financial system to the real economy from the perspective of the interaction between these two sides. This mechanism comprises credit demand and credit supply. From the perspective of credit demand, faced with negative impact, the real economy will witness large-scale asset depreciation, resulting in higher probability of corporate loans default. High bad debt ratio forces the financial system to curb credit scale, which poses larger difficulties to the operation of real economic entities. From the perspective of credit supply, faced with negative impact, the profitability of the financial system will decline. Depositors will run on the financial system to make sure that their funds are safe, and the asset scale of the financial system is thus compressed. As a result, the financial system will increase the financing cost of the real economy, and thus the investment and output will drop in the real economy.

Figure 1 shows the transmission mechanism of risk absorption and risk amplification effects of the financial system on the real economy.

Figure 1

The Transmission Mechanism of Risk Absorption and Risk Amplification Effects of the Financial System on the Real Economy

3.1 LASSO-VAR Model

This paper constructs VAR model to conduct quantitative study on the correlation between China’s financial system and real economy. To be specific, the N-ary VAR model of order P can be described as:

In this formula, Y _t is composed of three financial sub-sectors—banking, diversified financing and insurance—and 21 sectors in the real economy, namely 24 variables in total. v refers to 24×1 dimensional intercept vector; Φ_i refers to 24×24 dimensional coefficient matrix; p refers to lag order; and ε_t refers to 24×24 dimensional independent and identically distributed disturbance.

LASSO is a least squares method that imposes a penalty term on the regression parameters, which can estimate the parameters and select variables at the same time (Messner and Pinson, 2019). According to Nicholson et al. (2017), the estimated expression of the LASSO-VAR model is as follows:

Among them, || A ||_F is the Frobenius norm of Matrix A, which is the sum of the squared absolute value of the elements of Matrix A. λ_i refers to the penalty parameter, which is estimated by sequential cross-validation. || Φ_i || ₁ refers to penalty term l₁, which is the sum of the absolute values of all the elements in the vector. By assigning the least significant element to 0, it can reduce the number of intercept coefficients.

3.2 The Construction of Spillover Index

Following the idea in Diebold and Yilmaz (2012), this paper uses generalized variance decomposition to construct the spillover index. In the generalized variance decomposition of the advance H-step prediction, the contribution degree of the j-th variable to the i-th variable is expressed as follows:

Among them, Σ refers to the variance matrix of error ε; σ_{j, j} refers to the standard deviation of the error in the j-th equation; the value of the i-th term in vector e_i is 1, and the values of other terms are 0. As the contribution of all endogenous variables in generalized variance decomposition does not add up to 1, or in other words,

∑j=1Nθi,j(H)≠1, each row of data needs to be standardized:

Therefore, ∑j=1Nθ˜i,j(H)=1 and ∑i,j=1Nθ˜i,j(H)=N.θ˜i,j(H) refers to the j-th endogenous variable is the impact, and the i-th endogenous variable predicts the variance decomposition of period H. Based on θ˜i,j(H), we can construct the spillover index between the financial system and the real economy.

To facilitate this construction, this paper takes the data of sectors in the real economy as the first N-3 endogenous variables, the banking industry as the endogenous variable N-2, and the insurance industry as the endogenous variable N.

3.3 The Construction of Sectoral Risk Variables

To distinguish the specific risks of different sectors, this paper uses the rolling regression method in the calculation process to remove market returns and common influencing factors. The construction of sectoral risk variables is divided into three steps.

The first step is to calculate the profitability of each sector. The formula for calculating the logarithmic return rate of sector i is as follows:

In this formula, P_i,t and P_i,t-1 refer to the closing price of sector i in periods t and t-1.

The second step is to remove the market returns rate by using rolling regression, and to take the residual as the returns rate of each sector.

In this formula, r_i,t refers to the returns rate of sector i, and r_m,t refers to the returns rate of Shanghai Composite Index. Residual ε_i,t is taken in this paper as the returns rate of each sector after removing market returns rate, and is used to calculate sectoral risks.

The third step is to use the univariate GARCH (1,1) model to calculate risk variables of each sector.

Based on Fang et al. (2019), this paper adopts a general autoregressive conditional heteroscedasticity (GARCH) volatility model, and uses dynamic volatility rate to characterize risk variables of each sector. From this, it can be seen that sectoral risks mentioned in this paper refer to specific volatility rates of each sector, rather than the risks with common source in the real economy.

3.4 Data Description

Based on the availability of data, this paper selects the Wind secondary industries classification index to examine the risk spillover effect of the financial system and the real economy on each other, and conducts robustness analysis on the basis of financial data. The advantage of secondary industries classification index lies in its consideration on the heterogeneity between the sub-sectors. On the one hand, this paper uses secondary industries index, rather than primary industries index, because the finer the industries classification, the richer the information contained. Moreover, on March 1, 2007, Wind adjusted its primary industries classification index, separating the real estate sector from the financial industry. Therefore, using primary industries index will result in incomparability between data before and after March 2007. On the other hand, this paper uses secondary industries index rather than tertiary industries index, because it needs to strike a balance between information and data quantity. To be specific, Wind secondary industries classification index includes 24 sectors, three of which are sectors of the financial industry, namely banking, diversified financing and insurance, and 21 sectors in the real economy. Finally, this paper takes Shanghai Composite Index as market returns rate. The sample interval selected in this paper is from January 1, 2000 to August 31, 2019, and the data frequency is weekly. Data sources include Wind database, CSMAR database and the National Bureau of Statistics.

4.1 Static Risk Spillover Effect of the Financial System and the Real Economy on Each Other

Table 1 demonstrates the mutual risk spillover between the financial system and the real economy. First, the financial system mainly absorbs risks in the real economy. In particular, the risk spillover effect of the financial system on the real economy is 42.17%, smaller than the spillover effect of the real economy on the financial system, which is 55.79%. In the overall sample interval, the real economy was the major source of risks, and the financial system absorbs risks in the real economy through resource allocation, risk management and inter-institutional risk sharing.

Table 1

Mutual Risk Spillover between the Financial System and the Real Economy Unit: %

1. Note: “→” means the risk spillover from the former sector to the latter sector; and “←” means the risk spillover from the latter sector to the former sector. The shady part shows the risk spillover effect within the financial system.

In addition, it can be seen from the spillover between financial sub-sectors and the real economy that the banking sector has the closest relationship with the real economy, but its risk absorption capacity is the poorest. Whether for the spillover from the real economy or to the real economy, the banking sector has the closest relationship with the real economy. This matches with China’s financing structure: the main source of financing for China’s real economy is the banking sector (Han and Han, 2017). The risk absorption capacity of the diversified financing sector is the strongest, that of the insurance sector comes the second, and that of the banking sector is the poorest. The reason may be that with the closest relationship with the real economy, the banking sector suffers the most from the risk spillover from the real economy. Higher risk spillover undermines the risk absorption capacity of the banking sector.

Finally, within the financial industry, diversified financing sector has the strongest risk spillover, while that of the banking sector is the weakest; and the correlation between the banking sector and the insurance sector is the closest. In particular, the shady part in Table 1 demonstrates the risk spillover among all the sub-sectors within the financial system. One, the risk spillover from diversified financing sector to the banking sector and the insurance sector is 2.65% and 3.26% respectively, higher than the spillover from the banking sector to the diversified financing sector (2.09%) and the spillover from the insurance sector to the diversified financing sector (2.87%). This indicates that the diversified financing sector has stronger risk spillover to the banking sector and the insurance sector than the reverse direction of spillover. Two, the risk spillover from the insurance sector to the banking sector is 3.73%, slightly higher than the risk spillover from the banking sector to the insurance sector (3.58%). This indicates that compared with the risk spillover from the banking sector to the insurance sector, that of the reverse direction is slightly stronger. Three, if the mutual risk spillover of these three sectors are compared, it can be seen that the risk spillover between the banking sector and the insurance sector is stronger than that between the diversified financing sector and the other two sectors respectively. This conclusion indicates that compared with the diversified financing sector, the banking sector and the insurance sector have closer relationship.

4.2 Periodical Features of the Overall Risk Spillover in China’s Economic and Financial System

This paper uses a rolling method to dynamically examine the time-order features of China’s financial and economic system. The rolling sample period in this paper is 96 weeks (about two years), and the prediction period is four weeks (about one month). To reduce the impact of noise, the following results are obtained by taking the monthly average of weekly data and converting them into monthly data.

Figure 2 shows that the overall risks in China’s financial and economic system has periodic characteristics. With the setting that “low point—increase—high point—decrease—low point” forms a complete period, we can divide the overall risks in China’s financial and economic system into four periods. The relevant statistical results are shown in Table 2.

Figure 2

The Time-Order Features of Overall Risk Spillover in China’s Economic and Financial System

From Figure 2 and Table 2, it can be seen that risks in China’s economic and financial system were the highest in Period IV and the lowest in Period III; and that risks were higher in Period II than in Period I. In Period I, overall risk in China’s economic and financial system featured rapid fluctuation, and peaked in February 2004. This period witnessed the early development of China’s financial industry, with frequent non-standard behaviors and financial regulatory actions. Therefore, uncertainties and volatilities were both high. In February, the State Council released nine opinions on the development of capital market, which was conducive to the sound and steady development of the financial system. After that, the overall risks in the economic and financial system were gradually mitigated. In Period II, the overall risks in China’s economic and financial system kept at a high level. This period covered global financial crisis and euro debt crisis. Also impacted by this external event, China witnessed high level of risks. In Period III, the overall risks in China’s economic and financial system were at a low level. But it should be noted that a small peak occurred in June 2013. At that time, “liquidity shortage” happened to China’s banking sector. This event had a heavy impact on the financial system, but as the liquidity shortage was a short-term issue rather than liquidity fracture, it had less impact on the real economy. The overall risks in the economic and financial system were low. In Period IV, the overall risks in China’s economic and financial system fluctuated rapidly again, and reached the historical highest point from September 2015 to April 2017. The “abnormal fluctuations” in the stock market in September 2015 and the China-US trade frictions started in April 2017 have posed high uncertainties to China’s economic development, and thus the risk spillover between the financial system and the real economy aggravated. Risks in the economic and financial system were high. In addition, compared with the “abnormal fluctuation” in China’s stock market, China-US trade frictions had smaller impact on the overall risks in China’s economic and financial system.

4.3 Comparison between the Risk Spillover within the Financial System and the Real Economy

This part studies the risk spillover within the financial system and the real economy. Hereinafter, the risk spillover among all sub-sectors within the financial system will be called financial system risks, and that among all sub-sectors within the real economy will be called real economic risks.

To accurately analyze financial system risks and real economic risks, this paper excludes the spillover effect of each sector on itself when calculating the risk spillover level. Meanwhile, this paper measures the risks of these two systems with the average spillover excluding the differences in the number of sectors, in order to precisely compare financial system risks and risk economic risks. Table 3 shows the average of financial system risks and real economic risks in different periods. Figure 3 shows the time-order features of financial system risks and real economic risks. The following conclusions can be drawn from Table 3 and Figure 3.

Figure 3

The Time-Order Features of Financial System Risks and Real Economic Risks

Note: The solid line (dotted line) represents the average risk spillover level of all sectors within the financial system (real economy).

First, the real economy is the major source of risks in China’s economic and financial system. Real economic risks reached the highest level in Period IV, followed by Period II and Period I, and the risk level in Period III was the lowest. Comparing Figure 2 and Figure 3, we can see that the trend of real economic risks is almost the same as that of China’s economic and financial system risks. This indicates that real economic risks are the major driving force of China’s economic and financial system risks. Financial system risks reached the highest level in Period I, followed by Period III and Period IV with the former slightly higher than the latter, and the risk level in Period II was the lowest. From this, it can be seen that financial system risks and risks in China’s economic and financial system have differences in time-order features. In Period IV, when risks in China’s economic and financial system reached the highest level, financial system risks merely ranked three among all the four periods. In Period III, when risks in China’s economic and financial system were lowest, financial system risks ranked second highest among all the four periods.

Second, generally speaking, real economic risks are higher than financial system risks. In particular, In Period I, financial system risks were higher than real economic risks. Apart from this period, real economic risks were higher than financial system risks in all other periods. In addition, the gap between financial system risks and real economic risks was the largest in Period II, followed by Period IV and Period III, and the lowest in Period I. This means that the extent to which financial risks higher than real economic risks in Period I was much lower than the extent to which real economic risks higher than financial system risks in other periods. This conclusion indicates again that real economic risks are the major contributor to risks in China’s financial and economic system.

4.4 Dynamic Risk Spillover of the Financial System and the Real Economy

This part studies the dynamic risk spillover between the financial system and the real economy and the role of the financial system in China’s economic and financial system in different periods.

First, in the four periods, the financial system mainly played the role of risk absorption, but at some specific times, it amplified risks. In Period I, the financial system mainly absorbed the risks, with its risk absorption effect first gradually strengthened and then gradually weakened. But in February to March 2002, October 2002 to April 2003, the financial system amplified risks. In Period II, the financial system mainly absorbed risks, and its risk absorption effect was first strengthened, then weakened, then strengthened again, and then weakened again, forming an “M” shape. But in April to December 2006, March to June 2010, the financial system amplified risks. In Period III, the financial system absorbed all the risk, but its absorption effect was weak. In Period IV, at the earlier stage, the financial system amplified risks; while in the later stage, it absorbed risks, and the risk absorption effect was gradually strengthened. In particular, in December 2014 to June 2015, November 2015 to March 2016, the financial system amplified risks. And at other times, the financial system absorbed risks.

Second, the strength of risk absorption of the financial system is relevant to the risk level of economic and financial system. The comparison of Period I, Period II and Period III shows that the risk absorption of the financial system was the strongest in Period II, followed by the later part of Period I, and the weakest in Period III. The reason is that in Period III, the overall risk level was low, and the risk spillover from the real economy to the financial system was weak. Therefore, the volume of risks that needed to be absorbed by the financial system was small. But in the later part of Period I and Period II, the overall risk level was higher, and the risk spillover from the real economy to the financial system was strong. Therefore, the volume of risks that needed to be absorbed by the financial system was larger. It should be noted that in the earlier part of Period IV, risks in the financial system were at a high level, while those in the real economy were at a lower level. At that time, the financial system amplified risks in the real economy.

This result can be proven by Table 4. In the former three periods, when the financial system absorbed risks, the absorption rate was positively correlated with risk level of China’s economic and financial system. The correlation coefficient between the two was 0.763 in Period II, followed by 0.603 in Period I and 0.078 in Period III. This shows that the intensity of the risk absorption rate of the financial system was positively correlated with risks in China’s economic and financial system. In Period IV, the risk absorption rate of the financial system was negatively correlated with risks in China’s economic and financial system. The main reason is that in the earlier part of Period IV, abnormal fluctuation of the stock market in 2015 led to risk spillover, and thus the financial system amplified risks.

Third, risk absorption of the financial system is positively correlated with risk level of the real economy, and negatively correlated with risk level of the financial system. It can be seen from Table 4 that risk absorption rate of the financial system is negatively correlated with risk level of the financial system. Except for Period IV, risk absorption of the financial system was positively correlated with risk level of the real economy. Observing the former three periods, in which the financial system absorbed risks, we can see that the absolute values of the correlation coefficient between the financial system risk absorption rate and the real economic risk as well as financial system risk was the highest in Period II, followed by Period I and Period III. This is consistent with the aforementioned conclusions. It indicates that the intensity of risk absorption of the financial system is not only correlated with economic and financial system risks, but also correlated with the risks in the financial system and the real economy respectively. The reason is easy to find: risks in the economic and financial system are the superposition of risks in the financial system and in the real economy. It should be noted that as the intensity of the risk amplification of the financial system was strong in the earlier part of Period IV, risk absorption and risk amplification in this period were balanced.

Fourth, when the risk of the financial system is higher than (lower than) the risk of the real economy, greater gaps would mean stronger risk amplification (absorption) effect of the financial system. In particular, when the risk of the financial system is higher than the risk of the real economy, the larger the gap, the stronger the risk amplification effect of the financial system. For instance, based on Figure 3 and Figure 4, before June 2004 and on the eve of “abnormal fluctuation” in the stock market in 2015, risks in the financial system were much higher than the risks in the real economy, and the risk amplification of the real economy was strong. When the risk of the real economy is higher than the risk of the financial system, the larger the gap, the stronger the risk absorption effect of the financial system. For instance, based on Table 3, Figure 3 and Figure 4, in Period II, the gap between the risk of the real economy and the risk of the financial system was the largest, and in this period, the risk absorption effect of the financial system was the strongest. In addition, during the global financial crisis, euro debt crisis, “liquidity shortage” of China’s banking sector and China-US trade frictions, the gap between the risk of the real economy and the risk of the financial system was relatively larger, and the risk absorption of the financial system was stronger. The possible reason for these results is that when risks in financial system are higher, the risk spillover from the financial system to the real economy is stronger, and thus the risk absorption rate of the financial system is low, and the risk amplification of the financial system is strong. On the contrary, when risks in real economy are higher, the risk spillover from the real economy to the financial system is stronger, and thus the risk absorption rate of the financial system is high, and the risk absorption of the financial system is strong.

Figure 4

The Time-Order Features of the Risk Absorption Rate of the Financial System to the Real Economy

Note: The dotted line represents that risk absorption rate is 0.

4.5 Dynamic Risk Spillover of the Financial Sub-Sectors and the Real Economy

To find which financial sub-sector contributes to the risk amplification and risk absorption of the financial system to the real economy, this part studies the risk spillover between financial sub-sectors and the real economy. Table 5 shows the average risk absorption rate of financial sub-sectors in all periods. Figure 5 shows the time-order features of risk absorption rate of financial sub-sectors.

Figure 5

The Time-Order Features of the Risk Absorption Rate of Financial Sub-Sectors

Note: The upper part represents the risk absorption rate of the banking sector to the real economic risks, the middle part represents that of the diversified financing sector, and the lower part represents that of the insurance sector. The dotted line means risk absorption rate equals zero.

First, risk spillover of financial sub-sectors in different periods showed heterogeneity. The risk absorption of the banking sector showed similar intensity in the former two periods, and the intensity in Period III was the weakest. The possible reason is that the “liquidity shortage” in Period III mainly happened to the banking sector, and with the heaviest impact, the functions of banking sector were most severely undermined. The risk absorption of the diversified financing sector was gradually strengthened from Period I to Period III, and was weakened to the lowest level in Period IV. This shows that the “abnormal fluctuation” in China’s stock market in Period IV had the heaviest impact on the diversified financing sector, and thus undermined its function. The risk absorption of the insurance sector was the strongest in Period II, followed by Period IV and Period III, and was the weakest in Period I. The possible reason is that in Period I, the insurance sector was boosting. In 2006, the national premium income reached RMB 564.1 billion, 1.8 times that of 2002. The incomplete development of the insurance sector affected its risk management function.

Second, comparing the risk absorption and risk amplification of financial sub-sectors in different periods shows the following findings. From the static results in Table 5, it can be seen that generally speaking, the diversified financing sector has the strongest risk absorption function, followed by the insurance sector, and the banking sector has the weakest. This is consistent with the results in Table 1. In particular, the banking sector had the strongest risk absorption in Period I, and in Period IV it had risk amplification effect. The diversified financing sector had strong risk absorption in Period II, Period III and Period IV, with risk absorption rate of 0.474%, 0.673% and 0.261% respectively. The insurance sector had the weakest risk absorption in Period I, and in other periods, its risk absorption was weaker than the diversified financing sector, but stronger than the banking sector. The possible reason is that in Period I, the development of the banking sector was more mature, while that of the insurance sector was incomplete. Therefore, the risk absorption of the banking sector was the strongest, and that of the insurance sector was the weakest. In other periods, the diversified financing sector and the insurance sector had matured. The emergence of diversified financial products enabled investors avoid risks through financial speculation and hedging. The risk management of this sector effectively mitigated real economic risks.

From the dynamic results in Table 5, it can be seen that both the risk absorption and the risk amplification of the financial system are closely related to the three financial sub-sectors. But in different periods, the dominant sub-sectors are different. In Period I, the trend of the risk absorption of the financial system was similar to that of the insurance sector, which indicates that in this period, the risk absorption of the financial system was dependent on the insurance sector. During the global financial crisis and euro debt crisis, the banking sector and the insurance sector both had two long peaks, which was consistent with the trend of the risk absorption rate of the financial system. This indicates that in this period, the risk absorption of the financial system was dependent on the banking sector and the insurance sector. During the “liquidity shortage” of China’s banking sector, the trend of the risk absorption rate of the financial system was similar to that of the banking sector, which indicates that the banking sector had a heavy influence on the risk absorption of the financial system. During the “abnormal fluctuation” of China’s stock market, the risk absorption rate of the diversified financing sector reached its minimum, and its risk amplification effect was obvious. This indicates that the risk amplification effect of the financial system is dominated by the diversified financing sector. The possible reason is that this sector covers investment banks, brokerage and diversified capital market, which are the most closely related to the stock market.

4.6 Robustness Test

^[1]This paper uses four methods for robustness analysis. First, it uses correlation coefficient method, which is more intuitive, to replace Diebold-Yilmaz model to measure risk correlation. Second, it uses the financial data of the real economy to calculate the real return on capital investment of all sectors, in order to examine the correlation among the return rates of sectors in the real economy and to demonstrate the rationality of data in this paper. Third, it takes financial sub-sectors as examples to further discuss whether the risk absorption effect of the financial system will increase its own risks. In most periods, the risk absorption of the financial system increased its own risks, and risk spillover reduced its own risks. Fourth, it uses the financial data of banks to verify the risk absorption mechanism of the financial system for the real economy. This paper finds that the higher the risk level of banks, the lower the risk absorption rate of banking sector to the real economy. The empirical research results of this paper are an intuitive reflection of theoretical mechanism rather than a logical mapping of the selected data.