Studying the determinants of corporate bond credit spreads is not only conducive to the management of credit risk and the pricing of primary issuance of credit products, but also helps secondary investors to judge the relative investment value of corporate bonds relative to interest rate varieties such as government bonds and financial bonds, thus providing research support and decision-making basis for them to seize investment opportunities and choose investment varieties.
Although China's credit bond market started late, investors have also experienced a relatively complete economic and monetary cycle in recent years. We hope to summarize some operating characteristics and laws of credit spreads suitable for the characteristics of China's market, and explore and refine the main influencing factors of credit spreads in China, so as to provide some reference for investors to grasp the changes of credit spreads and make investment decisions in the secondary market.
A summary of credit spread theory
Credit spread refers to the spread between credit bond yield and market risk-free yield, which can be understood as compensation for default risk borne by investors. The usual calculation method abroad is the yield of credit bonds MINUS the yield of government bonds with the same term. The research on credit spread can be traced back to Fisher's (1959) pioneering research on the determinants of credit risk premium. After that, Beaver made a quantitative study on credit risk, and verified for the first time in 1966 that the company's financial data contained abundant default information. Since then, black &; Scholes put forward a classical option pricing formula, which Merton introduced into bond credit risk pricing, called "structural model". On the basis of synthesizing the previous research results, the decomposition model developed after 2000 has gradually become the frontier in the research field of credit spread in American bond market, including Edwin J. Elton(200 1), Driessen(2003), Amato(2003) and Phihpov(2004). The model holds that the main factors affecting the credit spread of enterprises come from default risk factors, tax factors, liquidity risk factors and other systematic factors. However, due to the complexity of influencing factors of credit spread, foreign research has not found a general model to explain the changes of credit spread.
Although foreign theoretical models of credit spreads still cannot accurately decompose credit spreads and explain their causes, this does not prevent us from simply dividing credit spreads into risk premiums, expected losses, liquidity premiums and tax spreads, so as to study the main factors that may affect these parts respectively (see figure 1).
More importantly, in order to better understand the factors affecting credit spreads, we must first understand the differences and connections among expected default losses, actual default losses and risk premiums. The expected default loss is mainly determined by the historical average default rate, which is a relatively stable sequence corresponding to the rating. However, the default does not occur every year according to the historical average default rate, but has the characteristics of "intensive" and "outbreak", that is, the actual default rate every year is lower than the average default rate for many years, but it is obviously higher than the average default rate during the intensive default period. Therefore, in order to avoid absolute credit loss, credit bond investors must consider the whole macro trend and credit cycle in addition to reference rating.
The expected default loss in credit spread reflects the difference and change of rating, that is, the coverage of historical average default rate, while the premium required by macro credit cycle and large-scale default expectation is mainly reflected in risk premium. In addition, investors' demand for additional risk compensation for credit risk is also related to the systemic contagion characteristics of credit risk and the skewness characteristics of credit bond investment income. These characteristics make it difficult to disperse credit risk, so investors will tend to demand compensation beyond expected losses.
Due to the above problems, when investors invest in credit bonds, they are not priced according to the "historical average expected loss", but according to their expected loss rate, which may lead to the fluctuation of risk premium far greater than the actual default rate. To sum up, the expected loss is similar to the intrinsic value of the stock, while the credit spread is similar to the real trading price of the stock, and the difference between them is mainly the risk premium (see Figure 2).
A Review of the Historical Trend of Credit Spread in China
The most direct and necessary way to explore the operating law of credit spreads is to "take history as a mirror" and observe how credit spreads have been manifested under the influence of different supply and demand factors in different historical periods in the past. As China's credit bond market has been established for a relatively short time, the historical performance of credit spreads may be severely restricted by various characteristics in the early stage of market development, and historical data has limited guiding significance for the future. However, we believe that sorting out the factors that affect historical spreads and judging the possible changes of these factors in the future can still play a very good guiding role in predicting the long-term trend of future credit spreads. In this part of the review, according to the trend of the benchmark interest rate, we roughly divide the historical samples into nine segments. The direction of benchmark interest rate in each segment is basically the same, and the next segment will be divided after the direction is reversed. In this way, we can better observe the relationship between the change of credit spread and the trend of benchmark interest rate, and it is easier to find the similarities and differences of the driving factors of credit spread change in similar macro environment.
(1) Second half of 2008
In this magnificent bond bull market, the credit spread first narrowed and then expanded. The yield of high-grade credit bonds decreased obviously under the downward trend of interest rate bonds, and the credit spread at the end of the year and in the middle of the year was not much different. Low-grade credit bonds have different performances. For example, the yield at the end of the year is higher than that in the middle of the year. Although the yield of other bonds at the end of the year is lower than that in the middle of the year, the credit spread is obviously larger than that in the middle of the year, and the spread between ratings is obviously larger, which reflects the typical characteristics affected by default risk.
(2) The whole year of 2009
At the end of 2009, compared with the beginning of 2009, the inter-bank credit spread between AAA and AA slightly expanded, showing the characteristics of being gradually pushed up by the benchmark interest rate. During the period when the yield of interest rate bonds rose rapidly, the credit spread was passively compressed, but after the interest rate bonds were relatively stable, the yield of credit bonds continued to rise and the credit spread actively expanded. In addition to the upward trend of the benchmark interest rate, the substantial increase in supply is also an important factor. Because 2009 was the first year of large-scale expansion of the credit bond market, after winning the bid in 2008, it was issued in large quantities in 2009. Non-financial credit bonds issued in the whole year were close to 990 billion yuan, with a net increase of 700 billion yuan, 2.7 times that of 2008. The release of supply has further increased the pressure to push up credit spreads.
(3) 8 months before 3)20 10/0
The market is sandwiched between rising inflation and worries about a double dip in the economy, and interest rate bonds lack a clear bull market direction. However, in the case of slow fluctuation in the yield of interest rate bonds, good interest margin income has become the object sought after by investors, and credit bonds have been favored. In addition, due to the substantial expansion of credit bonds in 2009, the yield rose sharply, the supply of 20 10 credit bonds declined, and the improvement of the overall supply-demand relationship led to a greater decline in the yield of credit bonds than that of interest rate bonds, and the credit spreads and spreads between ratings narrowed significantly.
(4) September 20th10 to September 20th11.
At this stage, almost all the factors that are unfavorable to credit spread are gathered together. On the one hand, the benchmark interest rate has risen sharply, on the other hand, the supply of credit bonds has rebounded from the low level of 20 10. In addition, the rise of capital interest rate center and the increase of volatility lead to the narrowing of interest margin protection space, which superimposed serious default risk concerns in the later period. The yield of credit bonds rose sharply compared with the beginning and end of the period, and the trend of credit spreads basically showed an upward trend. The credit spread between high and low ratings climbed to the highest level in history, and the spread between ratings also expanded significantly.
(5) From June 20th11to June 20th12.
Driven by the easing of default risk concerns and the easing of funds, the yield of credit bonds has fallen sharply, and the trend of credit spreads reflects the typical characteristics that credit spreads and inter-rating spreads have narrowed at the same time in the downward process of benchmark interest rates. In addition to the cooperation between the downward trend of benchmark interest rate and the downward trend of capital interest rate, the expansion of financial disintermediation in 20 12 years, especially the expansion of products such as bank wealth management, is an important reason for promoting the significant improvement of the demand for credit bonds with higher coupons and narrowing the credit spread from a high level.
(6) the second half of 2012
20 12 The credit spread expanded slightly in the second half of the year, mainly in the third quarter. At that time, the upside of interest rate bonds was relatively large. In the third quarter, the interest rate center of monetary funds was tighter than that in the second quarter, and the overall supply pressure of credit bonds was greater, which led to the active expansion of credit spreads during the bull-bear period. After entering the fourth quarter, the market sentiment was relatively stable, the yield of interest rate bonds only fluctuated slightly upward, the fund surface eased compared with the third quarter, the spread value recovered, and the credit spread changed little.
(7) 5 months before 2013
In the bull market, the credit spread continued to narrow, and the stability of the capital interest rate center also contributed. In addition, the Notice on Regulating the Investment Operation of Commercial Banks' Wealth Management Business (No.8 Document) issued by CBRC in March 2065438+2003 requires banks to control the investment ratio of non-standard assets, which increases the demand for alternative allocation of standard credit bonds and also promotes the narrowing of credit spreads to some extent.
(8) 6 months after 2013
In this bear market, the credit spread widened, the demand for credit bonds shrank due to institutional stop loss, the interest rate center of funds moved up and the volatility increased, and non-standard investment squeezed bond investment funds. Concerns about the default risk brought by the economic downturn also make investors demand a higher risk premium for credit bonds. Based on the benchmark interest rate of national debt, the credit spreads of high-rated and low-rated credit bonds have exceeded the high point in the fourth quarter of 2008, which is close to the most serious point of default risk in the third quarter of 201/kloc-0. The credit spread of low-rated corporate bonds in the exchange has also climbed to the highest level in history, and a number of high-yield bonds with yields exceeding 10% have emerged.
(9) First half of 20 14
The trend of credit spreads also clearly reflects the narrowing of the bull market. The important factor for narrowing the credit spread at this stage is that the bond allocation funds have returned to the disintermediation track after the non-standard "gold absorption method" of 20 13. Bank deposits have been diverted at an accelerated pace, and various wealth management forces such as bank wealth management, securities firms and fund accounts have sprung up. Due to the high capital cost and limited non-standard allocation resources, these institutions can only turn to allocate credit bonds, especially low-grade credit bonds with high coupon, in order to obtain profit space. Therefore, driven by the improvement of risk appetite and the change of investor structure, this round of bond bull market has experienced a perfect rotation from interest rate bonds to high grades and then to low grades. Of course, unlike similar cycles in previous years, substantial defaults began to break out, which led to the unsatisfactory compression of low-rated credit spreads, and the fluctuations during the period were spectacular.
The main factors affecting the credit spread in China
(1) Default risk and investors' expectation of default risk
The factors related to default risk not only affect the expected default loss, but also affect the risk premium. Among them, the former is mainly reflected by the difference of ratings, while the latter is mainly determined by the change of investors' expectation of default risk. First, the credit spreads of bonds with different ratings are different, and the spreads of low ratings are always higher than those of high ratings, and the spreads fluctuate more. Second, rating adjustment will lead to changes in credit spreads, and the negative impact of rating downgrade is usually greater than the positive impact of rating upgrade. Especially in China, the qualification of repurchase of investable bonds and exchange bonds is linked to the main rating, and the negative rating action has a great influence on the credit spread. Third, due to the centralized and explosive nature of default, it is difficult to effectively disperse the credit risk. When the default risk comes, the risk premium will be greatly pushed up and the spread between ratings will be significantly expanded. In China, due to the scarcity of substantial default, there are not many periods when the risk of default significantly affects the credit spread. But once it happens, the impact on credit spreads will far exceed other negative factors, such as the second half of 2008, the third quarter of 20 1 1, the second half of 20 13 and the first quarter of 20 14. Among them, in the third quarter of 201/kloc-0, due to the superposition of many negative factors, such as the sharp rise of the benchmark interest rate, the increase of supply pressure, the rise of the capital interest rate center, and the negative feedback between the fear of default risk and the liquidity crisis caused by the withdrawal of institutional funds, the credit spread adjustment was large and lasted for a long time.
(2) Liquidity compensation
In the theory of credit spread, liquidity risk premium refers to the relatively weak liquidity of credit bonds relative to national debt. For a single bond, its liquidity premium in the secondary market is determined by its issuance scale, holder structure, turnover rate and debt age (old or new). As far as different types of bonds are concerned, because the liquidity of credit bonds is obviously weaker than that of interest rate bonds, investors need to demand a certain premium compensation for this weakness. Similar to the expected default loss can not cover the default risk in a specific period, and the liquidity premium required by investors can not cover the liquidity risk with great uncertainty in a specific period, so traders tend to demand a relatively high liquidity premium when the market liquidity is tight. The typical cycle is the third quarter of 20 1 1. The initial sharp increase in its yield came from the concern about credit risk, but then institutions concentrated on selling, which led to the deep shortage of the secondary market, and finally evolved into a collapse effect caused by the concentrated outbreak of liquidity pressure, which made the yield and credit spread of investment bonds in the exchange city rise sharply to the highest level in history.
(C) the use of operational convenience and interest margin space
In China, it is relatively easy and common to leverage credit bonds. When the spread between the yield of credit bonds and the interest rate of funds is large, the funds are relatively stable, and pledge repurchase is convenient, investors will increase leverage to allocate credit bonds, otherwise, if leverage is removed, the demand will decrease at a very fast rate. Leveraged investment model is a double-edged sword, which actually amplifies the fluctuation of market demand and credit bond yield and spread. When institutions increase leverage, the yield of credit bonds will accelerate downward, thus narrowing the credit spread. When institutions deleverage, the yield of credit bonds can easily soar, thus greatly expanding the credit spread. For example, in the first half of 2009, the first half of 20 12, and the first half of 20 14, when the interest rate of the capital center went down and the volatility dropped, the spread space was at a high level, and the credit spread narrowed all the way from the high level. However, in the third quarter of 20 1 1 and the second half of 20 13, when the credit spread rose to a historical high, we can observe the high cost of capital and the obvious increase in the volatility of capital interest rate (see Figure 3).
(D) Changes in the structure of investors
The main demand subjects of domestic credit bonds are banks, insurance and generalized funds, among which banks and insurance have lower risk preference, while generalized funds have higher risk preference and more rigid demand for coupon. Since 20 12, the background of financial disintermediation has greatly increased the proportion of investors such as bank wealth management, funds and special accounts in the structure of credit bondholders (see Figure 4), which has promoted the trend decline of investors' overall risk preference and credit spread, especially the spread between ratings. Since 20 14, the credit spread has been lower than the historical average and has been continuously compressed, which is closely related to the fact that all kinds of wealth management funds have become the absolute main force of credit bond investment groups under the background of financial disintermediation. Due to the high cost of financial management funds, there are fewer and fewer bonds that can cover the cost, and the risk exposure is very limited. Financial investors can only obtain relatively high spreads through passive credit sinking, thus promoting the continuous compression of spreads between ratings.
(E) Credit bond supply pressure
The supply of credit bonds is mainly determined by the financing willingness of enterprises and the comparison between bonds and other financing instruments. Among them, financing willingness comes from the investment demand and refinancing demand of enterprises. Compared with other financing tools, it mainly depends on what financing channels enterprises can choose and whether bonds have obvious advantages compared with other channels. In China, because the credit bond market is in the early stage of development, the bond approval policy and the innovation of issuance varieties are also important factors affecting supply. It should be noted that potential supply is the essential factor affecting the relationship between supply and demand and credit spread, and actual circulation is the result of demand test. The decrease in actual circulation caused by sales difficulties has reflected the deterioration of supply and demand, rather than the reduction of supply pressure. From the historical experience, the serious negative impact of supply on credit spreads mainly occurs when new products are innovated and greatly expanded, and other stages are reflected in the long-term relatively moderate impact. In history, the main stages of credit spread expansion caused by typical supply increase were the second half of 2009 and the second half of 20 12, and the supply of new products jumped sharply in these two stages. In other years, the negative impact of supply is relatively mild, and the effect will be slower. In addition, the fluctuation of credit bond supply is also pro-cyclical, which is easy to strengthen the bear market or bull market trend of the bond market.
The Historical Relationship between China Credit Spread and the Trend of Benchmark Interest Rate
Judging from the history of credit bonds in China since 2008, the relationship between credit spreads and benchmark interest rates can be summarized as follows:
(1) In general, the trend of credit spreads and benchmark interest rates is consistent.
In other words, in a bull market where the benchmark interest rate goes down, the credit spread will generally narrow, while in a bear market where the benchmark interest rate goes up, the credit spread will generally expand.
(b) When the benchmark interest rate changes rapidly, credit spreads may have a lagging response.
Its performance is that the credit spread is opposite to the benchmark interest rate in a short time, and then turns to the same direction with the benchmark interest rate after a period of time, that is, the credit spread will deviate from the benchmark interest rate in a short time.
(3) Without the influence of credit risk, low-rated bonds usually perform better than high-rated bonds, regardless of bear market or bull market.
In other words, the yield of low-grade bonds in bear market is usually lower than that of high-grade bonds, while the yield of low-grade bonds in bull market is usually higher than that of high-grade bonds. This reflects the strong defensive nature of high coupon. Examples of bear market are the second half of 2009 and 20 12, and examples of bull market are the first August of 20 10, the first half of 20 12 and the first May of 20 13. We believe that the formation of this law is related to the lack of market default and the change of investor structure in the process of financial disintermediation.
(d) In the event of default concern or real credit risk, both low-rated credit spreads and inter-rating spreads will expand regardless of the trend of the benchmark interest rate.
If the expectation of default is enhanced, the negative impact on the yield will be even greater if the interest rate bond adjustment and the shortage of funds are superimposed. Historically, the stages affected by default factors are mainly in the fourth quarter of 2008, 20 1 1 (especially in the third quarter), the second half of 20 13, and the first half of 20 14 (especially in March). Among them, the benchmark interest rate was in a bull market environment in 2008 and 20 14, and in a bear market environment in 20 1 1 year. Spread spread caused by default is not long in history, but it is fatal. If these high-risk periods cannot be safely passed (avoiding a large number of defaults and bearing liquidity shocks), low-grade and relatively high coupon returns cannot be achieved.
Possible long-term changes in China's credit spread in the future
As mentioned above, although we have analyzed the reasons and laws of credit spread fluctuation in China history from many angles, if we look at it from a longer historical period, many existing laws may be related to the primary development stage of China's credit bond market. In the future, with the continuous development and maturity of the market, the dominant factors affecting credit spreads may also change. From the current point of view, we believe that in the longer historical period in the future, the following factors may push the credit spread to change differently from the past:
(1) The "rigid redemption" is gradually broken, and the impact of default-related factors on credit spreads will be significantly enhanced.
Before 20 13, the credit spread of low-grade credit bonds in China was lower than that in the international market. The main reason is that there was no substantial default event in China before this year's super-Japanese debt default, and the market lacked sufficient awareness of premium protection for credit risk. With the development and maturity of the market, this rigid payment situation is difficult to maintain for a long time, and it is not conducive to the optimal allocation of resources and the reduction of financing costs for high-quality enterprises. Although large-scale and centralized defaults still depend on the fluctuation of macroeconomic cycle, the occurrence of partial and individual defaults may become the norm, so the impact of expected losses on credit spreads is bound to become more and more important, which will push up the center of spreads between ratings and aggravate the fluctuation of the yield of low-qualified securities.
(2) The implementation of IRB is beneficial for banks to reduce the spread requirements for high-grade bonds.
At present, only six banks have been approved to use the internal rating method to calculate the risk weight, and most banks adopt a mechanical weight classification method for all kinds of credit assets, that is, basically all credit bonds adopt the risk weight of 100%, and there is no difference between different ratings. Therefore, when banks buy bonds issued by the most qualified enterprises, they must occupy a lot of capital, so they must demand higher interest spread returns to cover capital consumption. In the future, as more banks adopt the internal rating method, investing in high-rated bonds will have more advantages in saving capital, and banks' differentiated requirements for high-and low-grade credit spreads will become more obvious.
(c) It is more difficult to make use of low-rated credit bonds, which may lead to a wider spread between ratings.
As mentioned above, leverage operation is a common operation strategy in China credit bond market. In the long run, banks will be increasingly reluctant to accept low-and medium-rated credit bonds as financing collateral from the perspective of controlling capital consumption and default risk due to stricter restrictions on leverage operation in the bond market by regulators. This will make it more and more difficult to use low-rated bonds as collateral to improve leverage, which may push up the spread between high and low ratings, but the fluctuation range of yield and spread brought by leverage operation may slow down accordingly.
(d) The investor groups are more diversified, and the further development of derivatives such as CDS to hedge the default risk will make the credit risk pricing more sensitive and accurate.
On the whole, the risk preference of credit bond investors in China is relatively low, which is related to the zero tolerance atmosphere of risk in China's bond market, and on the other hand, to the imperfection of China's legal system to deal with default and bankruptcy. In the future, with more and more institutions with stable financial strength and in-depth credit research ability, and with the improvement of default disposal mechanism and legal system, risk pricing will be more real and sensitive. In addition, with the emergence of material default, credit default swaps and other risk derivatives may gradually develop and mature in the long run. Because credit default swap derivatives separate credit spreads and give them liquidity, default risk information can be reflected in the market more purely and timely. In the future, with the diversification of participants, the pricing of credit spreads will become more accurate and detailed, and the volatility and liquidity will gradually increase, which will promote price discovery in the market and provide more trading opportunities for the market.