Changes in Government Bond Pricing Anchor

Historical data analysis indicates that, compared to the 1-year MLF (Medium-term Lending Facility) rate, the 7-day reverse repo rate is the actual anchor for the pricing of the 10-year government bond yield.

Since 2024, China's bond market has shown a significant bull market trend.

On January 9th, the yield of the 10-year government bond broke through its original pricing anchor—the 1-year MLF operation rate of 2.50%—and continued to decline, repeatedly setting new lows not seen since April 26, 2002, for over 22 years.

The anchoring effect of the MLF rate on the pricing of the 10-year government bond has been weakened.

On the other hand, the central bank emphasizes that the 7-day reverse repo operation rate should be the main policy rate.

The interest rate cut that began on July 22nd of this round was initiated by the 7-day reverse repo as the policy signal, with the MLF rate adjustment at the end of the chain "7-day reverse repo rate -> SLF (Standing Lending Facility), LPR (Loan Prime Rate) -> deposit挂牌利率 -> MLF."

Advertisement

So, how will the pricing anchor for the 10-year government bond yield be determined in the future?

Looking at the historical trend of the relationship between government bonds and MLF, since 2016, the monthly average trend of China's 10-year government bond yield has basically fluctuated around the 1-year MLF operation rate as the central axis.

From 2016 to July 2024, there were 103 months.

Among them, the monthly average of the 10-year government bond yield was higher than the MLF for 52 months, with an average positive spread of 25 basis points (BP); it was lower than the MLF for 51 months, with an average negative spread of -15 BP.

Looking at the entire sample period, the monthly average of the 10-year government bond yield was 5 BP higher than the MLF rate on average, basically fluctuating around the 1-year MLF rate as the central axis, which also indicates that "1-year MLF rate + 5 BP" is the long-term central pricing for the 10-year government bond yield.

Phase by phase, from May 2017 to June 2018, the spread between the 10-year government bond yield and MLF exceeded 30 BP, reaching a maximum of 72 BP, but then the spread between the two was basically narrowed to within 30 BP.

From July 2018 to July 2024, the monthly average of the 10-year government bond yield and the MLF rate formed a more stable relationship.

During this period, the probability of the 10-year government bond yield operating within the [-30 BP, 30 BP] range of the MLF rate was as high as 95%, and the probability of operating within the [MLF-20 BP, MLF+20 BP] range was 82%.

Overall, from 2016 to July 2024, the monthly average of the 10-year government bond yield operating within the [MLF-30 BP, MLF+30 BP] range was 80%.

The monthly average trend of the 10-year government bond yield is not only related to the 1-year MLF rate but also highly consistent with the trend of the 1-year AAA-grade interbank certificate of deposit (CD) maturity yield, especially after July 2018.

Due to the lower frequency of changes in the MLF rate, the frequency of changes in the interbank CD rate, as a market long-term capital rate, is significantly higher than that of MLF, so the timeliness of the impact of the CD rate on government bonds may be significantly greater than that of the MLF rate.

The trend of the 1-year AAA-grade interbank CD maturity yield is highly correlated with R007 and DR007.

Both R007 and DR007 are short-term capital rates, while the 1-year interbank CD is a long-term capital rate, and there is a significant correlation between the two.

The monthly average trend of R007 and DR007 is largely influenced by the central bank's 7-day reverse repo operation rate.

In addition, the monthly average of the 1-year government bond yield is highly consistent with the trend of R007 in both direction and level, and the short-end yield of government bonds has always been highly related to the money market.

The 1-year government bond yield is also basically consistent with the trend of the 1-year interbank CD rate.

Both are 1-year varieties and have a certain competitive substitution relationship.

The difference in their absolute levels was larger during the period from 2016 to June 2018, but after July 2018, the spread between the two narrowed significantly, and their trends became more synchronous.

From a long-term trend perspective, the trend of the 1-year government bond yield fluctuates around the 7-day reverse repo rate, especially after July 2018, which is more pronounced.

Based on the above historical trend analysis, we find that the pricing anchor for the 1-year government bond yield is the 7-day reverse repo rate.

In addition to fluctuating around the 1-year MLF operation rate, the 10-year government bond yield is also highly consistent with the trend of the 1-year AAA-grade interbank CD maturity yield, which in turn is influenced by the transmission of the 7-day reverse repo rate.

The pricing of the 10-year government bond yield may also have a deeper relationship with the 7-day reverse repo rate.

Who is the real pricing anchor for government bonds?

We calculate the correlation coefficients between government bond yields and different interest rates to examine the real pricing anchor for government bond yields.

Looking at the entire sample period from 2016 to July 2024, the correlation coefficient between the 1-year government bond and the 7-day reverse repo is 71%, higher than its correlation coefficient with the 1-year MLF rate of 68%; the correlation coefficient between the 10-year government bond and the 7-day reverse repo is 82%, higher than its correlation coefficient with the 1-year MLF rate of 79%.

It is evident that government bond yields, whether short-term or long-term, have a higher correlation with the 7-day reverse repo rate.

Compared to the 1-year MLF rate, the 7-day reverse repo rate should be more appropriately considered as the pricing benchmark for the 10-year government bond.

During the sample period, the average spread between the 10-year government bond and the 7-day reverse repo rate is 80 BP, and the average spread between the 1-year government bond and the 7-day reverse repo rate is 19 BP.

Therefore, "7-day reverse repo rate + 80 BP" can be regarded as the long-term central pricing for the 10-year government bond yield, and "7-day reverse repo rate + 20 BP" can be considered as the long-term central pricing for the 1-year government bond yield.

The spread between the two is 60 BP, which can be seen as the average term premium from 1-year to 10-year.

To measure which of "7-day reverse repo rate + 80 BP" and "1-year MLF rate + 5 BP" is the real central pricing for the 10-year government bond yield, the author calculated the sum of squared errors between the monthly average of the 10-year government bond yield during the sample period and the pricing central.

The results show that the sum of squared errors with "7-day reverse repo rate + 80 BP" is 5.99, lower than the sum of squared errors with "1-year MLF rate + 5 BP" at 6.29.

This indicates that compared to the 1-year MLF rate, the 7-day reverse repo rate is the actual anchor for the pricing of the 10-year government bond yield.

Therefore, the establishment of the 7-day reverse repo rate as the main policy rate in the current monetary policy control system, with the weakening of the MLF policy rate's role, will not cause the 10-year government bond yield to lose its pricing anchor but will prompt its pricing to become more apparent to the real anchor—the 7-day reverse repo rate.

The long-term pricing central for the 10-year government bond in the future will be benchmarked by "7-day reverse repo rate + 80 BP."

As of the end of July, the 10-year government bond yield has fallen to 2.15%, setting a new low since April 27, 2002, only 45 BP away from the 7-day reverse repo operation rate of 1.70%, and has obviously deviated from the central "7-day reverse repo rate + 80 BP"; the 1-year government bond yield is 1.42%, 28 BP lower than the 7-day reverse repo operation rate, and there is a 48 BP gap from the original long-term price central "7-day reverse repo rate + 20 BP."

It is evident that the current government bond yield is deviating from its long-term price central and is closer to the 7-day reverse repo rate.

The "asset scarcity" is the reason for the current government bond yield deviating from the long-term pricing central.

The current weak real entity financing demand is an important reason for the continuous decline in bond yields.

At present, corporate financing demand is weakening, social financing is increasing less, and asset supply is seriously insufficient.

Among the social financing increase, 59% is bank credit.

As of the end of June 2024, the 12-month rolling new credit scale decreased by 3.14 trillion yuan from the end of the previous year to 19.1 trillion yuan.

At the same time, the scale of bank bond holdings increased by 3.14 trillion yuan, and the demand for bonds is very strong.

In terms of broad funds, the prohibition of manual interest supplementation and the reduction of deposit挂牌利率 have driven bank deposits to move to bank wealth management and funds, and the scale of bank wealth management products and fund shares has increased significantly.

Both mainly allocate funds to bonds.

According to the latest data, as of the end of June 2024, the scale of broad fund bond holdings increased by 3.24 trillion yuan from the end of the previous year to 48.38 trillion yuan.

Financial institutions have abundant liquidity and pressure to allocate assets, but there is little real entity financing demand, and asset supply is limited.

A large amount of funds chase limited assets, thereby driving bond yields to continue to decline.

And extending the duration can obtain greater price difference income, leading to the 10-year and 30-year government bond yields frequently setting historical lows.

Although the government bond yield is close to the 7-day reverse repo rate, the frequency of change of the 7-day reverse repo rate is still relatively low compared to market rates.

The rates more highly related to the 1-year and 10-year government bond yields are the 1-year interbank CD maturity yields, with a correlation coefficient of 96% with the 1-year government bond yield and 92% with the 10-year government bond yield, both significantly higher than the correlation coefficients between other rates.

Although the 1-year interbank CD maturity yield is a secondary market rate, it is an important reference for the issuance rate of bank interbank CDs.

This indicates that the bank's active liability cost—the interbank CD rate—has a significant impact on the pricing of government bond yields.

According to the bond custody data of the Central Treasury, Shanghai Clearing House, and Shanghai Stock Exchange, as of the end of June 2024, the scale of bank-held government bonds is 20.96 trillion yuan, accounting for as high as 68% of the market.

Therefore, if the issuance rate of 1-year interbank CDs continues to decline, there is still room for government bond yields to decline; otherwise, government bond yields face the risk of a callback.当然可以，请提供您需要翻译的内容。