Group 1 - The core view of the report indicates that the spread of credit bond ETFs is primarily influenced by two factors: market sentiment and calendar effects. Recent trends show that the spread of credit bond ETFs continues to recover due to strong certainty in the short to medium credit market [7][8]. Group 2 - In the primary issuance segment, net financing has increased. For the week of March 16 to March 20, 2026, short-term financing issued amounted to 1190.2 billion yuan, while maturing debt was 1100.1 billion yuan. The total issuance of major credit bond varieties reached 3824.3 billion yuan, with net financing of 949.7 billion yuan, an increase from the previous week [16]. Group 3 - In the secondary trading segment, transaction volume increased, and the overall term spread widened. During the same week, major credit bond varieties had a total transaction volume of 9214 billion yuan, an increase of 865 billion yuan from the previous week. The yields on medium-term notes decreased overall, with the 3-year AAA medium-term note yield down by 1.9 basis points to 1.78% [19][20].

Report Industry Investment Rating The report does not mention the industry investment rating directly. Core Viewpoints of the Report - The investment value of credit - bond ETFs with a larger discount may depend on the investor's entry timing, and timing is more important than choosing the ETF. In the interest - rate decline stage, these ETFs may have higher potential returns and safety cushions, but are limited by liquidity. In the interest - rate increase stage, they may carry greater potential risks and more prominent liquidity disadvantages. [74][112] - In the future, as the products mature, the premium - discount rate of credit - bond ETFs may change from unilateral discount to bilateral discount and premium. The premium - discount rate of market - making bond ETFs is expected to stabilize earlier. However, the current low redemption caps of some products may pose liquidity risks in the long run. [66][67] - The difference in the premium - discount rate of credit - bond ETFs is mainly due to differences in redemption mechanisms and liquidity. Full - cash redemption ETFs have a lower absolute value of the premium - discount rate, and higher - liquidity ETFs can more effectively smooth out the premium - discount rate. [44][56] - Based on the daily - frequency premium - discount rate, the quantitative arbitrage models of single - credit - bond ETF and cross - ETF of the same index are constructed. The performance of different models and strategies varies among different types of credit - bond ETFs. [115][131] Summary According to the Directory 1. What is the premium - discount rate? - The premium - discount rate, also known as the discount - premium rate, measures the deviation between the secondary - market price of an ETF and the true value of its underlying assets. When the market price is higher than the net value, it is a premium; otherwise, it is a discount. For credit - bond ETFs, the market generally compares the official daily closing net asset value (NAV) because the underlying credit bonds are numerous and less actively traded, and the IOPV is not regularly disclosed. [9][13] - The deviation between the secondary - market price of credit - bond ETFs and the underlying asset NAV is the result of the combined action of market microstructure, investor sentiment, and arbitrage mechanisms. [12][13] 2. What are the characteristics of the premium - discount rate of credit - bond ETFs? - As of March 6, 2026, the total scale of credit - bond ETFs reached 524.4 billion yuan. Most credit - bond ETFs are currently in a discount state, but the characteristics of the premium - discount rate of each credit - bond ETF vary significantly. [21] - Old - three credit - bond ETFs: They were listed earlier. In the early listing period, due to factors such as a small market capacity, few participating institutions, and an imperfect market - maker system, the discount was large, and then gradually converged. As of March 6, 2026, the premium - discount rates of the short - term financing ETF, urban investment bond ETF, and corporate bond ETF were - 0.4/5.0/3.0 BP, respectively, at a medium or relatively high historical level. [26] - Market - making bond ETFs: They were slightly discounted in the first week of listing, then in a slightly premium state for a long time, and have been in a discount state since July 2025. As of March 6, 2026, the median premium - discount rate was - 9.1 BP, at a medium historical level. [31] - Science and technology innovation bond ETFs: Except for the Fuguofund on the listing day, they were in a premium state, but most turned to a discount state after one week and have been in a discount state since then. The absolute value of the premium - discount rate of 8 full - cash redemption products is relatively low. As of March 6, 2026, the median premium - discount rate was - 10.8 BP, at a relatively high historical level. [36] - The reasons for the differences in the premium - discount rate of credit - bond ETFs are mainly the differences in redemption mechanisms and liquidity. The difference in redemption mechanisms is an important reason, and full - cash redemption ETFs have a lower absolute value of the premium - discount rate. Liquidity difference is the core reason, and higher - liquidity ETFs can more effectively smooth out the premium - discount rate. [44][56] - In the future, the premium - discount rate of credit - bond ETFs may change from unilateral discount to bilateral discount and premium. Market - making bond ETFs are expected to achieve stability in the premium - discount rate earlier, but the low redemption caps of some products may pose liquidity risks in the long run. [66][67] 3. Are credit - bond ETFs with a larger discount more cost - effective? - The investment value of credit - bond ETFs with a larger discount depends on the investor's entry timing. In the interest - rate decline stage, these ETFs may have higher potential returns and safety cushions, but are limited by liquidity. In the interest - rate increase stage, they may carry greater potential risks and more prominent liquidity disadvantages. [74][112] - In the short term, credit - bond ETFs with a larger discount may provide a tool for increasing returns during a phased decline in interest rates, but attention should be paid to the potential restriction of liquidity disadvantages on the strategy capacity. In the long term, it is more important to focus on the fund manager's net - value management ability and the product's own liquidity, and credit - bond ETFs with stable discount - premium may have greater long - term exploration advantages. [106][111] 4. Daily - frequency quantitative arbitrage strategies for credit - bond ETFs based on the premium - discount rate 4.1 Single - credit - bond ETF daily - frequency premium - discount rate arbitrage model - A daily - frequency premium - discount rate arbitrage model for credit - bond ETFs is constructed based on the premium - discount rate of the closing price on day T relative to the NAV on day T. The basic model uses the rolling 20 - day quantile of the daily premium - discount rate of each credit - bond ETF to determine portfolio position operations. [115][116] - The basic model is improved by adding opening - price adjustment and interest - rate timing. The back - testing results show that for old - three credit - bond ETFs, various arbitrage strategies have poor effects; for market - making bond ETFs, the premium - discount rate arbitrage strategy and the premium - discount rate & opening - price adjustment strategy have insignificant thickening effects, but the premium - discount rate & interest - rate timing strategy significantly thickens the risk - return ratio; for science and technology innovation bond ETFs, the first - batch science and technology innovation bond ETFs have obvious thickening effects in various arbitrage strategies, and most science and technology innovation bond ETFs have good thickening effects under the interest - rate timing strategy. Credit - bond ETFs with a larger discount may have better potential performance - thickening effects in arbitrage strategies. [117][121] 4.2 Daily - frequency premium - discount rate arbitrage rotation model for different credit - bond ETFs of the same index - A daily - frequency premium - discount rate arbitrage rotation model for different credit - bond ETFs of the same index is constructed, aiming to buy the "cheapest" credit - bond ETF of the same index to obtain potentially greater returns. There are "absolute" (the largest discount) and "relative" (the lowest quantile) schemes. [131] - The back - testing results show that for Shanghai market - making credit - bond ETFs and Shenzhen market - making credit - bond ETFs, various arbitrage strategies have poor performance - thickening effects, possibly due to high trading costs caused by frequent position changes; for Shanghai AAA science and technology innovation bond ETFs, various rotation arbitrage strategies based on the absolute low of the premium - discount rate have obvious performance - thickening effects; for China Securities AAA science and technology innovation bond ETFs, the rotation arbitrage strategy based on the absolute low of the premium - discount rate performs well, and the performance thickening after adding interest - rate timing is obvious. The Shanghai AAA science and technology innovation bond may be a better choice for cross - ETF rotation arbitrage of the same index. Adding interest - rate timing can improve the arbitrage effect, indicating that interest - rate timing may be more important than choosing the index and ETF. Attention should also be paid to the potential restriction of liquidity on the strategy capacity. [140][173]

1. Report industry investment rating - This report does not provide an investment rating for the industry [1] 2. Core View of the Report - The secondary market discount may lead to a temporary increase in the cost - effectiveness of credit bond ETFs. If the trading sentiment recovers in the future and the ETF premium - discount rate reverts to the central level, it may drive the ETF price to recover. It is recommended to moderately focus on the temporary trading opportunities brought by credit bond ETFs entering the discount range [2] 3. Summary by Relevant Catalog 3.1 Credit bond ETF performance - From July 21 to July 25, affected by the strong performance of the equity and commodity markets, the benchmark - made credit bond ETFs and science - innovation bond ETFs were under pressure. The net worth and closing prices of 8 benchmark - made credit bond ETFs and 10 science - innovation bond ETFs declined for 4 consecutive days from July 21 to July 24 and showed differentiation on July 25. As of July 25, 2025, the closing prices of 10 science - innovation bond ETFs fell below 100 yuan, with cumulative declines ranging from 0.28% to 0.57% since listing; the closing prices of 8 benchmark - made credit bond ETFs also fell below 101 yuan, with declines ranging from 0.45% to 0.66% from July 17 to July 25. On July 28, the net worth of both types of ETFs recovered slightly [5] - Since the listing of science - innovation bond ETFs on July 17, the net capital inflow has rapidly declined. From July 23 to July 25, the benchmark - made credit bond ETFs saw a large - scale capital outflow. On July 24, the single - day net outflow of benchmark - made credit bond ETFs reached 2.52 billion yuan, the largest since the beginning of this year. More than half of the science - innovation bond ETFs had a small net capital outflow on July 24, just one week after listing [4][6] 3.2 How to view the discount trading opportunities of credit bond ETFs - Credit bond ETFs can be traded in the secondary market. Currently, 8 benchmark - made credit bond ETFs use physical subscription and redemption methods, and only 3 out of 10 science - innovation bond ETFs use cash subscription and redemption methods [12] - In the physical subscription and redemption model, investors are more likely to sell credit bond ETFs in the secondary market during market adjustments, which may lead to short - term discounts. After a short - term discount occurs, there are two possible directions: the market continues to adjust and the discount deepens, or the premium - discount rate reverts to the central level due to arbitrage activities [13] - Regarding the influencing factors, from July 17 to July 25, the valuation adjustments of the underlying component bonds of benchmark - made credit bond ETFs and science - innovation bond ETFs were not significantly different from those of non - component bonds, and no additional factors causing excessive valuation adjustments were observed [14] - From July 21 to July 25, the discount of benchmark - made credit bond ETFs deepened, and science - innovation bond ETFs gradually turned to a discount state. As of July 24, the discount of 8 credit bond ETFs ranged from 25 to 47 BP, significantly higher than the historical average; as of July 25, the discount of science - innovation bond ETFs ranged from 8 to 25 BP [17][18] - Although both types of ETFs are in the discount range, the trading sentiment may recover in the future, pulling the premium - discount rate back to the central level. The discount may bring a temporary increase in cost - effectiveness, and it is recommended to pay attention to the trading opportunities [19]