2025年债市启示录：框架的贫

Group 1: Report's Industry Investment Rating - The industry rating is "Overweight", indicating an expected increase of over 10% compared to the benchmark index in the next 6 - 12 months [50] Group 2: Report's Core View - 2025 was a more "uncomfortable" year for the bond market than 2017, with a worse investment and holding experience. The market should shift from "asset pricing research" to "liability behavior research" and avoid being confined by existing frameworks [3][7] - The traditional bond - market research frameworks are not reliable, as they are mostly established in the past few years and do not adapt well to rapid changes. The focus should be on trading and liability behavior research [45][46] Group 3: Summary by Directory 1. Technology Bull Disrupting Bond Bull (February) - The first bond - market adjustment from February to March was due to the "liability shortage", which was actually the prelude of the year's adjustment. The technology - stock rally diverted funds from fixed - income to equities, and it was also a link between the 2024 "924 stock - bond seesaw" and the 2025 July stock - bond seesaw [7] - The view of "stock - bond double bull" was a fallacy, and the industry's attitude towards technology - stock positions began to change, affecting the stock - bond balance strategy [8][10] 2. Liability Shortage and Seasonal Anomaly (March) - Seasonal patterns in the bond market failed in March, November, and December. This was caused by the seasonality of bank - deposit maturities and the reversal of wealth - management institutions' performance - chasing behavior [11] 3. Trade Disturbance Interlude (April) - The bond - market rally in April was mainly due to the oversold rebound after large banks sold OCI bonds for profit settlement, not trade disturbances. Comparing 2025 with 2018, asset - pricing models overly relying on export forecasts are not robust [14][16] - As of November, exports were a significant positive contributor to GDP, and there are high expectations for next - year's exports [17] 4. Starting Point of Story Reconstruction (June) - Signs of bond - market adjustment could be seen in the equity market in June, such as the failure of the real - estate recovery in the second quarter, the increase of bond - fund duration to a high - level, and investors' misunderstanding of the equity market [19] - The market should change its subjective value judgment on data, and previous interest - rate decline stories were not reliable, such as "Japanization" and "de - globalization" [21] 5. Anti - Involution: Building High and Entertaining Guests (July) - The "anti - involution" market started with policy - catalyzed price "reflexivity". After July, commodity prices rose, and inflation expectations improved, which led to bond - market adjustment [22] - Later, only the new - energy sector could support price "anti - involution", while commodity prices in other sectors fell back, but bond yields did not [23] 6. Which is Primary: Trading or Fundamentals? (September) - In September, the expected improvement in supply - demand from "anti - involution" did not materialize, and two investment logics were in a "voting competition": one based on the micro - trading structure and the other affected by market "news" [25] - Bond trading is primary, and over - emphasizing research frameworks while ignoring trading itself is inappropriate [29] 7. Bond Market Recovery with Limited Impact from Bond Buying (October) - In October, the bond market had a small recovery due to loose expectations and a second - round of trade disturbances. However, several events indicated that the market did not have a long - term valuation recovery [30] - Trading institutions were optimistic and increased duration, while banks' liability - side behaviors showed "redemption during recovery", and bonds with maturities under 10 years were more stable [30][31] - The total bond - investment funds were stable, with banks' bond investments moving from off - balance - sheet to on - balance - sheet. This led to discussions on "bond - market supply - demand" from November to December [33][34] 8. Bond - Market Supply - Demand Issues Becoming Systematized (November) - In November, as the bond market fell after the October recovery, economic data weakened, and the stock - bond seesaw effect failed. The market began in - depth research on "liability behavior" [35] - Insurance - premium structure changes, bank EVE limitations, and the rigidity of interest - rate bond issuance affected bond supply - demand, especially for long - term interest - rate bonds [35] 9. New - Year Opening, Allocation - Oriented Institutions' Linear Allocation, and Withdrawal of Trading - Oriented Institutions (December) - In December, the bond market fluctuated after an initial sharp decline due to concentrated long - term bond supply. Short - term bonds were strong due to central - bank support, accelerated year - end fiscal expenditure, and large - bank net lending [37][38] - Long - term bonds had significant long - short divergence. The mainstream view is that long - term bond supply - demand issues and spread - widening risks exist, and the long - end New - Year market is not worth expecting [40][44] Bond - Market "Traditional Framework" Failure Theory: Leveraging Thinking is "Knowledge Debt" - Most so - called "traditional frameworks" in the bond market are not reliable, as they are newly established and do not adapt well to rapid changes. Over - abstracted bond - market methodologies can lead to "cognitive overload" [46] - The bond market should learn from 2025 and shift from "asset pricing research" to "liability behavior research" in 2026 [48]