AI智能总结
Bryant XuStrategist To understand the relative value in Asia local currency government bonds, we havedeveloped a rich/cheap monitor based on the Nelson-Siegel-Svensson (NSS)model. This systematic monitor seeks to provide actionable insights into bond Jalaj Singh Research Associate The NSS model fits a smooth, continuous yield curve on a universe of observed cashbond yields. This serves as a critical mechanism for eliminating noise attributableto individual bond idiosyncrasies, thereby establishing a robust "fair pricing" Sameer GoelMacro Strategist+65-642-36973 Perry KojodjojoStrategist We note that deviations from the NSS model do not always signify a marketdislocation. Rather, these discrepancies can arise from legitimate factors such asdivergent liquidity conditions (e.g., highly liquid on-the-run bonds frequentlyexhibit lower yields compared with off-the-run bonds of equivalent maturity),specific investor demand (e.g., 10Y FAR IGBs demonstrate substantially lower Vaninder Singh, CFAMacro Strategist Joey ChungStrategist Hazel LaiMacro Strategist Our NSS model-based monitor incorporates several key features to enhance its nData sourcing. We utilize Bloomberg as our primary data source, capturingessentialinformation such as cash bond pricing,maturity dates,outstanding sizes, and bid/ask spreads. This transparent data sourcing nDefining the bond universe. We use a tightly defined bond universe withthehelp of several exclusion criteria.Bonds are excluded if theiroutstanding amount is below $2bn (or $1bn for RPGBs); if they exhibit poor liquidity (e.g., excessively wide bid/ask spreads or infrequent tradingactivity); or if they were issued more than 10 years ago.1 This approach nCustomizable Python-based modeling. Our NSS model, calibrated usingPython, provides granular control over all parameters and facilitates highlycustomized adjustments. This capability is particularly valuable in marketssuch as KTBs, which has a distinctive 'belly' twist, with higher yields in thecurve's mid-section compared with its front and back ends. By adjusting nRich/Cheap charts. The monitor includes two distinct rich/cheap charts -one covering bonds with maturities ranging from 1 to 10 years, and the Highlighting extreme valuations. The monitor clearly highlights the five Our rich/cheap monitor covers eight major Asian local currency governmentbond markets: China, India, Indonesia, Korea, Malaysia, Philippines, Singapore,and Thailand.For details on the underlying mathematical framework and technical Model takeaways The latest readings of our rich/cheap monitor provides several observations on key nChina: Within the CGB curve, a clear contrast in valuations exists: CGBs inthe 11-29 year segment are largely cheap relative to the NSS model, likelyreflecting an illiquidity premium due to subdued trading. Conversely, therichest bonds are typically benchmark or nearby issues in the 7Y, 10Y, and30Y tenors. Their superior liquidity profile, which facilitates easier repotransactions for leveraged investors, drives this higher demand and richervaluation against the NSS curve. An interesting case isCGB 2.88 Feb-33;despite not being a benchmark, it's the richest bond, likely due to its older nIndia: Distinct deviations are observable within the IGB curve. Mostnotably, the two 10Y FAR bonds,IGB 6.33 May-35andIGB 6.48 Oct-35,stand out as the richest points. This richness is attributed to robust demandfrom foreign investors, which likely suppresses their yields compared tocomparablematurities.Interestingly,IGB 6.33 May-35 achieves thisrichness despite not being the benchmark bond. In contrast, other IGBswithin the 9-11 year tenor band exhibit market yields largely consistent with nIndonesia: The valuation (richness/cheapness relative to NSS curve) ofIndoGBs broadly correlates with prevailing liquidity conditions. Evidencingthis, the two richest bonds -INDOGB 6 ½ Apr-36andINDOGB 6 ¼ Jun-36 - are both found within the highly liquid 10Y tenor, with the former serving 20 years consistently appear relatively cheap against the NSS curve, acharacteristic largely attributable to diminished liquidity in that longer- nKorea: The KTB yield curve displays a distinctive shape within the Asiancontext, marked by an elevated 'belly' in the 10-20 year sector compared toboth the front-end (under 10 years) and back-end (exceeding 20 years)maturities. A notable inversion is observed within the 20-30 year segment,primarily driven by strong demand from life insurers for long-dated duration nMalaysia: On the MGS/MGII curve, the richest bonds are predominantlybenchmark bonds in the 5Y and 10Y tenors, a characteristic likely driven bytheir superior liquidity profile. Conversely, bonds with maturities exceeding nPhilippines:RPGBs'market yields generally align with NSS yields.However, the 10Y benchmark bond notably stands out as the richest point nSingapore: The SGS market yield curve closely tracks the NSS curve formaturities under 10 years