A Modified Arbitrage-Free Nelson-Siegel Model : An Alternative Affine Term Structure Model of Interest Rates

A Modified Arbitrage-Free Nelson-Siegel Model : An Alternative Affine Term Structure Model of Interest Rates
复制标题

修正的无套利尼尔森-西格尔模型:另一种利率仿射期限结构模型

DOI:
10.1007/s10690-014-9191-x
复制
发表时间:
2015
期刊:
Asia-Pacific Financial Market
影响因子:
--
通讯作者:
大西匡光
大西匡光
中科院分区:
--
文献类型:
--
作者:
SIM DARA;大西匡光

文献摘要

相似文献

本文推导了一个修正的无仲裁Nelson-Siegel(AFNS)模型,其中用CIR过程代替了AFNS模型中的高斯水平因子。通过这样做,得到的模型是仿射动态期限结构模型的一个子类,试图保持Nelson-Siegel(NS)属性。从经验上讲,我们发现,对于美国财政部数据,我们的模型比AFNS更适合,但这被较长期限的较弱预测性能所抵消。对于日本政府债券零收益率数据,我们发现我们的模型和AFNS模型之间没有显著差异。此外,对于这两个数据集,我们的模型在一定程度上保持了NS属性。我们认为,由于我们的模型和现有的AFNS模型之间的相似性,为建模的目的,它可以被用作替代模型。据我们所知,本文是第一个研究的经验表现和性质的AFNS模型的日本国债零息票收益率。
This paper derives a modified version of the Arbitrage-Free Nelson–Siegel (AFNS) model in which the Gaussian level factor in the AFNS model is replaced by a CIR process. By doing so, the resulting model is a subclass of theclass of affine dynamic term structure models, that tries to maintain the Nelson–Siegel (NS) property. Empirically, we found that for the US treasury data, our model fits better than AFNS, but that is offset by weaker forecasting performance for the very long maturities. For the Japanese Government Bond (JGB) zero-yields data, we found that there is no significant difference between our model and the AFNS model. In addition, for both data sets, our model to some extent, did maintain the NS property. We argued that because of the similarity between our model and the existing AFNS model, for modeling purpose it may be used as an alternative model. To our knowledge, this paper is the first to study the empirical performances and properties of the AFNS model for JGB zero-coupon yields.