羅玉 肖麗雯 楊澤鵬



[摘 要]近年來,金融平臺及其發行的城投債在經濟增長和社會發展方面發揮了非常重要的作用。然而,市場上與城市債務有關的信用風險并不明顯,而且外部評級機構在發行的城投債方面普遍虛高評級。在這種情況下,通過建立一個邏輯系統,分析城投債償債風險,盡可能選擇具有代表性的指標,通過BP神經網絡等智能算法來實現城投債評級,以計算模型預測的準確性。歸根結底,從結果來看,神經網絡模型BP已超越了傳統方法的界限,并通過模擬非線性模型來模擬實際數據,從而超越了對線性模型的依賴。
[關鍵詞]城投債;matlab ;BP神經網絡;內部評級
[中圖分類號]F810
[文獻標識碼]A
[文章編號]2095-3283(2019)10-00102-03
The Insolvency Risk of City Construction Investment Bonds Based on Machine Learing Algorithm
Luo Yu Xiao Liwen Yang Zepeng
(Jiangxi University of Finance and Economics,Nanchang Jiangxi 330013)
Abstract: City Construction investment bonds issued by Financing platform have played a very important role in Chinas economic growth and social development in recent years. However, due to the particularity and the opacity of local government, the credit risk of city investment bonds in the market has not been clearly exposed, and the rating of city investment bonds is generally high. Under such back ground, we attempt to establish a set of logical system for analyzing the insolvency risk of City Construction investment bonds, select representative indicators as much as possible, achieve stock rating through artificial neural network model, and calculate the accuracy of the model prediction. In the end, from the results, the BP neural network model gets rid of the limitations of the traditional methods, breaks the restriction of the dependence on the linear model, simulates the actual data with the nonlinear model, and realizes the innovation in the method of stock rating application research.
Keywords: City Construction Investment Bond; Matlab; BP Neural Network; Internal Rating
[作者簡介]羅玉(1998-),女,漢族,江西南昌人,本科生,研究方向:會統核算:肖麗雯(1998-),女,漢族,江西南昌人,本科生,研究方向:金融統計;楊澤鵬(1998-),男,江西南昌人,本科生,研究方向:金融學。
[基金項目]江西財經大學2019年度江西財經大學學生實踐創新訓練項目(項目編號:201910421022)。
一、引言
中央政府和地方政府在國家分稅制改革以后,明確規定了財政權力和事權地方財政收入增長以及經濟發展在一定程度上得到提升。與此同時,隨著國家迅速城市化、地方經濟快速增長、地方政府財政支出的增加以及資金缺口逐年擴大,地方政府通過建立金融借貸平臺,在很大程度上彌補了基礎設施融資缺口。
城投債是介于標準企業債和地方政府債務之間的一種準市政債,是由地方政府投資融資方案發行的企業債券作為主要來源,主要用于資助地方基礎設施項目和公共利益項目。……