针对bootstrap方法的中国银行业效率变动分析.pdf

硕士学位论文摘要中国银行业效率变动实证分析是当前学界研究的热点。多数研究采用基于 DEA方法的Malmquist指数分析方法。但运用该方法进行银行效率测算,只能根据样本数据给出一个点估计。为了更稳健地估计银行效率变动和分析银行业效率的影响因素,本文采用了Bootstrapped 年间的效率变动进行测算和分解分析。文章选取存款总额、固定资产和管理费用作为投入指标,税前利润、贷款和利息收入作为产出指标。先是运用Bootstrap方法估计了22家银行的技术效率变化及其置信区间,并通过Bootstrap纠偏提高了效率测度的准确性。实证表明, Bootstrap方法在效率测算中具有明显的优越性,经过Bootstrap方法纠偏后的效率值可更清楚的区分决策区间的效率。接着文章计算了Bootstrapped Malmquist指数,并对此进行了分解研究,详细分析了银行业效率的纯效率变化、规模效率变化、纯技术变化和规模技术变化等几个方面的变化情况和趋势。分析表明:,中国银行业效率变动呈现出先上升后下降然后再次上升的波动趋势,但总体上中国银行业的效率是进步的;纯效率变化、规模效率变化和纯技术变化对银行业效率的提高有着正的贡献,其中纯技术进步是推动银行业效率进步的主要动力,技术规模变化则最不明显,显示我国银行业有向CRS边界靠近的趋势。关键词:银行效率;Bootstrap方法;Bootstrapped Malmquist指数 Abstract Currently,some research isfocusing on theanalysis ofefficiency changes inthe Chinese banking most studies use approach to estimate the efficiency of ,this method can only give a point estimation from order toestimate theefficiency changes ofbanks and the effect factors of banking industry in a robust way,this article employs the Bootstrapped Malmquist approach tocalculate and analysis theefficiency changes of 22 banks inchina over theperiod 2001 to2010. This article selects deposits&short term assets and overheads as inputs,profits beforetax,loans interest e bootstrap to estimate thetechnical efficiency scores and confidence intervals ofall22banks atthe beginning,and also improve the accuracy oftheefficiency estimation by correcting thebias intheoriginal results show thatthe bootstrap approach has obvious advantage intheefficiency estimates,the bias—corrected efficiency scores can behelpful indistinguishing between decision units. Then the article calculates the Bootstrapped Malmquist indices and pose theMalmquist index into pureefficiency,scale efficiency,pure technology change and changes inthescale oftechnology toanalyse theefficiency changes and trends of banking industry results obtained show