Composite Indicator of Financial Development in a Benefit‐of‐Doubt Approach

Abstract

We use data by the World Economic Forum (WEF) to build a Composite Financial Development Index (CFDI) alternative to the WEF financial index. Unlike the WEF index, the CFDI optimally combines seven dimensions with non‐fixed weights determined endogenously without recourse to subjective opinions of experts. The CFDI is obtained by applying a Data Envelopment Analysis linear programming model with proportion restrictions on weights calculated in a Benefit‐of‐Doubt approach. In this way, the CFDI scores allow us to group 60 countries by different levels of financial development identifying dimensions that contribute to good or poor financial performances, taking into account the different characteristics of the financial systems in the countries under investigation.

Beyond the Z‐score: A novel measure of bank stability for effective policymaking

“…1 For example, Despotis (2005) possible for each bank regardless of the weighting scheme used (Giambona & Vassallo, 2013). Lastly, Models (1) and (2) eliminate the subjective aspect of weight selection, thus providing relatively objective performance scores for each sampled bank (Zhou et al, 2007).…”

“…It is not possible to get a dimensional/composite indicator value greater than one if another bank uses this weight set (Thanassoulis et al, 2016). Fourth, the weights are optimised to get the best results possible for each bank regardless of the weighting scheme used (Giambona & Vassallo, 2013). Lastly, Models (1) and (2) eliminate the subjective aspect of weight selection, thus providing relatively objective performance scores for each sampled bank (Zhou et al, 2007).…”

A non-parametric index of corporate governance in the banking industry: An application to Indian data

“…In the BoD model, the composite index is constructed by treating all the dimensions/indicators as outputs, thereby considering no inputs in the model (Lovell et al 1995). In fact, a BoD model compares the actual performance of the unit with an internal benchmark rather than an external benchmark that could not be realistically achievable in the specific local context (Giambona and Vassallo 2013). In fact, in the absence of true weights, BoD automatically assigns the benefit-of-the-doubt weights determined by the data to each characteristic in order to build up the composite score, one for each unit (Witte and Rogge 2011).…”

“…U represent lower and upper bound on endogenous weight assigned to th i dimension for j th unit. The similar model is used by Badasyan et al (2011) and Giambona and Vassallo (2013). In the present study, we set lower bound as 10 percent, and the upper limit is assigned accordingly.…”

“…Accordingly, twelve alternative corporate governance indices are calculated (see Table C2 in the Appendix C for details on CGI scores corresponding to lower bound from 5 percent to 16 percent). Similar procedure is adopted for conducting a sensitivity analysis by Giambona and Vassallo (2013) in the construction of financial development index. From Table 5, we note that unlike traditional equal weight or unweighted approaches, constrained BoD approach rewards for policy priority by banks and penalizes under-performance in one or more dimensions.…”