The averages we use in the Global Custodian Agent Bank Surveys can tell us much about clients’ current perception of their service providers, but like all statistics they do not and cannot tell the whole story. Over the years, we have toyed with different ways of presenting the results and, to be honest, this is still a work in progress.
Below, for example, are two ways of presenting the results at a market level. The first column – and the traditional presentation in the GC surveys – shows the average score awarded across all survey responses for the categories rated. At first glance this shows that Relationship Management is the best regarded service category at the upper end of ‘Good’ range. Pricing, by contrast, brings up the rear, though remaining in ‘Good’ territory.
The second column shows the median rating for each category – essentially the midpoint with half of individual ratings above and half below this level. Looking at the two columns, they show the same relative considerations, but with a few differences that might change our perception. For example, we can now see that while the average view of pricing is that it remains ‘Good’ and therefore less of a concern for providers, half of our survey sample view the pricing policies of their service providers as ‘Satisfactory’ at best.
Table 1: Global Category Scores
Asset Safety and Risk Management
Cash Management and FX
Regulation and Compliance
Table two provides a different perspective again. These ‘box and whisker’ plots provide a quick “on the whole” view with an idea of outliers. Glancing at this chart, one can quickly see that there is a fairly wide spread of opinion about Securities Lending even among the majority of respondents and outliers at both ends of the valuation, suggesting that with commoditisation of many core services, this may be a fruitful area of potential differentiation to explore.
All of this confirms my assertion in an earlier opinion piece that while maths is like baking, requiring precision quantities and temperatures for the recipe to work, statistics as a practice is more akin to cooking: you see what ingredients you’ve got and try to put them together in a way that tastes okay…and passes the smell test.