The use of data to answer questions is widespread. However, the most beneficial role of data are not just the answers provided, but the questions raised by the ''new'' facts. Last week we took a look at data using a macro approach to analysis. This week we will look at the institution performance and member level analyses.
The Institutional Performance Level
For most managers, the most intensive analysis is still around performance at the institutional level. One assumption in most models is that larger organizations have competitive advantages that smaller credit unions do not. Much of the data supports this observation. For example, the largest 100 credit unions — when compared with the remaining 10,103 credit unions in 2001 — reported:
- Higher loan and share growth
- Lower expense ratios
- Higher member savings and loan balances
- Higher dividends to members
- Higher ROA’s and faster capital growth
These were not one-year outcomes. The gaps were consistent for at least the last decade.
Performance comparisons can also show where there may be efficiencies. For example, six credit union officials in the Top 100 with superior performance results for recent years appeared on the same speaking platform. Their collective performance in 2001 included share growth of 22.1%, loan increases of 13.3% and capital growth of 17.7%. The loan and capital growth rates are almost 100% higher than the credit union average.
By almost any performance criteria, these credit unions are an elite group. But within the group there are wide differences in productivity measures. For example, the loans per employee ratio ranged from a high $4.2 million to $2.9 million. Revenue per employee showed a similar gap from $502 thousand at the high end to $297 thousand at the low. The data provoke more questions, such as “How can organizations operating in the same business have such wide differences in productivity—and still report overall results that place their organizations in the 95th percentile or higher.”
The Member Level
Because credit unions are organizations of members, a great deal of investment has been made in collecting and using member data more effectively. Most credit unions utilize some kind of member information database that augments transaction histories. A number of credit unions are even extending this data gathering function on members’ behalf by offering account aggregation.
More and more analysis today is focused on member profitability, share of wallet, or even more comprehensive modeling such as the net present value of a member’s relationship. The goal is to become more tailored in pricing, lending and even fee decisions at the member level. One pricing strategy no longer fits all from either marketing or an equity point of view.
Member analysis is vital, but it often may lead to higher order questions about business and product strategy. Another example from the six credit unions above: one credit union has average income per member of $917 whereas the bottom end of the range is $720, or almost a $200 difference. With the level of interest rates the same for most institutions, how is one credit union generating almost 28% more income per member?
The gap is even wider when looking at net income per member. The highest outcome for one credit union was $188 and the lowest was $108, or only 57% as “profitable” as the leading credit union. What is the difference in business strategy?
The Addictive Power of Numbers
Working with data, especially comparative information, can be habit forming — and it should be. Collecting relevant data is not the end, but the beginning of wisdom..
A danger, however, is putting too much faith in data. For example there are an increasing number of “models” that seem to “underwrite” performance improvement outcomes. Some of these are member profitability analysis, product profitability measures or even ALM models that forecast a credit union’s net economic value (NEV) under different interest rate and cash flow assumptions.
These models are useful and can illustrate new ways of looking at data on a summary level. But all rely on assumptions. In every case these assumptions are driving the factual “data” outcomes.
Data is nothing more than a tool. Its value is neutral. Relevant data can be a good tool and empower wise users. But put data in the hands of a fool, and the outcome may just be a greater, not wiser, fool.
Last week we took a look at data using a macro approach to analysis. This week we will look at the institution performance and member level analyses.