Credit risk analytics is a technical discipline and the temptation to recruit analysts purely on the strength of their technical expertise is often overwhelming. However, ‘accurate analytics’ is not always the same as ‘value-creating analytics’. Accurate analysis must be combined with sound business strategies before real value is created. So, if your’s is an organisations wanting to implement the next generation of value-creating analytical techniques – profit model analytics, test-and-learn analytics, etc. – it is important to hire analysts that posses technical skills as well as a deep understanding of the business environment.

The best way to identify candidates with this mix of abilities is to make case studies an integral part of the recruiting process – particularly case studies based on the profit model. Case studies should form part of a larger three-pronged recruitment strategy which should determine the candidate’s compatibility with the organisation’s culture, the candidate’s technical ability and the candidate’s business understanding.

The first two prongs of the strategy are achieved using traditional recruitment techniques. Each organisation’s culture is different and so those parts of the recruitment strategy designed to test for cultural compatibility will vary from organisation to organisation. Usually though, competency based interviews and one-on-one discussions with existing team members will suffice. The candidate’s technical abilities can be determined by a thorough analysis of their resume and, potentially, a series of mathematical tests. These steps are vital and case studies should be a compliment to them, not a replacement.

Profit model case studies, then, are designed to test the candidate’s ability to apply analytical techniques and business insights to solve a problem. Because technical competency is proven separately, the design of these case studies should emphasise the logic of the profit model above mathematical complexity. The standard profit model case study follows a simple template. The case will always start with a brief introduction to a business scenario which, to put candidates at ease and to emphasise the fact that the case is not testing for pre-existing banking knowledge, should ideally not be banking related. Some common scenarios include selling second-hand golf balls, operating a passenger ferry and delivering pizzas.

The first questions should be kept general and should test a candidate’s breadth of thinking and their ability to work with ambiguous and/ or limited information. A candidate must be able to identify key profit levers in the business and come up with logical ways to measure and manage those profit levers. The candidate should also be able to estimate reasonable values for one or two of these measures in an environment of limited information.

Consider a case that deals with a business selling second-hand golf balls. The candidate would be asked to identify those ratios which they would measure to determine the desirability of such a business. What they are actually being asked to do is to identify the key profit levers in the business – number of balls, price of balls, cost of retrievals, etc. Once they have identified these profit levers, the candidate should suggest a logical way to estimate, for example, the number of golf balls in a particular water hazard – a function of the age of the course, the number of players, the likelihood of hitting a ball in that hazard, etc. At this stage it is common for a strong candidate to already be showing signs of a logical thought-process. However, it is the numerical questions that follow that most clearly differentiate candidates.

These questions test for two critical abilities: the ability to construct an equation and the ability to manipulate an equation. An equation is a numerical representation of a logical thought. In this case, the equation being constructed is a profit model. Candidates should be provided with the values for key profit levers which they should then use to determine the current level of profitability for the pertinent business.

Continuing the example, a candidate could be asked to calculate the profitability of the second-hand golf ball business assuming there are 5 000 balls in a particular dam on the local golf course which can be sold for a dollar each but, in order to be allowed to retrieve the balls, there is an obligation to pay a royalty to the club of 5% of total sales and to pay a diver a fixed cost of five hundred dollars per retrieval.

There are two approaches a candidate can take to solve this problem. The first approach is to construct an equation. In this case profit is equal to revenue (sales price multiplied by the number of balls sold) minus variable costs (the cost of the royalty) and fixed costs (the cost of the dive). Populating and solving this equation quickly reveals the answer. The second approach – reminiscent of accounting formats – is to calculate each component separately before combining them at the end. There is nothing expressly wrong with this approach and most candidates will still get the correct answer. However, candidates who think and work in equations will almost always do much better in the more difficult questions that follow.

The second set of numerical questions should oblige the manipulation of equations. These questions provide the most insight into a candidate’s ability to visualise a business problem in terms of a dynamic numerical relationship between various profit levers – i.e. to visualise a profit model. In these questions one of the factors in the original equation should be adjusted and the candidate asked to calculate the implications of that change. A candidate might, for example, be told that the royalty is set to increase and be asked to calculate the maximum level this royalty could reach before the business made a loss or they might be told that prices are dropping and be similarly asked to calculate the break-even selling price. By seeing business problems as a series of dynamic numerical relationships which can be represented and analysed using equations, strong candidates prove themselves comfortable with the concept of profit levers and profit models even though they may be unfamiliar with those specific terms.