Originally published: Winter 2018 RC02 Newsletter
The story behind the financial crisis is relatively well known. By cutting-up and repackaging mortgage risks, banks could make more profit. At the same time, they amplified their financial losses which expanded to financial institutions all over the globe. Ten years after the global financial crisis, a new one is never far away but where will the risks come from next time? In order to know that, as economic sociologists we need to know more of the object itself: financial risks.
Capital requirements, internal risk assessments and stress tests all give an image of a world in which risks are controlled. The capital is a buffer to accounted losses, internal risk assessments would determine if investments follow the risk appetite of an organization, and stress tests would show the resilience of financial institutions and overall financial stability. What is behind all these numbers? In this short piece, I argue in favor of ethnographic research on financial risks. The environment of the object’s creation makes direct observations the only way to fully understand them.
Ethnographies or the methods of observations have the ability to go beyond representations. Both quantitative data and less intrusive qualitative data can hide the inner-workings of financial risks. Aggregated numbers found in balance sheets and supervisory risk reports are the result of many a discretionary choice taken by the makers of these figures. Similarly, even though to a lesser extent, interviews and document analysis both lack the ability to obtain information about the exact process. They remain representations of either the interviewee or the writer of the document. On the other hand, observations, participant or not, allow for access to the details of the creation behind the risk representation.
Risks are not only economic or technical objects; they originate from specific social and organizational settings. While accumulations of risk numbers, balance sheet entries and communications to supervisors, through for example stress tests, all give a representation of risks, they lack the understanding of their creation. Ethnographic observations with an understanding of the local knowledge structures, such as economic sociologists have known to do, are the only way to understand financial risks. The reasons to understand the details behind the risks are threefold. There is the possibility of involved actors to lose face, the malleability of financial figures and their creation within the organization.
First of all, let me explain the aspect of losing face. All actors involved in the risk creation can lose face if they show the difficulties of the process. The outcome of the risk namely determines the survival of the organization. Therefore, everything that inhibits that presentation of health can lead to a bad reputation. For an organization to openly report its risk practices, including the negative events or difficulties, incentives need to be very high.
The creation of risk numbers inside financial organizations is not only crucial to investments; the organization’s life depends on them. Regulators, investors and accountants all demand compliance to their expectations in both reporting and financial results. All these actors have the possibility to bring down the organization. Ample examples exist, for instance through the European supervisors “Failure” or “Likely to Fail” decision or by pulling out investments. Risk creation is thus organizationally sensitive, making it likely to be kept under relative secrecy.
Not only can the organization suffer, so can the creators of risks, the risk managers. I call them the creators because they are, in the final instance, the people in financial organizations who make the risk assessments, allowing them to exist. The risk managers have the responsibility to avoid negative events for the organization. If such an event would actually take place, they would not necessarily do a good job. Therefore, to openly speak or report about such negative possibilities would entail a direct loss of their reputation. Thus, one cannot directly expect the actors that make financial risks to openly speak about the difficulties they have in the creation of the numbers.
In my research on financial market risk managers, the difference between interviews and observations was extremely striking. The interviewees in semi-structured interviews explained their work by emphasizing their successes at limiting problems and giving general descriptions of processes. Even the more candid interviewees did not show the nitty-gritty decisions that made their risk assessments. The observations, on the other hand, showed the messiness of the creation of the numbers. None of this fit the expectations of a rational, impartial and well-thought through risk measures. By explaining to me their processes, and being involved myself, I saw the difficulties of the risk managers to make their assessments and have them stick.
Secondly, financial risks are part of the financial reporting and market practices. Critical accounting studies have shown the ability of actors to change accounting figures (Vollmer 2007). The term used is malleability, whereby the number adapts to the circumstances in which it is created. Malleability initially meant the ability to let numbers change. Accounting in that sense is not a fixed set of values, the use of the rules is fluid.
Malleability is crucial in the understanding of financial risk that both allude to accounting values and financial market estimates. One can relatively easily change the assumptions of a model or the accounting classifications. Therefore, the assessment as represented is only a reflection of the environment in which it was created at the time.
When I sat behind the desks of the risk managers, they showed me the small decisions they made in their assessments. From running a computer program in a different way, changing the data because the numbers did not make sense for interpretation or by changing accounting categories for costly assets, all of these choices fell inside the rules yet could completely change the final representation.
Not only did the (smaller) choices of the risk managers lead to changeable risk assessments, (larger) choices made in and between teams led to a similar malleability of the figures. I was able to sit in meetings and write down the interactions behind the final creation of risk numbers. Discussions ranged from contestations on the use of software programs to the necessity to obtain a very specific outcome.
Organizational Environment of Risk Creation
The environment in which the risks are created is the third reason for the need for observation data. Namely, the organizations in which financial risks are created have their own local logics. While the borders of organizations are porous, they still exist. The actions that people take beyond them, hidden from sight, make the financial market risks.
While the technology of making a risk number might seem relatively simple, their organizational dependence is not. Local rationalities, employees who are ‘making out’ and the legitimacies created in a team all make these risk assessments (Burawoy 1979, Friedberg 1997). Since these situations are behind the doors of the organization, to know them means going beyond those same doors. Either by participating or by simply observing, the local rationalities can be studied rather than just their representation.
The meetings I attended are a great example of these local rationalities. Take for example a meeting on the risks related to derivative models I was able to observe at Bank F. One of the risk managers, Janice, had suggested some improvements to the models. The theoretical discussion was supposed to be one between models based on historical market data or expectations of how markets react. However, the underlying discussion was based on one simple argument: that the three men who also worked on the models, in risk management and elsewhere, disliked Janice. They even acknowledged this. Therefore, Janice’s suggestions only encountered negative criticisms from her counterparts. Finally, the outcome of the discussions and the proposed changes to the methodologies reflected this social stand-off. To summarize, the local rationality of dislike of one person determined how Bank F valued its derivative models. The risk creation thus directly depended on the interactions of the people on the inside of the organization.
Entering the Financial Organizations
Many will say that access to financial organizations is difficult. That is true. It requires convincing power, generally a trade of some sorts and most importantly the skills necessary to comprehend the technicalities of the world. However, bankers and financiers are not the boogeymen we easily think they are. Similar to any other setting, the multiplicity of people also leads to a diversity of ideas and interests, who can also relate to the quests of a sociologist.
In my case, it took nine months of talking to bankers and insurers to find an entry in an organization. That road to the observation included multiple dead-ends. Take for example the time where one of the advisers to the CEO of a large European bank promised me a couple of weeks of observations in the trading room. Sadly nothing came of the promise. However, I just needed one location to observe but ended up with two. Chance was involved but also persistence and an ability to convince the people inside of the organizations on the relevance of my work.
While the above only restates that access to finance is difficult, it is possible. Not only did I manage to obtain it, so have many other social scientists. If we fully want to understand the financial world, as I believe, we need to know what goes on in the darkest depths of these markets. Not making the effort would mean we seem to give up one of academia’s ultimate goals: to know more about the world in which we live. Given finance’s clout on the rest of the economy and society, it is time we fully understand its risks. Without observations, this becomes difficult.
While many steps have taken to understand the behemoth that is global finance, we nearly do not know enough. Especially the empirics lack on the practices of financial actors, including financial organizations.
Burawoy, M. (1979). Manufacturing consent: Changes in the labor process under monopoly capitalism. Chicago, IL, USA: University of Chicago Press.
Friedberg, E. (1997). Le pouvoir et la règle. Dynamiques de l'action organisée. Le Seuil.
Vollmer, H. (2007). How to do more with numbers: Elementary stakes, framing, keying, and the three-dimensional character of numerical signs. Accounting, Organizations and Society, 32(6), 577-600.
Anne EA van der Graaf defended her PhD at Sciences Po, Paris, France last September. Her research focuses on risk management of financial markets, insurance, and banking. Her latest publication is The risk fluctuation "The consequences of avoiding interest rate risk", in the collected volume The Making of Finance, edited by Isabelle Chambost, Marc Lenglet and Yamina Tadjeddine. The book came out in 2018 and gives an overview of the French social studies of finance.