Harald Stieber, Economic Analysis and Evaluation Unit, DG FISMA, European Commission
The financial crisis of 2007/08 was not caused by complexity alone. It was caused by rapidly increasing financial leverage until a breaking point was reached. While the mostly short-term debt used for leveraging up consists of “run-prone contracts“, the precise location of that breaking point had to be discovered in real time and space rather than in a controlled simulation environment. Also, the complex dynamic patterns that emerged as the crisis unfolded showed that little had been known about how an increasingly complex financial system would transmit stress. The sequence of markets being impacted and the speed of risk propagation across different markets and market infrastructures was not known beforehand and had to be discovered “on the fly”. Our ignorance with respect to these static and dynamic properties of the system reflects deep-rooted issues linked to data governance, modelling capabilities, and policy design (in that order).
From a policy perspective, the crisis revealed that several parts of the financial ecosystem remained outside the regulatory perimeter. As a result, the public good of financial stability was not provided any longer to a sufficient degree in all circumstances. However, the regulatory agenda that followed, under a principles-based approach coordinated at the level of the newly created G20, while closing many important regulatory gaps, also created increasing regulatory complexity.
Regulatory complexity can also increase risks to financial stability. Higher compliance cost can induce avoidance behaviour, which makes financial regulation less effective as regulated entities and agents will engage in regulatory arbitrage as well as in seeking to escape the regulatory perimeter altogether via financial innovation. Until recently, at least the largest financial institutions were considered to “like” regulatory complexity.
However, the perception of complexity in the financial industry is changing. Complexity cannot be gamed any longer as compliance cost and risk of fines have increased. One of the clearest statements in that direction came in the form of a letter from financial trading associations that we at the European Commission received (together with all main regulators) on June 11 2015. In their letter, the associations called for coordinated action in the area of financial (data) standards that would reduce complexity to a level that could again be managed by the sector.
The European Commission’s Better Regulation agenda has at its heart the principle that existing rules need to be evaluated in a continuous manner to assess their effectiveness as well as their efficiency. Under this agenda, the Commission launched a public consultation in 2015 calling on stakeholders to provide evidence on 15 issues with a strong focus on the cumulative impact of financial regulation in place. The purpose was to identify possible overlaps, inconsistencies, duplications, or gaps in the financial regulatory framework which had increased considerably in complexity. The area of (data) reporting emerged as a major area where responses pointed to important possible future gains in regulatory effectiveness and efficiency.
Regulatory reporting has seen massive changes as the lack of relevant data at the level of supervisory authorities had been identified as a major source of risk during the crisis. Especially, legislation in the area of financial markets such as the European Market Infrastructure Regulation (EMIR), but also MiFID/R, employed a different approach to regulatory reporting compared to existing reporting obligations for regulated financial institutions (e.g. COREP, FINREP). EMIR puts the focus on the individual financial transaction (of financial derivatives traded over-the-counter rather than on a regulated exchange), with reporting at the most granular level of the individual financial contract. Reporting under EMIR started to be rolled out in several phases from February 2014 and is still ongoing, starting from the most standardized contracts and continuing to the least standardized ones. This approach is extended to a broader class of instruments under MiFID/R.
This granular approach to regulatory reporting holds tremendous promise from a complexity science perspective. It could, at some point, allow the mapping of the financial ecosystem from bottom-up, as well as further the development of a Global Systems Science policymaking process. However, to arrive at more evidence-based, data-driven policies, data governance, and more precisely financial data standards, will have to be adapted to the increasingly granular data-reporting environment.
Data governance requires robust financial data standards that keep up with technological change. We see a few precise implications at this stage what standards need to do in that respect. Financial contract data is Big Data. Financial data standards produce small data from Big Data. They add structure and scalability in both directions.
In a follow-up project to the call for evidence, we are therefore looking at different ways how financial data standards and regulatory technology can help achieve Better Regulation objectives. These possible ways comprise the definition of core data methodologies, the development of data point models, exploring the use of algorithmic standards, as well as possible uses of distributed ledger and decentralized consensus technologies. We cannot say at this stage if the vision of a “run-free financial system” is within our reach in the medium-term. But the resilience properties of the internet are one possible guide how technology could help regulatory reporting achieve its objectives in a much more powerful way in the future that will at the same time acknowledge the complexity of our subject matter.
OECD-EC-INET Oxford Workshop on Complexity and Policy, 29-30 September, OECD HQ, Paris: Click here to register
 Effectiveness: Does the impact observed on the ground correspond to the outcome aimed for by the EU co-legislators?
 Efficiency: Is the desired regulatory outcome achieved at lowest possible compliance cost?