A single rogue algorithm can trigger a cascade of automated selling that crashes a stock market before a human can even see it happen. This research tackles a fundamental gap in how we understand financial risk. Most economic models treat market crashes as random events caused by outside forces—like a hurricane hitting a city. But the real danger is *endogenous* risk: risk generated from within the system itself, by the interactions of traders, algorithms, regulations, and political decisions all feeding back on one another. The Centre will combine finance, computer science, sociology, and evolutionary biology to model these feedback loops directly. If successful, this work could reshape how regulators design rules for high-frequency trading, capital requirements, and crisis management. Instead of reacting to crashes after they happen, policymakers could identify the conditions that breed systemic risk and cut off dangerous feedback loops before they spiral. The research also aims to expose how political capture and unintended consequences of regulation can themselves become sources of instability. The ultimate prize is a financial system that is more resilient to its own internal dynamics—not just to external shocks.
View original technical description
Systemic financial risk and the impact of financial risk spilling over to the real economy, is one of the greatest threats facing the UK economy. The Centre research on systemic risk (SR) will be guided by academic principles and rigour, drawn from decision sciences, sociology, evolutionary biology, political economy and computer science, in addition to our core areas of finance, statistics and economics. We would be tempted to label it the "Systemic Risk 2.0 Centre." We propose to focus on 4 interrelated themes. 1. Endogenous risk, financial policy and risk forecasting 2. Computer-Based Trading (CBT), High-Frequency Trading (HFT) and endogenous market crashes 3. The impact of rules and decision-making 4. Network risk While it is clear to any observer that the financial system is made up of individuals who in aggregate determine outcomes, surprisingly most models used assume the opposite, that risk is largely exogenous with extreme outcomes supposedly dropped onto financial markets by an outside force. The reason for this is that it is hard to model risk when it is endogenous, so most find the paradigm of exogenous, albeit large, risk to provide an acceptable balance between ease and reliability. But SR by definition involves the entire system with its myriad actors linked together through subtle and endogenous networks and reacting to endogenous news, prices, rules, laws and political processes. By applying endogenous risk and network analysis we can directly identify channels for the creation of SR via the various feedback loops inherent in the financial system. Experience suggests that endogenous and SR is ever present throughout the financial system and emerges at the confluence of different fields. The Centre research will therefore be interdisciplinary by design. A narrow research focus, based on a single discipline, is unlikely to provide much headway in the understanding of SR. Furthermore, the analysis of the behaviour of an entire system has a long history in the natural sciences. The Centre is able to draw on some of the leading experts in this field looking for a partner to applying their insights to finance. A similar natural fit can be found with computer science. Crashes through sudden coordinated selling have been endemic throughout history, but their danger is becoming more immediate with the presence of CBT as more trading is done by unthinking algorithms that follow mechanical rules at speeds making supervision physically impossible while little about the structure of markets itself is commonly known, leading to second guessing of second guesses. Such uncertainty is further deepened by the nature of the balance-sheet and trading networks which are only starting to be analysed. One of the highest concerns to society is the design of informed rules (e.g. laws, accountancy standards etc.) and regulations that prevent the build-up of such SR and allow feedback loops to be cut once they operate. Unfortunately, regulating markets is fraught with difficulties, not least because of the law of unintended consequences that pervades finance. It is tempting to argue that if any one actor acts safely, then the system altogether is safe. To go beyond this partial equilibrium based fallacy of composition forms a major impetus of our work and leads to robust regulations. Furthermore, such rules are themselves endogenous. It is impossible to understand SR without attention to the political aspects of decision-making that can either lead the authorities to be captured by special interests, stall excessively on decision-making or provide robust resolution of a crisis. In that sense, SR is significantly politically constructed as society has created complex that were all the outcomes of political struggles and compromises. Through our close relationships, the output of the Centre directly impacts policy making and supervision, and the Centre in return is itself informed by policy makers' needs and interests.
Christian Julliard (Co-Investigator)Eva Micheler (Co-Investigator)Jean Pierre Zigrand (Co-Investigator)Jon Danielsson (Principal Investigator)Julia Black (Co-Investigator)Kathy Yuan (Co-Investigator)Philip Treleaven (Co-Investigator)Robert Hancke (Co-Investigator)Ronald Anderson (Co-Investigator)Tomaso Aste (Co-Investigator)
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know