By Ron Rimcus, CFA*
The case of Jacobellis v. Ohio went before the US Supreme Court in 1963. The question the nine justices were asked to address: Did a risqué movie qualify as obscene?
In the decision delivered the following year, Justice Potter Stewart famously rendered his opinion on pornography, stating simply, “I know it when I see it, and the motion picture involved in this case is not that.”
With so much talk of systemic risk in recent years, there is surprisingly little consensus on what systemic risk actually is. Yet many people have strong views on the topic. Investors, regulators, bankers, government officials, and others are guided by their gut feelings — that is, like Potter Stewart with pornography, they know it when they see it.
But does it have to be this way? Much of the work performed on systemic risk thus far employs models that require a stream of time series data, usually focused on the pricing of various market assets. For instance, Rodney Sullivan, CFA, Steven Peterson, and David Waltenbaugh develop a sophisticated statistical model that focuses on extreme losses and market liquidity, volatility, and default risk.
Unfortunately, they neither ask nor answer the question, What happens if these are the wrong variables to monitor? Further, what happens if they are the right variables, but the market prices them incorrectly, as it did mortgage-backed securities in the mid- 2000s?
Simply fitting a model to a given sample of data only introduces the bias that the conditions present during the sample period will be the same in the future. So, as time goes by, the model will be less and less reliable (assuming it was reliable to begin with).
All the shortcomings of our models notwithstanding, they haven’t stopped the government from monitoring and regulating systemic risk. As Mark Van Der Weide demonstrates, Dodd-Frank financial reform legislation requires regulators to identify and mitigate systemic risk. But what happens when policy itself creates systemic risk, as it did in the 2008 crisis? Can we really expect the government to identify the government as the problem? And if the government can’t identify the problem, can we expect it to identify an appropriate solution? And with vast differences in approaches to modeling and monitoring systemic risk, how can we trust that the regulators got it right?
So often in the investment business, we look for answers in quantitative models. Systemic risk is 19.2 — time to hedge! Systemic risk has fallen to 7.9 . . . Phew, we can all breathe easier now! Alas, if only it was so simple. There is a quote, often and perhaps erroneously attributed to Albert Einstein, “Not everything that can be counted counts, and not everything that counts can be counted.” Apocryphal or not, it’s true in all walks of life and certainly true in evaluating systemic risk.
For instance, what if a given country starts to abandon the rule of law? What if monetary policy leads to major misallocations of capital? What if the data you need isn’t captured in time series format? What if the buffers that exist within the financial system are slowly whittled away as players (both public and private) in the system push for growth only to find themselves approaching the end of the runway in time? What if the public fails to recognize these changes and hence is absent from security prices? These examples do not lend themselves to time series data. Nevertheless, they are enormously important.
Fortunately, these conditions can be analysed. All we need is a framework for thinking about it.
Systemic risk is the potential for a large-scale failure of a financial system during which providers of capital (depositors, investors, capital markets, etc.) lose trust in either the users of capital (banks, borrowers, leveraged investors, etc.) or in a given medium of exchange (e.g., US dollars, Japanese yen, gold, silver, etc.).
Perhaps the most important component of systemic risk is contagion: It can be passed from unhealthy to otherwise healthy institutions through a transmission mechanism. If it were not for the potential for infection, then the risk would not be “systemic” at all and we would only deal with the problems at the “micro” level institutional. The transmission mechanism is defined by the combination of leverage, interconnectedness, and safeguards in the financial system. The various players in the financial system each have exposure to the three components of systemic risk as noted in the matrix below:
Systemic Risk Matrix
As noted in the table, not all of these variables lend themselves to numerical data, but when assembled as a whole they tell a compelling story.
All of this begs the question: How big a threat does systemic risk pose to markets today?
We polled readers of CFA Insitute Financial NewsBrief to find out. Of the 635 respondents, 47% indicated that systemic risk is high or that a systemic crisis is imminent. A full 45% said that systemic risk is moderate (i.e., medium), while only 8% indicated that systemic risk is low.
Though there is not a single conclusive systemic risk model, respondents are expressing an uneasiness about the current state of affairs.
Intuitively, they seem to know without necessarily knowing precisely why. A framework, like that presented above, will help you connect the dots, so that when you evaluate systemic risk, “You’ll know it, when you see it.”
This content was supplied by CFA Australia/New Zealand. It was orginially posted here.