Accurately quantifying financial risk calls for innovative approaches, says Professor Jonathan Li.
Current risk management practice can come with unpleasant surprises – the result of excessive risk-taking and a false sense of security among investors. Quantitative finance experts, or “quants” as they are known in the trade, masked the true picture of financial risk in the run-up to the 2007-2009 financial crisis using models that failed to connect to human behaviour. With his studies focused on improving the conventional measures of risk,
Professor Jonathan Li is among a handful of researchers seeking to close a significant gap between the theory and the practice of risk management in quantitative finance.
The crux of any financial risk management is reaching consensus about the tolerance for risk, whether between a money manager and an owner, a portfolio manager and a client, or a bank manager and a stakeholder. But it turns out that getting two parties to spell out what they actually mean by risk, let alone reach a consensus, is much harder than people realize. As the business press has also discovered, this problem continues to vex money managers and encourage future financial instability. So there’s an urgent need for a risk measure capable of reaching such a consensus.
“The development of a new risk measure must go beyond the conventional mathematical models,” says Li. “We are not throwing out the quants altogether; we’re saying their models don’t properly reflect people’s actual tolerance for risk. We really need to address this, or else we miss an opportunity to take the right lessons from the financial crisis.
“Instead of serving the purposes they were intended for, current risk-measurement models tend to do just enough to get by regulators. The financial industry’s position is ‘no risk, no return.’ But how do you define acceptable risk? We argue that risks should not come with so many unexpected surprises.”
What is claimed to be a risk-adverse solution may actually be a very risky one for a client, according to Li. So the critical question is how to elaborate a risk measure that reaches the limits of risk tolerance – but goes no further.
Using business analytics tools like optimization and statistical learning, Dr. Li and his colleague show for the first time that it is possible to account precisely for the risk tolerance of an investor when comparing and optimizing financial decisions. Their methodology, described in an article in the FT50 journalManagement Science, presents the opportunity to make financial decisions better aligned with individuals’ risk tolerance.
Li acknowledges that in any situation where people must make decisions based on incomplete information, there is always a role for ‘gutsy decisions.’ But they need not be fraught with risk.
“It’s not an either/or issue," says professor Li. "You can make gut-level choices and stay clear of your personal risk tolerance limits and worse-case scenarios.”