An Interview with Gordon Woo of RMS

AD: Your book Calculating Catastrophe discusses the dynamic nature of catastrophe risks, and you have written elsewhere about how longevity risk is also a dynamic risk.  What is the key to dynamic risk analysis--imagination and informed judgment based on multiple disciplines and scenarios?  

Gordon Woo: All the risk analysis at RMS is based first and foremost on identifying the causal structure of the underlying phenomenon.  Generating ensembles of alternative scenarios in a model that reflects the randomness of the events at hand is what we do at RMS.  The randomness is crucial to take into account with modeling.

Regarding longevity, what we are trying to do is open a window onto the future on the types of events that might take place such as advances in medical science. GASP, for example, is an acronym we have created which stands for Geroscience Advancement Stochastic Process.  Longevity is essentially a stochastic process.

AD: It seems that heuristics play a large role in forward looking longevity risk analysis.  Is longevity risk analysis currently more art than science?

Gordon Woo: Heuristics are fundamental to longevity risk analysis as scientific discovery is premised on heuristic research.  This is why it takes time to make medical discoveries--no one can tell how long it takes.

Our model and approach specifically takes into account this non-linear reality of medical progress.

A key part about having a model for longevity is to have a non-linear, meta-model of longevity that reflects the reality of how scientific discoveries are made.  Medical progress is represented by serendipity and fact is that much progress has been through blind luck and accident. This is embedded in our model and is fundamental to our process.

Regarding use of medical information, there is some debate about how medical information is used.  For example, when asking when we will see a cure for cancer, there are inevitably many opinions but no consensus among medical community.  Our view is that there is no consensus because there is such a large element of randomness to it.  

The random walk concept is a key part of our insight.  Randomness has not been properly accounted for in previous assessments of longevity risk.  Medical discovery is not linear.  It  (a realistic representation of the time frame and uncertainty around medical progress) is such a key aspect of our model and approach to longevity risk analysis.

AD: Is it possible that forward looking longevity risk will be significantly different than what is currently being priced?  If so, what are the implications for long-dated contracts that are being pricing based on retrospective analysis?

Gordon Woo: What our model consists of is a series of stochastic projections.  There is no one particular pathway to the future--it is a random walk.  

With the random walk principle, there are a number of paths for longevity increases.

Clearly some contracts will be mispriced with respect to some of these longevity scenarios.  It’s impossible to say which contracts will be affected, but looking at the ensemble of future scenarios it is clear that some will be mispriced.

The weather or natural hazard analogy is a good one as it reflects this type of inherent uncertainty.  Like weather forecasters, we come up with a range of possible longevity scenarios.

AD: Are U.S. financial services consumers being presented with a unique opportunity to lock-in pricing based on retrospective longevity risk analysis that could prove inadequate in the future?  Stated differently, is there an arbitrage opportunity available to income annuity buyers if current prices for longevity risk transfer are, in fact, relatively attractive in light of potential positive shifts in future longevity trends?

Gordon Woo: There is of course the interest rate aspect to this and rates are very low at the moment, so this could present a less than optimal timing for annuity purchases.

There may come a time in the future when interest rates start rising and those who are offering annuities may think twice about how generous they want to be with longevity-based guarantees.

It’s a very interesting situation right now with low interest rates, and will be interesting to see how the future unfolds.

AD: Could positive longevity factors that enhance resilience and life expectancy such as high levels of engagement in social and productive activities have a greater impact on longevity than reductions in negative health conditions that increase mortality?

Gordon Woo: Attention has focused primarily on frailty and the number of health conditions that people have.  The more conditions, the more frail you are and the shorter your life-span is likely to be.

Complementary to frailty is robustness.  

Resilience, however, is something entirely different.  Resilience is about bouncing back from frailty.  

One example that is telling involves an elderly Japanese lady who is frail.  What she said in an interview is what she really needed is someone to talk to everyday.  Her health depended crucially on the amount of meaningful contact she had.  

The notion of having “a reason for living or getting up in the morning” is powerful.  It would be interesting to have the “what gets you up in the morning” box on any annuity quotation.

This is a universal situation that does not receive the attention it deserves.

AD: Is anyone currently incorporating the positive longevity factors and resilience variables you have discussed into annuity pricing and underwriting?

Gordon Woo: I’m giving a workshop on the topic this coming month in Edinburgh on the 11th of November, and in January I will be giving a similar talk at the Living to 100 Symposium that takes place from January 8-10 in Orlando, Florida.  

My presentation is titled “Cognitive, Psychological and Social Drivers of Longevity.”  One of the key insights of the presentation is that cognitive engagement is critical at certain age ranges.  Cognitive functioning is an essential element in coping with some of the health problems that inevitably arise.  At advanced ages, social networking and support--having people around you--is critical.  

AD: What is the state of the art with current annuity underwriting--particularly in light of the opportunity to “capture vulnerability data at high resolution”?

Gordon Woo: This is an endeavor that one can compare with weather forecasting.  One can see the same things happening in longevity forecasting that have happened in weather forecasting.  The whole area is very positive.

AD: Which underlying cause of mortality improvement (lifestyle, health environment, advances in medical science) has the greatest potential to drive a meaningful shift in future longevity trends?

Gordon Woo: I would make a distinction between greatest potential and most likely--they are not synonymous.

Greatest potential would be in regenerative medicine--stem cell therapy.  For any baby born now, they essentially have a life of spare parts when they reach middle age and need some new organ.

In terms of most likely, there is a long way to go in terms of lifestyle and simply better, healthier ways of living.  

Smoking, drinking, a vegetarian diet, weight management are just such obvious longevity factors that can be addressed.  The potential with lifestyle changes is very real and tangible--the examples exist with certain individuals and community.  This is in contrast to the medical advancement factors which are more speculative.

AD: What are your thoughts on the potential impact of rapid technology development and, in particular, outlier longevity scenarios such as those advanced by singularity advocates?

Gordon Woo: In our approach, the way that medical progress is made involves a key element of luck and serendipity rather than some linear advance.

Also, a key limiting factor is the need to be cautious around any medical innovation with respect to potential adverse side effects.

Testing and approval of any discovery to go from lab bench to the drugstore is a key variable in the time-frame involved with medical discovery.

AD: Are longevity trends socially regressive and is this likely to remain the case in the future?

Gordon Woo: Of course the answer is yes--particularly relating to specific types of treatment.  

Inevitably results will be hugely regressive with new, niche treatments when the market is small and expensive.

Lifestyle gains need not be socially regressive, but the reality is unfortunately often the opposite.

AD: Are capital markets a critical source of capacity for longevity risk transfer?

Gordon Woo: RMS is really pioneering the way in this area and hoping to plug a gap.

Our value in the market is the independent viewpoint that is medically-based and different from many of the existing models in the market.

We will see in coming years increasing amounts of longevity risk passed along to capital markets.

AD: Thank you so much Gordon.