RMS Model Provides a New Perspective on Longevity Risk

Risk Management Solutions (RMS) is well known as a leader in the area of catastrophe modeling and analytics.  The company provides its services to a broad range of insurers, reinsurers, consultants and capital markets participants who are active in the property and casualty insurance industry.

More recently, the company has developed resources that focus on life and health risks.  Though initially focused on mortality risks, RMS recently launched a longevity risk model.  The company was instrumental in the first longevity bond offering that recently came through Swiss Re.

We had the opportunity to discuss the longevity model with Andrew Coburn who is the Senior Vice President of the Life & Longevity risk modeling team in the London office of RMS.

Annuity Digest: Can you provide some background information on the longevity risk-related activities within RMS?

Andrew Coburn: We had some initial success with pandemic influenza and infectious disease modeling, and a few years back we had a big break with a Swiss Re mortality bond.  Many clients now hire us to model their excess mortality risk.

We then had clients suggesting that longevity modeling is actually their biggest challenge, so we began development of a longevity model around three and a half years ago.

The model development has taken a bit longer than expected because this stuff is complicated.  The work required our research team to understand a range of different causes of death and the drivers of mortality change—what we now refer to as “vitagions.”

RMS now has a medically-based specialty team of about a dozen people.  Most of this team is in the San Francisco Bay Area and some are in London.  The group members have a blend of backgrounds and skill-sets ranging from epidemiology, and mathematical biology to qualified physicians and financial modelers. 

Annuity Digest: RMS seems to believe that rates of mortality improvement over past several decades will abate somewhat over next couple of decades, correct?

Andrew Coburn: Most actuarial models today assume the recent high levels of mortality improvement are not sustainable.

RMS has an attenuation of mortality improvement like most actuarial models.  Further out, though, RMS projects more of an uptick in mortality improvements than most conventional actuarial models.

Annuity Digest: The bearish longevity view seems somewhat contrarian and a bet against what many would consider structural trends that are both exponential and inherently uncertain (i.e. the pace of technology development).  What are the key factors that support the view that current perceptions of longevity risk are overestimated?

Andrew Coburn: What we basically are doing in the model is showing the constraints that a new technology need to go through in terms of their time to market impact.  For a new treatment to be brought onto the market it requires approval by the Federal Drug Administration – this can take 15 to 20 years before physicians start thinking about prescribing it to their patients.

It is entirely possible that new techniques will be developed.

Annuity Digest: Can you comment on how the model methodology differs from a conventional actuarial approach?  How are the projections that are being made different from simply projecting based upon past statistical trends?

Andrew Coburn: Historically, most mortality analytics have looked at past mortality data and projected it forward.  We would characterize this as a backward looking, more trend projection modeling.

What RMS has developed is essentially a meta-model of medical advances.  Most of the value of the RMS model is helping people see the realistic constraints on mortality improvement.

We use precedents from past medical development to project future development.  For example, we would look at how long the development of antibiotics took to have an impact on the general population.  A more recent example is Herceptin, and our focus would be the amount of time involved from lab to testing to market, etc.

We have tracked a whole series of “precedents” or how long medical technologies have taken to come to fruition.  The distribution is driven by time—specifically the speed at which new mortality improvements may evolve from technology improvements.

Annuity Digest: Why is the model better at predicting medium and long-term dynamics of mortality change?

Andrew Coburn: If you trend recent historical mortality, the persistence of recent experience is relatively short and the uncertainties become relative large if you go more than a few years out.

We have added information to constrain these uncertainties.

We relax the constraints in the modeling far more as one goes further out in time to mitigate the uncertainty of making more far reaching projections.

Annuity Digest: Is the model calibrated largely to measure the likelihood and impact of extreme longevity scenarios?  If so, can clients use the model to manage more benign and likely scenarios as well?

Andrew Coburn: RMS has a strong background in catastrophe risk, so we naturally focus on the tail of the distribution.

Clients have an interest in expected outcomes—driven in large part by the need to determine prices in the swap market, so we’ve had to develop detailed views of this aspect too.

Annuity Digest: How does RMS account for the risk that the current set of vitagions or individual sources of mortality improvement do not contemplate future developments that may prove meaningful?  For example, what if a future mortality factor falls outside of the areas of cancer treatment, regenerative medicine or ageing science? 

Andrew Coburn: We have made our classes pretty broad.  For example, retardation of ageing captures “whacky” or out there possible improvements and the uncertainties are pretty large around this.  Retardation of ageing is our catch-all for the “unknown unknowns.”

We are constantly monitoring the relative change in technologies as they come through.

Annuity Digest: Do any of the vitagions have a greater impact than others?

Andrew Coburn: Yes, medical intervention is the big, primary category for the moment (e,g, smoking, cardiovascular disease treatment, cancer treatment).

We think the whole category of regenerative medicine will play-out in a very meaningful way over next 2-3 decades.

Different vitagions are more important for different eras.

Annuity Digest: Will the list of vitagions evolve and expand?

Andrew Coburn: Yes, most definitely.  We are constantly monitoring and tracking.

Annuity Digest: Can you comment on Kortis Capital and what you believe to be the potential of capital market capacity for longevity risk as well as longevity as an asset class?

Andrew Coburn: We are proud to be part of the Kortis transaction and to be a modeler of first successful longevity bond. 

There is a huge amount of interest in longevity risk as a new investment asset class.  I think there will be large demand for transactions of this type. 

At the moment, longevity as an asset class still requires some pretty significant spreads to make investors comfortable.

There is clearly lots of supply (of longevity risk) in pension schemes, and many people are attempting to de-risk in an attempt to optimize corporate balance sheets.

I think it will be a big area.