Peter Nakada is a Managing Director, capital markets at Risk Management Solutions (RMS). We had a chance to speak to Peter about longevity risk, structural modeling and the RMS LifeRisks model at the recent RMS annual conference.
AD: Can you explain why using an actuarial approach or “pure math” will not work when trying to understand the future of longevity. Specifically, why are actuarial approaches unable to capture “regime shifts” if those regime shifts are present in the historical time series?
Peter Nakada: The issue involves both not capturing the regime shifts and also the ability to connect the real world to your view of risk. In other words, with the statistical model, you really can’t say what would happen if there is a cure for cancer. The statistical distribution just sort of blurs it in there. With the RMS LifeRisks model, you can ask this question.
We are now building in the ability for users to incorporate their view of risk and then play back their implied bets—they can run their own scenarios and play back the implications. The RMS model also allows the user to conduct reverse stress tests: users choose an extreme scenario and the RMS model outputs scenarios that would cause a loss of a certain percentage of the capital base.
AD: You mention that pension de-risking and deferred annuities are a sweet spot for the LifeRisks model. Can you comment on the applicability to deferred annuities? Wouldn’t the “pricing issue” come into play with deferred annuities?
Peter Nakada: The model is differentially better the further away you are from your present situation, either in time or in risk. Out of the money risk transfers, where you have to think about the tail rather than the fat part of the distribution, are where the RMS model has advantages over actuarial approaches.
AD: Why has pension plan de-risking gained greater traction in the UK than the US? Was there similar inertia in UK?
Peter Nakada: Accounting. The UK has for a longer time been forced to mark their liabilities to market in a more regular way through the income statement which is more visibly painful in an accounting sense.
AD: Was RMS involved in either GM or Verizon de-risking deals?
Peter Nakada: No. These were at the money deals. More of the action has been in the fat part of the curve. Having said that, we are in discussions with pension buy-out providers about using our model to help them price their deals.
AD: You mention that the number of life insurers or “longevity capacity providers” that do not have access to longevity risk is relatively large. Why do you think this is the case if there is a clear economic case putting longevity capacity to work? Can you comment on scope of opportunity in US?
Peter Nakada: Partly it’s regulation. There is a regulation that says you have to be the safest available provider. This limits the number of companies that can do full buyouts.
Also, the annuity market is not that big which presents a scale issue. The bias is towards big and scale. This means there are hundreds of smaller companies that are potential buyers of this risk.
One scenario is that the bigger players who are closer to full-up on longevity risk would cede to the smaller players through something like a swap.
In terms of scope, we calculate that roughly $2 trillion of the $7 trillion in defined benefit liabilities should optimally be transferred to the balance sheets of U.S. life insurers. This would take maximum advantage of the offset between life insurance and annuities.
AD: What are your thoughts on private equity players in US fixed annuity market?
Peter Nakada: They are not as efficient a holder of that risk because to my knowledge they don’t have the mortality offset.
Alternative products to income annuities, such as indexed and variable annuities, are less clean and the mortality/longevity offset is weaker.
AD: How compelling has RMS’s ability to model and quantify the mortality-longevity offset been to longevity risk capacity providers?
Peter Nakada: There are a couple of life insurers that have literally changed their top-level strategy in light of this issue. Others understand it but it just hasn’t clicked. There has been quite a bit of inertia over past couple years.
The real key is whether or not the C-level executives understand. If they delegate down, it’s too different and people are afraid. The opportunity could be as big as the rest of RMS. The natural catastrophe or “nat-cat” market is so much smaller. Out of the money or 1-in-100 risks in the nat-cat market are roughly $200 billion whereas 1-in-100 longevity loss for western pension funds alone is $1 trillion. The economics are very compelling even if the opportunity is limited to the defined benefit space.
AD: How granular is the RMS longevity model in terms of looking at detailed composition of books of life insurance business and particularly policyholder behavior?
Peter Nakada: The model operates on broad segments, such as age, gender and country. One can tailor and adjust model factors for things such as socioeconomic factors. We’re working on the more granular, individual implications.
Country-wise, the model covers the US, UK, Japan, Netherlands and Canada – but if population mortality data exist, we can develop a new country model in a matter of weeks.
AD: Any thoughts or comments on variable annuity hedges/guarantees and has RMS thought about bringing structural approach to analysis of living benefit guarantees within VA books of business?
Peter Nakada: We have not brought model there yet. It would be interesting and a good application—we just haven’t gotten there yet.
AD: What do you see as the sort of key catalysts that could expand capital markets as a source of capacity for longevity risk?
Peter Nakada: The most likely near-term catalyst would be a rise in interest rates.
There will be a lot of movement in longevity risk the moment rates go up enough for someone to write DB pension plan managers a check rather than them having to write a check (for insurance).
It’s really a funded status issue.
There could be a rush to the door if the 10-year reaches 5 percent.
AD: Thanks very much Peter.