Actuary Professional Code of Conduct

The Real Code

 Ruark Code of Conduct

As part of the continuing education that is required for me to maintain my US actuarial credentials, I typically spend about an hour each year meditating on the

Code of Professional Conduct.  Unlike most of the regulations, practice standards, guidance notes, memoranda, presentations, and work products with which we so often must grapple in our daily work, the Code is concise, easy to read, and vital.  And therein lies its power to clearly define what it means to be an actuary, and hence, what is not an actuary.

The Code has been effective since January 2001, replacing prior versions which dated back many years.  Yet like returning to the great literary classics, subsequent readings of the Code continue to offer new insights into timeless challenges, and received wisdom to apply to new challenges.  For this, actuaries, their clients and employers, and the public owe the committee of authors an enduring debt of gratitude.

With actuarial work increasingly reliant on, and sometimes competing against, computer-based algorithms, artificial intelligence, and myriad techy buzzwords (including at my own company!), I think that the Code is more valuable than ever.  But don’t take my word for it.  It’s only four pages -- read it for yourself.  Read it for your clients and employers.  Read it for the public.  Then ask yourself -- am I living up to this high standard of conduct?  Can algorithms, computers, or other merely math-savvy people replace this professionalism?

I used to read the Code with a rote objective of fulfilling my continuing education requirements.  Now I know better.  I read and reread it as the animating spirit of our profession, so that no matter what my work is, or what technological tools (dare I say computer code) and professional judgment I use to carry it out, I am ever mindful of the high standards I must maintain to be an actuary and the importance of my work to the greater good.


Ruark Behavioral Analytics Advisory Council - next call June 23 at 2pm Eastern

Our next meeting is coming up soon.

This is by invitation only, so please RSVP if you have not done so already, or contact us if you are interested.  We plan to share 3 key things:

  • Starting the modeling process with relevant industry data, then tailoring to each company, typically leads to better assumption models than what an individual company can do using only its own data. We will demonstrate this with model factor selection, standard error terms for coefficient estimates, and actual-to-expected ratios.
  • Integration of models across behaviors is important. We will demonstrate this by showing the impact of partial withdrawal history on surrender behavior.
  • In collaboration with the UCONN Goldenson Center for Actuarial Research, we have developed a very powerful modeling approach for partial withdrawal behavior, particularly for GLWBs and the like, reflecting the impact of partial withdrawal history as well as behavioral variations between large ad hoc withdrawals and true income-related withdrawals. We will outline our process for developing these models and how to tailor them to each company, and share sample output.

Friendly reminder that our 2017 FIA industry studies and spring VA industry studies have been completed and are available for purchase, along with related assumption model development and benchmarking services.  If you are a data-contributing client and know what you would like to purchase, please contact us and we will arrange delivery and discussion. Or if you have any questions that would aid in your decision-making, please let us know. Our 2017 service and pricing summary can be found here.


Our 2017 plans for behavioral analytics

platform_ruarkbg

 

Are your assumptions informed by credible industry experience?

 

checkWe provide a powerful combination of industry- and company-level experience studies, predictive modeling, traditional analytical techniques, and expert judgment, based on seriatim monthly data since 2007, covering approximately 70% of the annuity industry.

Is your analytical framework robust as new data emerges?

checkWe work with you to customize and implement our behavioral analytics framework, with transparent linkage from experience data to assumption models, naturally suited to regular updates for inforce and new business.

Are your analytics granular enough to mitigate anti-selection and proactively manage changes in your business mix?

checkWe address the many factors of influence and their changes over time, including product and guarantee type, surrender charge period and duration, moneyness of guarantees on actuarial and nominal bases, contract size, tax status, age, gender, distribution channel and compensation structures, and income utilization and efficiency.

Is speed important to you?

checkIt is to us too. We provide the timely and immediately actionable results you need to efficiently manage your company’s behavioral risks.

 


We aim to be the platform and industry benchmark for principles-based insurance data analytics and risk management.


 

2017 VA and FIA Behavioral Modules

For each of: Options
VA Surrenders       VA Income Utilization       VA GMIB Annuitization       FIA Surrenders       FIA Income Utilization 1 2 3
Experience Studies – industry results in aggregate, along with your company results, in a detailed report with numerical exhibits covering key factors, cohorts, and dynamics
Customized assumption model – initially calibrated to industry results, and tailored to your company based on credibility techniques
Review of your current assumptions, and comparison to the customized model above
Benchmarking of your results relative to peers
Presentation and discussion with our team
Membership on our Behavioral Analytics Advisory Council

 

va-fia_timeline

 

Would you like to learn more about implementation and pricing?

Contact: 

Timothy Paris
860.866.7786


How well can you model GLWB behavior?

In collaboration with UCONN's Goldenson Center for Actuarial Research, we have developed a remarkable approach to modeling policyholder behavior for deferred annuities (particularly GLWB income utilization), with strong goodness of fit to historical data and high predictive power. All of this is fueled by industry-level data, and then tailored to each company using credibility procedures. Actuaries, A/E ratios in the 99-101% range?! Yes, it can be done. Hit us up.