|
|
|
From the Readers
Proceedings: Yea or Nay?
Dear Editor:
In February's Actuarial Review, Clive Keatinge tenders our willingness to cease publication of the Proceedings. In his view, [it does] not support the CAS's pursuit of the Centennial Goals. I see no conflict there; in fact I am more inclined to the view that the Proceedings reinforce these goals. Clive asserts that few actuaries outside the U.S. read the PCAS. I imagine that just about as many actuaries outside of the U.S. who want to, read the PCAS. They are online; there are no barriers. Lastly, he claims that the Proceedings are not timely in today's fast-paced world. Perhaps they are not. However, the CAS has plenty of timely, responsive venues for its members to get their reactions, comments, and input on important current issues.
Clive suggests two alternatives to the Proceedings—The North American Actuarial Journal and the creation of a worldwide journal. In my view, either (or both) would actually dilute the intent of the first Centennial Goal, which focuses specifically on casualty actuaries and casualty actuarial science.
The Proceedings keep us focused on our art and science.
Ed Shoop, FCAS
Clive Keatinge responds:
The problem with the Proceedings is not one of access; it is indeed easily accessible. The problem is that too few outside the CAS are interested in access.
The Proceedings is our only refereed publication, and unrefereed publications such as the Forum and the Discussion Paper Program are not a substitute. Most of our current literature is now unrefereed and thus unreliable. If we expect to attain the stature required by the CAS Centennial Goal, we must address this.
As I stated in my article, the worldwide journal I suggest would indeed focus on casualty actuarial science. The North American Actuarial Journal would be an appropriate venue for papers of interest to a wider audience than just casualty actuaries.
Fortunately, the problems with our publications have not gone unnoticed. I encourage AR readers to read the report of the CAS Task Force on Publications when it is released.
A Good Thing
Dear Editor:
Regarding the rating agencies' articles about actuaries, I shared Paul Lacko's initial reaction of outrage ("In My Opinion," The Actuarial Review, February 2004), and I agree that our leadership dealt with it swiftly and appropriately. I am writing to share my view (now that the outrage has passed and I can think clearly) that this is actually a good thing for the actuarial profession.
First off, the evidence would not support the charges, so they should not harm the profession. For example, reserve deterioration, by itself, does not imply unethical actuaries, or that the profession does not do enough to advance the science. As others have pointed out, the people who have made these accusations don't seem to have any idea who members of our profession are: highly mathematical insurance experts who have dedicated their careers to advancing the body of knowledge of casualty actuarial science.
What the articles really say is how critical the work of actuaries is and has become. The insurance industry spends a lot of money on actuarial work and these articles would support spending even more. The financial condition of an insurance enterprise has a high degree of uncertainty compared to other industries (it is more dependant on future contingent events). That is precisely why actuaries are needed in the industry. Even with the best possible actuarial work, we cannot eliminate this uncertainty. As you said, no forecaster can guarantee a right answer. As a profession, we should not get defensive when these charges are brought because we know that they cannot be supported. Rather, we should use it as an opportunity to explain how complex the issues are, how much judgment and uncertainty is involved, and all the advances that top experts in the field have been making. The sentiments in the articles can actually lead to increasing the importance placed on our analyses.
Just as good actuarial work is critical to success, bad actuarial work can kill a company. That statement is not critical of the profession, but rather justification for the high positions and incomes demanded by many of our members. When we defend ourselves to unfounded charges, we need to be careful not to damage the profession's reputation by implying that actuaries are not important enough to do any harm.
In this country, people are free to charge us with whatever they want. But, upon reviewing the evidence, I am confident the public would conclude that actuaries are highly ethical professionals devoted to constantly improving the process and doing the best possible work. I, for one, am glad I work in a profession that not only is so important that it is scrutinized when the industry has problems, but also requires a great deal of judgment. While initially upsetting, the rating agencies' articles may lead to an increased reliance placed on actuarial work. This is a good thing for the profession.
Michael Dubin, FCAS
What's Next?
Dear Editor:
In her "From the President" column (The Actuarial Review, February 2004) Mary Frances Miller asks, "where is our research on how to anticipate and reflect changing economic conditions in pricing and reserving?"
Well, you don't need a whole lot of research for that. All you need to do is some low-level sensitivity testing of your process. I suspect most (if not all) thorough and responsible actuaries do this now. However, having done it, the issue is, "what do I do next?" One of the problems is that there is no adequate, accepted public venue for the disclosure of these results.
Let's say you're somehow able to determine that your estimate reflects an underlying, overall, embedded average (I know there's a problem here) rate of change of x%. So you plug in (X +/- 1)%, and re-run. Similarly for (X +/- 2)%. You get a good idea what a one percent change is worth, real fast. Then you can say something useful, like "my best estimate reflects an X% rate of change, the underlying average in the data. If the past repeats itself, there's no reason why this estimate ought not to be pretty good. However we know the past is not going to repeat itself. Therefore, I have also calculated the liability at (X +/- !)% (etc.), and it is worth Y millions" (etc.). Relatively unsophisticated users can get a pretty good idea of the sensitivity of the result to a change in the input real fast.
This communicates two aspects of reserve estimates very quickly, simply, and clearly: they are sensitive to small changes in highly variable assumptions, and for that reason (amongst others) they are highly uncertain. It lets users decide for themselves if this particular insurance product (homeowners, for example, or workers comp) is for them.
As I wrote, actuaries are almost certainly doing this right now. The problem is, who are they doing it for, and where (and possibly when) are they doing it? If any actuary is doing this solely for him (her) self, that's good, but it's clearly not nearly good enough—and I maintain this is one of the huge problems of our profession. We tend to cultivate a mistrust in our profession by pretending to know what we do not—and cannot—know.
With respect to whom it is being done for, if it's only internally for management, that's good, too, but again, that is not good enough. Management has its own agenda, which may or may not coincide with the full contingent of actuarial publics. If it's being done for outside dissemination, that's very good. Then the question tilts towards "who is paying attention?" If this sort of advice is widely available, responsive (timely), and easily obtained, then the fault lies with the user. However, I suspect this is not the case. Actuarial pronouncements on reserves tend first to stay inside. Even the Board report is protected. If they do somehow make their way outside, their distribution tends to be carefully controlled.
My preference is for us to make our work and our results—crucial to almost any insurance-related financial statement user—completely and swiftly available and easily obtainable. I urge the reserves study group Mary Frances identified to make strong recommendations in this direction.
Edward C. Shoop, FCAS
P.S. If you teach people how to do reserves, which in the simple case is largely picking appropriate and defensible loss development factors, you can always tell them, "Well if you don't like this particular answer, pick your own LDFs. I'll just write 'em up in my report as yours, not mine, and YOU can own the result." Here's a good example: "historical LDFs for 12 to 24, eight years running, were: 3.15, 3.20, 3.08, 3.13, 3.23, 3.19, 3.11, and 3.17. I tested 3.16, the (rounded) average. They produced a reserve of $515M. Management thought 2.23 was a better idea. It produced a reserve of $383M. That's what's in the financials."
Of course, you should also go on to say, "Even though I feel 3.16 is a good representative average if history repeats, if it does NOT, and we actually end up experiencing say a half point higher LDF at 3.65, here is the impact."
Him or Her?
Dear Sirs:
Notice the salutation. I did not use "Dear Ms." This letter is to inform your contributors, particularly Walter C. Wright and John Robertson, that the convention is to refer to the unknown person as if he were male, not female ("25 Years Ago in the AR" and "It's a Puzzlement," The Actuarial Review, February 2004). It is jarring to see the feminist proclivities of your authors injected into a supposedly objective article. Must I go through each word and rub out each "s" from "she" and edit each "her" to "his" or "him?" Whatever happened to our grammar school education? Does one swallow politics with one's A B C's these days?
Louis Spore, FCAS
Editor's reply:
The Actuarial Review editors use The New York Times Manual of Style and Usage as the referee in disputes about acceptable writing standards. The manual says, "He or she may be used as a last resort to avoid an unwanted assumption of maleness or femaleness in a general reference. But preferred solutions are those that spare the reader all traces of a writer's struggle. Try the plural construction...Or rewrite the sentence so that no pronoun is required.…"
This does not apply, unfortunately, to either of the items you cited. Walt Wright was quoting directly, and correctly, from an article published in the AR 25 years ago. John Robertson's puzzle in the November 2003 issue clearly indicated that the agent is female. To refer to the agent as "he" in the February 2004 issue would be ludicrous as well as wrong.
In a letter to the editors, "Dear Editor" or "Dear Editors" is perfectly acceptable. Alternatively, use "Letter to the Editors" as the subject line of your e-mail and skip the salutation.
