투자회사가 대학생에게 학자금을 주고 그학생의 미래 수입을 증권으로 취득한다.
그 대학생이 취업한 후 발생하는 급여에서 일정 몫(stake)을 직접 받는데, 그 학생(피투자자)가 실업일때는 못받게된다.

미국 Purdue University에서 학부생을 대상으로 진행되고 있다는데, 보험의 원리가 흠뻑 스며들어가 있다.
결국 실업률이라는 통계가 보험의 사망률을 대신하는 것 같은데, 영어 전공자에게는 취업후 매월 급여의 4.52%를 약10년(116개월) 동안 회수하는 조건으로 1만달러(약1140만원)을 투자한다.
한편, 취업률이 높은 컴공과 전공자에게는 훨씬 좋은 조건인 2.6% 88개월이다.

출처 : https://www.bloomberg.com/news/articles/2019-04-09/college-grads-sell-stakes-in-themselves-to-wall-street?fbclid=IwAR0T2EBUWryqZ4sBhi8BTm-Ce8vwBVOOZK2Sq24F72F-4EOix8jnIH3yHhQ

College Grads Sell Stakes in Themselves to Wall Street

Instead of taking out loans, students can agree to hand over part of their future earnings in return for investment.

By

Claire Boston

2019‎년 ‎4월‎ ‎9‎일‎ ‎오후‎ ‎6‎:‎00

Amy Wroblewski

Photographer: Maura Friedman for Bloomberg Businessweek

To pay for college, Amy Wroblewski sold a piece of her future. Every month, for eight-and-a-half years, she must turn over a set percentage of her salary to investors. Today, about a year after graduation, Wroblewski makes $50,000 a year as a higher education recruiter in Winchester, Va. So the cut comes to $279 a month, less than her car payment.

If the 23-year-old becomes a star in her field, she could pay twice as much. If she loses her job, she won’t have to pay anything, and investors will be out of luck until she finds work.

Wroblewski struck this unusual deal as an undergraduate at public Purdue University in West Lafayette, Ind. To fund part of the cost of her degree in strategy and organizational management, she sidestepped the common source of money, a student loan. Instead, she agreed to hand over part of her future earnings through a new kind of financial instrument called an income-sharing agreement, or ISA. In a sense, financiers are transforming student debtors into stock investments, with much of the same risk and, ideally, return.

In Wall Street terms, Wroblewski, a first-generation college student, is more small-company stock than Microsoft. Her mother works as a waitress; her father, as a quality control inspector in a car dealership’s body shop. With a strong work ethic, Wroblewski always held down at least two part-time jobs in school, working as a Purdue teaching assistant, a Target cashier, and an Amazon seasonal worker. Showing potential for leadership—not to mention earnings—she rose to vice president of Delta Sigma Pi, a business fraternity.

Those qualities impressed a company called Vemo Education, which vets students at Purdue and a handful of other schools on behalf of potential investors. More important, perhaps, Wroblewski believes in herself and her ability to make good on the contract. “Even with all my other loans, I knew I could make it work,” says Wroblewski.

Americans owe $1.5 trillion in higher education debt, a burden that weighs down their dreams and the U.S. economy. The Federal Reserve says millennials are now less likely to buy homes than young people were in 2005, and even senior citizens find themselves still making payments on their student loans. Wall Street sees the crisis as an opportunity. College graduates on average earn $1 million more over their lifetimes. Investors could capture some of that wage premium for themselves..

Total Student Loan Debt Outstanding

Data: Federal Reserve Bank of New York

“I envision a whole new equity market for higher education in the next five years where today there’s only debt,” says Chuck Trafton, who runs hedge fund FlowPoint Capital Partners LP, which has invested in ISAs, including Purdue’s. ISA experts say they have fielded calls from some of the world's largest investment managers that are considering investing in the contracts. And Tony James, executive vice chairman of money manager Blackstone Group LP, formed the Education Finance Institute to help schools study and develop ISAs.

For now, the market for income-sharing agreements can be measured in the tens of millions, a tiny sum compared with the $170 billion in outstanding asset-backed securities created from student loans. Only some schools let outside investment firms buy a stake in students. Others seek out individual donors, mostly wealthy alumni, or use money from their own endowments.

Along with Purdue, which started its program in 2016, some smaller private schools such as Lackawanna College in Scranton, Pa., and Norwich University in Vermont are offering ISAs. The University of Utah recently announced a pilot plan.

ISAs raise all kinds of questions. How many students will lose their jobs and be unable to pay? How much should Wall Street demand as compensation for the risk? Investors typically ask for a smaller slice from students with more lucrative majors. At Purdue, for example, English majors borrowing $10,000 pay 4.52 percent of their future income over nearly 10 years; chemical engineers, 2.57 percent in a bit over seven years.

Major Decision

Estimated payment schedule for a $10,000 income-share agreement made through Purdue University in a student’s senior year, by major

Data: Purdue University, Vemo Education

Purdue set up its program to be competitive with many student loans for the typical borrower. Consider a junior economics major who needs $10,000. Through a private loan, she’d likely pay $146 a month, or $17,576 over the course of 10 years. Through an ISA, a student with a starting salary of $47,000, Purdue’s estimate for its 2020 economics graduates, would pay $15,673, assuming 3.8 percent annual salary increases. That would be a good deal. But, if she found a $60,000-a-year job, she’d have to fork over $20,010.

Financial firms and for-profit colleges have been known to prey on college students’ financial naiveté to sell them high-priced private student loans, rather than steer them toward more favorable government-backed ones. While schools offering ISAs say they will offer them only after government loans with the most favorable terms are exhausted, some students may end up again with regrets.

“There’s a level of enthusiasm that’s overstated,” says Julie Margetta Morgan, a fellow who studies higher education at the Roosevelt Institute, a think tank focused on reducing income inequality. “It’s pretty darn near impossible to say whether an ISA is better or worse for an individual.” Morgan dislikes that ISAs require arbitration, which means students give up their right to sue in court.

The last big ISA experiment—at Yale University in the 1970s—ended up as a cautionary tale. Yale pooled all borrowers, and they owed the school a percentage of their incomes for 35 years, or until everyone paid back what they owed. The idea was that graduates who ended up with high-paying jobs in finance would subsidize those who chose public service.

But many students defaulted, leaving the remaining borrowers on the hook longer than they’d anticipated. Other wealthier students exited the pools via large one-time buyout payments. The remaining students tended to be lower-income. Some stopped paying altogether. Yale ultimately bailed out the borrowers, winding down the whole program in 2001.

Juan Leon, who sells business jets for Dassault Aviation SA, graduated from Yale in 1974 with a degree in urban studies. He borrowed $1,500 through the college’s “Tuition Postponement Option.” By the late 1990s, he’d paid back $8,000. “We didn’t read the fine print,” Leon says. “It was quite, quite onerous.”

Students have more protection under newer plans. Purdue, for example, caps total payments at 2.5 times what a student borrowed, so the most successful don’t feel gouged. And students making less than $20,000 a year won’t be charged at all, as long as they are working full time or seeking work. Those who are working part time or not seeking work will only have their payments deferred, which means that they’ll owe for a longer period of time.

Purdue has arranged more than 700 contracts worth $9.5 million and closed two investment funds totaling $17 million. David Cooper, Purdue’s chief investment officer, helped to develop the program and pitch it to investors after almost a decade of overseeing investments for Indiana’s retirement system. He says the funds are drawing more interest now that the oldest contracts have over 20 months of repayment data. “We feel like we’ve got the pricing for the students at a pretty good spot,” Cooper says. “At the same time, it’s a reasonable return for the investors.”

Purdue’s early funds attracted investments from wealthy individuals, as well as nonprofit Strada Education Network and INvestEd, a nonprofit lender and financial literacy organization in Indiana. Cooper says ISAs may make most sense for socially conscious investors, but he points out that even funds seeking lofty profits might one day be interested if they can juice returns with leverage.

Charlotte Hebert, 23, who graduated from Purdue in 2017, has mixed feelings about the $27,000 she took out from an ISA to pay for her senior-year costs. A professional writing major, she’s required to shell out 10 percent of her income for the term of the deal. That’s about 2.5 percentage points more than an engineer would pay. The daughter of a teacher and a nurse, she makes about $38,000 a year as a technical writer for an engineering firm in Lafayette, Ind.

She pays investors $312 a month. “I don’t think it’s the perfect solution,” Hebert says. “I am of the opinion that in a society where most of its workers need a college education, nobody should be paying this much to be what is considered a functional member of society.”

보험계리학 전공이 미국대학교에서 가장 핫한 전공라고 합니다.

계리사 평균 연봉이 약1억2천만원( $108,658)으로 매우 높은데다가 졸업만하면 보험계리사가 아니어도 자격증 여부와 상관없이 거의다 취업이 되고 있다고 합니다.

그래서 학부때 계리사자격시험 중 일부러 시험 몇개만 붙고 나머지 과목은 취업해서 월급받으면서 공부해서 패스한다고 합니다. 거의 대학원 진학은 안하구요

출처 : http://www.chicagotribune.com/business/ct-biz-college-major-actuarial-science-20180910-story,amp.html?__twitter_impression=true

The most valuable college major? Actuarial science

미보고발생손해액...IBNR 준비금을 계산할 때 어떻게 어떤 모델을 쓸것인가 등에 대해서 정의하고 모델구성하고 예제달고 평가하고... a부터 z까지 정리해 놓았다.

출처 : http://theactuarymagazine.org/anticipating-events/

Anticipating Events

Using member-level predictive models to calculate IBNR reserves ANDERS LARSON, JACK LEEMHUIS AND MICHAEL NIEMERGJune/July 2018

Predictive models have the potential to transform many aspects of traditional actuarial practice and change the way actuaries manage and think about risk. One common actuarial task where modern predictive models are not commonly used is the calculation of incurred but not reported (IBNR) reserves. Rather, IBNR has historically been calculated for pools of members using aggregate methods that utilize high-level assumptions without any sophisticated consideration of the risk factors of the individual members within the pool. However, by incorporating these risk factors into a predictive model, there is the potential to develop an informative alternative to the traditional actuarial approach. In this article, we’ll consider how a predictive model might be built to estimate IBNR at the member level. To demonstrate its efficacy, we’ll consider a case study from the group health care market.

IBNR Defined

Let’s first define what IBNR is. Essentially, IBNR is an estimate of the amount of claim dollars outstanding for events that have already happened but have not yet been reported to the risk-bearing entity.¹ For instance, if you break your arm and go to the emergency room, you will generate a claim on that date. Until you (or your provider) report that claim, your insurance company does not know about it. However, your insurance company is still liable for the claim. In fact, the risk-bearing entity is responsible for all incurred and unreported claims like this across its pool, and so it must set funds aside in its financial statements for the estimated amount of these payments. The challenge here is obvious: Because the insurance company doesn’t even know that you’ve gone to the hospital, the IBNR reserves held on its financial statement will always need to be estimated.

Traditional actuarial methods for IBNR estimation have many flavors, but they have largely revolved around aggregate estimations for entire pools of members. One traditional actuarial method, which we’ll refer to as the completion factor method, looks at the claims already received and estimates what percentage of incurred claims are believed to already be reported. This value is our completion factor. With an estimate of the total incurred claim cost, then the calculation of IBNR is as straightforward as subtracting the claims already reported from the total incurred claim costs, as shown in Figure 1. All the science and art of this method of IBNR estimation revolve around deriving good estimates for how complete the claims are for a given month.

An alternative actuarial approach, which we’ll refer to as the projection method, is to estimate the average incurred claim cost per member with no consideration of the amount of claims already reported. This is typically done by using the average incurred claim costs per member from a time period that is assumed to be 100 percent complete (or close to complete).² With an estimate of the total incurred claim cost per member in hand, we merely need to take the difference between this value and the average amount of the claims already reported per member to get the IBNR expressed on a per-member basis. Multiplying this value by the total number of members in the pool gives us our final IBNR estimate.

The projection method is a common approach for very recent months, and it relies on the assumption that the claims that have been reported to date in those recent months are not a good predictor of total incurred claims. The completion factor method is more common in months where the claim payments are assumed to be more mature.

Why Use Predictive Models at the Member Level?

Traditional methods like the previous example are technically predictive models, but they treat all individual risks the same. The benefit of such an approach is its simplicity and tractability. However, the underlying assumption that every person in the pool has the same historical payment pattern and propensity to have incurred and unreported claims seems unlikely.

An alternative to these traditional methods is to use predictive models at the member level. One of the strengths of predictive models is their ability to take high-dimensional data sets within which to segment and attribute risk more accurately, while appropriately handling any complex relationships between our prediction and the variables the model uses to make that prediction. Instead of relying upon aggregate completion patterns, predictive models can estimate IBNR for each member directly. These member-level IBNR predictions can then be summed together into an aggregate reserve amount for an entire employer group or pool of business.

Why use predictive analytics in this fashion? The biggest potential gain is in the accuracy of the estimate. IBNR can fluctuate wildly, particularly for small groups or payers with unstable payment patterns, and any additional pickup in predictive power can be helpful in estimation. An additional drawback of traditional methods is that it can often be difficult to develop IBNR estimates for different subpopulations. For instance, suppose you work at a small insurance company and you are interested in reviewing the incurred claims by month, including IBNR, for individually insured members ages 55 to 64 in a particular geographic region. Using a traditional approach, there would be two options:

1. Develop an IBNR estimate based on payment patterns observed specifically for this cohort. This involves additional effort, and the credibility of the estimates could be a concern if the population is small.
2. Apply completion factors developed from a larger pool of members. This approach is simpler, but it can also be problematic if the underlying payment pattern for this cohort is different from the larger pool.

Predictive analytics methods applied at the member level can solve this challenge by leveraging the credibility of the entire pool of members while accurately reflecting the risk characteristics embedded within any slice of the data. By producing estimates for each individual member, the estimates can be aggregated to any desired level.

The added sophistication of member-level predictive models is not free. Generally, estimating IBNR using aggregate methods can be done in a spreadsheet application after doing some data preprocessing in a language of your choice. The minimum data requirements for the completion factor method are simply a summary of claims paid for each combination of incurred month and reported month in the historical period (known as a lag triangle). Building predictive models at the member level is more demanding. First, you need to capture all the data elements required for your predictive model that perhaps you weren’t capturing at the individual level before (demographics, geography, risk scores, etc.). Second, you need to manipulate this larger data set into a format that can be fed into modeling software. Once the data is ready, you need to actually be scoring all these members on a platform capable of making predictions using a predictive model before finally aggregating and interpreting results.

Case Study: Our Model Building Approach

To assess the potential benefits of using predictive analytics to calculate IBNR at the member level, we performed an illustrative case study from a large, multiple-payer data set for 10 different employer groups ranging in size from approximately 400 to 7,000 members. In our evaluation, we looked at the performance of two popular machine learning methods: penalized regression and gradient boosting decision trees.³

We built separate models for each incurred month. For instance, one model was strictly focused on predicting IBNR in the most recent month, while a separate model was focused on predicting IBNR in the previous month. To train the models, we included a rich variety of features, including historical payment information (by incurred month and paid month), as well as demographic and clinical information such as age, gender and risk score. We also included some “leading indicator” features that helped the model identify potential large payments that had been incurred. For instance, one of these features indicated that a member had incurred a professional claim at an inpatient or outpatient facility during a given month, yet no facility claim had been reported for that month. During a hospital visit, there are typically separate bills from the facility and from the physician (or physicians). The physician (professional) bill is often processed more quickly and is generally much less expensive than the facility bill. The presence of only the professional bill is a strong indicator that there is a large claim that is yet to be reported.

For many modern machine learning algorithms, the relationships between features and predicted values can be complex. The waterfall chart in Figure 2 is a representation of the prediction development for a single member’s IBNR estimate for the most recent month, using a gradient boosting machine. For this member, the model started with a baseline estimate of $206, but this increased by approximately $1,576 as a result of the member having a “missing outpatient claim” (as described earlier). Other features pushed the prediction even higher, including high monthly costs over the past six months and a high risk score. Ultimately, the model predicted an IBNR of $5,186 for this member.

Case Study: How Accurate Were Our Models?

To keep our case study simple, our models only predicted claims that were incurred within the three months prior to the valuation date because these months constitute the bulk of the reserve. To evaluate the accuracy of our models, we split the data into two sets: a training set and the testing set. The model was built on the training set while the testing set was withheld for model evaluation and to ensure we weren’t overfitting.

To estimate overall performance, we compared the 10 group-level models for each algorithm to two traditional methods. We then compared the predicted results to the actual IBNR for each method or model, and we calculated the aggregate error across all groups, the average absolute percentage error for each group, and the standard deviation of the percentage error across the groups. These values can be seen in Figure 3. Overall, the gradient boosting decision tree model and the penalized regression model estimated the overall IBNR more accurately and had less variation than the traditional methods. These results suggest that predictive models have the potential to increase the accuracy of reserve estimates. We also found that the member-level predictions from the predictive models generally had a 30 percent to 50 percent correlation with actual results, compared with 20 percent to 30 percent when applying the group-level completion factors to individual members. The member-level correlation statistics are more complicated to aggregate across groups and lag months, so we excluded them from Figure 3.

Considerations

Using predictive analytics for the estimation of IBNR does not mean that actuarial judgment is no longer needed. Beyond the expertise needed in crafting the models themselves, adjustments to IBNR should still be made outside the model or as offsets within the modeling process. These adjustments can include handling new entrants without historical data, claim trends, or any staffing or technological considerations that could impact the backlog of claims.

One of the most important considerations in building a predictive model is which variables to include. Most of the increases in predictive power will not come from more powerful or refined techniques, but rather from more carefully considered and richer input data. For health care, some more obvious variables to consider (when available) are age, gender, plan design and geography of the member. In addition, the temporal nature of IBNR makes the timing of when things happen a key consideration. In designing variables for the model, this should be exploited where possible. For instance, the reporting of less expensive drug claims may precede more expensive inpatient and outpatient claims, or high claims in a prior period may indicate more claims are still outstanding.

Given enough feature creation and enough volume of data, a well-crafted predictive model should be able to discern the most pertinent relationships. As an example of some possible relationships a predictive model might uncover, consider Figure 4. In the first table, we see two variables and their joint impact on the IBNR within our case study (for simplicity we are only considering the amount of unreported claims in the month prior to the valuation date and paid within the next month, which we denote L0). The first variable is the member’s average monthly claims over the past year. The other variable is the “missing inpatient” indicator discussed earlier. Similarly, in the second table in Figure 4, we see another joint relationship that can stratify risk. This time the relationship is between the claims already paid in L0 and the risk score of the member.

The values shown in each cell represent the average observed IBNR for the most recent incurred month in our training data. As we can see in each chart, these variables are all strongly correlated with IBNR, but together we can stratify the risk more accurately than we can in isolation.

Before involving predictive models in your reserving process, many practical considerations are involved. The first and foremost should be a good understanding of the problem you are hoping to solve. While we mention two possible benefits to using predictive models—increased accuracy of the estimates and more accurate IBNR attribution to individual members within the pool—these benefits may not hold in all cases, depending on the availability of data and the line of business. For a list of potential considerations, see Figure 5.

One thing to keep in mind is that member-level predictive models need not completely replace traditional actuarial methods to be valuable. In fact, the completion factor method and the projection method described are often blended in practice. IBNR estimates created by member-level predictive models can be similarly blended with any traditional approach. They could also be used not for the results directly, but instead as a way to help understand the drivers of changing IBNR values. Regardless, until enough comfort and sophistication with predictive models is established, the most prudent course of action for any actuary is to do rigorous back-testing and results monitoring before replacing any traditional methods.

Conclusion

Overall, our findings indicate that using predictive models for IBNR estimation is promising. However, our analysis is not definitive; given the volatility in IBNR estimates and the sample size we tested, further research is warranted before concluding that predictive modeling techniques are superior to traditional methods. However, predictive analytics methods need not completely supplant traditional IBNR methods to be valuable. Instead, and more likely, the two approaches can supplement and complement each other. What our analysis does suggest is that this is a productive endeavor to explore. By incorporating predictive models into traditional actuarial methods we might not find the crystal ball that we seek, but with the steady incremental improvements it allows us, we can help advance actuarial practice.

출처 : http://www.msn.com/en-us/money/careersandeducation/10-bachelors-degree-jobs-that-can-pay-more-than-dollar100k/ss-

AAzbruw?ocid=sf

10 bachelor's degree jobs that can pay more than $100K

 Actuary

Tip : To Insure Promptness

17세기 런던 로이즈 커피하우스(Lloyd's coffee house)에서 선박보험 등을 취급하던 기업가와 상인들, 선원들이 몰려들면서 대기시간이 길어지자 웨이터에게 주문한 커피가 빨리나오도록 몇 푼씩 주었던 동전으로, 신속함을 담보하는 보험료라는 뜻 

==================================

미국에서는 밥을 사먹을 때 마다 팁을 내야 하는데 얼마를 내야 하느냐가 약간 고민거리다. 물론 10~20%를 내는 것이 기본이라고 하지만, 어떤 식당은 그냥 내가 가져가서 먹는데도 팁을 적어 넣게 되어 있어서, 카드로 선결제할 때 얼마를 적어야 할지 고민이 된다. 아예 어떤 식당은 영수증 아래에 음식값의 20%에 해당하는 금액을 예시하기도 한다. 그런데, 식당이 내라는 대로 낼 수는 없는 노릇이고…

그런데 오늘 출근길에 본 무가지 신문(Philly Metro 8p, March 25, 2014)에서 Lizzie Poster라는 친구(에티켓 전문가라는 Emily Post의 몇대 증손녀란다)가 팁을 얼마 줘야 미국 에티켓에 맞는지 설명하고 있는 것이 있어서 인용해 본다. 일단 이 친구는 팁이란 것을 “누군가가 먼저 내게 좋은 서비스를 해 주면, 서비스를 받고 난 후에 거기에 대해서 고맙다고 사례하는 것”으로 정의하고 있는 데, 이런 접근방식이 참 맘에 든다.

1.      음식이 식탁으로 서빙되서 올 때는 무조건 10~15%를 서빙해주는 사람에 줘라

2.      웨이터 앞 카운터에 앉아서 식사할 때는 웨이터가 음식을 내자리에 놔주어도 팁을 줄 필요없다

3.      좌석에 안내 받아 밥을 먹는 식당이라면 15~20%의 팁을 줘라. 서비스가 형편이 없더라도 최소한 15%를 주고 나서, 매니저를 불러서 정말 형편없는 서비스(it was really atrocious service!)라고 한마디 해라

4.      술집에서 잔술 먹을때는 한잔당 1~2달러를 팁으로 줘라

5.      택시를 탈 때, 택시기사는 팁으로 20~30%를 은연중에 요구하지만 그럴 필요 없다. 15%면 충분하다.

A1 : The best. The term originates in marine insurance, where A1 denotes a vessel that is in perfect conditon, the letter referring to the condition of the hull and number to the condition of the ship's trappings.

출처 : Dictionary of International Insurance & Finance Terms, John O E Clark, 5p

영국하면 떠오르는 것은 빅벤, 대영박물관, 마마미야 뮤지컬...

그런데, 보험에 관심있는 사람은 이곳을 가보고 싶어한다. 보험을 이야기하는 교과서의 서론에 등장하는 이곳! 로이즈(Loyd's of London).

로이즈의 큰 건물안에 수많은 브로커들이 각자가 하나의 협상테이블을 차려놓고 고객과 요율을 협상한다. 옛날 사람들이 모여서 해상보험의 요율을 논하던 모습을 이렇게 근사하게 현재까지 이어오고 있다. 파리의 퐁피두센터를 건축했던 리차드 로저스가 지었다는 로이즈 건물과 그안에서 진지하게 토론하고 계산하는 브로커와 고객들의 모습에서 근대보험의 효시였던 로이즈가 여전히 세계시장을 이끌어가고 있음을 확인한다.

흔히 A1하면 제일 먼저 생각하게 되는 것은 신용평가회사 무디스의 신용등급 체계에서 있는 A1이다. 그러나, A1이라는 등급을 처음 도입한 것은 신용평가사가 아니고 정육점도 아니다.

약 150년전인 1870년대 영국의 로이즈가 선박보험요율을 책정하면서, 선박의 선체 상태를 A, B, C 등으로 평가하고, 탑재능력을 1, 2, 3로 구분함에 따라 시작되었다.

Mortality table : An instrument by which the probabilities of life and probabilities of death can be measured. The basis is the ratio of the number of persons dying at any age to the number of persons alive at the beginning of the year of that age. Mortality and interest rate factors enable actuaries to produce life insurance 'net premium' calculations.

출처 : Dictionary of Insurance, second edition, C. Bennett, Prentice Hall, 212p

자녀가 있는 학부모가 주말에 서울에서 편하게 갈 수 있는 곳 중에 하나가 구세군회관 옆에 자리한 서울역사박물관이다. 서울을 축소해 놓은 입체모형도 재미있거니와 듬성듬성 서울의 근현대사를 시각적으로 표현해 놓은 것들이 무척 정감이 있다.

그런데, 보험이나 연금에 관심있는 사람들의 눈을 사로잡은 낡은 책이 하나 전시되어 있었는데, 다름 아닌 1955년 "간이총인구조사보고"이다.

종신보험이나 연금보험같은 생명보험상품을 만들기 위해서는 보험에 가입하는 사람이 생존하거나 사망할 확률을 알고 있어야 한다. 이 확률을 구하는 가장 직접적이고 원초적인 방법은 해마다 각 나이별로 몇명이 살고 있고 몇명이 죽었는지를 알아내는 것이다. 우리나라 생명보험역사의 초창기였던 1950년대에 내무부통계국에서 조사했던 인구표! 연령별로 생존자수를 table로 전개해 놓은 이 무미건조한 표가 우리나라 생명보험역사에 어마어마한 기여를 했으면 좋으련만... 보험상품을 만들때 일반적인 인구통계(국민통계)가 아닌 보험가입자들을 대상으로 하는 경험통계(경험생명표)의 사용이 선호되고, 우리나라 근대의 불안정한 이 통계대신에 안정적인 일본의 인구통계를 인용하는 경향이 있었던 것이 사실이다.

그러나, 이제는 충분한 통계량, 지속적이고 안정적인 축적으로 이것 저것 많이 활용하고 있다. 물론 경험통계를 직접 이용하지만 국민통계와의 비교는 필수이다.

아무튼 우리나라는 1982년 처음으로 경험생명표를 만든 이후에 2013년11월 현재 제7회 경험생명표가 사용되고 있다. 후손들은 잊지 않는다. 해마다 한칸 한칸을 채워나갔던 선대의 지루한 노력이 가치가 있음을...

Benefits : Compensation for loss and other services provided by insurers under terms of insurance contracts.

출처 http://www.irmi.com/online/insurance-glossary/terms/b/benefits.aspx

예전 우리나라 생명보험회사 이름에 베네피트라는 단어가 들어가 있는 경우가 있었다. 베네피트라고 영어사전에서 찾아보면 첫번째 뜻이 혜택, 이득(an advantage)이다. 이 생명보험회사에 가입하면 보험가입자가 혜택을 볼 수 있다는 뜻을 가지니 얼마나 좋은 이름인가? (사실 이 회사의 이름은 동양시멘트와 미국 Benefit 생명보험 상호회사가 합작해서 보험회사를 만들면서 그 앞글자를 따서 지어진 이름이니 이런 해석과는 전혀 상관없고, 나중에 미국회사의 지분이 빠지면서 베네피트라는 이름도 빠졌다.)

아무튼, 보험에서 benefit라는 용어는 그냥 댓가없이 받는 혜택도 이득도 아니다. 보험계약(terms of insurance contracts)에 따라 사고가 났을 때 보험회사가 주는 보험금으로 해석된다. 보험계약이 존재하려면 사람들은 당연히 보험료(premium)를 내야만 한다. 따라서, 내가 보험료를 내고 그 댓가로 보험회사로 부터 받는 돈이 보험금인 것이다. 그런데, 이 보험금을 영어에서 혜택, benefit이라니...

정말 어처구니 없게 생각될 수 있다. 특히, 보험사고가 발생했는데도 이런 저런 계약상의 제한조건 때문에 보험금을 받지 못하는 사람들의 입장에서는 열불이 날 일이다.

하지만, 그냥 보험이라는 것이 왜 생겼을까를 생각해보면 보험금이 Benefit이라는 것에는 충분히 동의할 수 있을 것같다. 여러사람이 십시일반(十匙一飯)해서 어려움에 처한 사람을 돕는 것이 보험이 생긴 이유이고, 어려움에 처한 사람이 자기가 낸 몫보다 많은 돈으로 생활의 곤궁이 다소 나아졌다면 혜택을 본 것 아닌가?

생긴지 30년 가까이 된 서울의 낡은 재건축아파트 출구에 버젓이 서있는 차량 안전용 볼록거울과 이를 설치한 어느 보험회사의 명칭 광고를 보면서, 복잡하기만 한 보험용어 "benefits"와 운전자의 안전을 보장해주는 볼록거울과 닮았다는 생각을 해본다.