Actuarial news and views from Cape Town and beyond

Leave a comment

Predictive Modelling


Predictive modelling refers to the development of advanced statistical models to analyse, extract insights and make inferences from large data sets. Statistical tools are used to separate organized patterns from random noise. Predictive modelling is used extensively in the actuarial field and generalised linear models (GLMs) are perhaps the most popular. GLMs are only one aspect of predictive modelling though. In recent times, the amount of data available to actuaries, coupled with better computing power has opened up new opportunities to build more accurate models for predictive purposes. Other, more sophisticated forms of predictive modelling techniques like the use of neural networks, machine learning, classification and regression trees (CART) and statistical clustering can now be employed by actuaries. The areas to which these techniques can be applied are diverse. These range from price setting to improving marketing efficiency.

These techniques however should not be used in isolation. Without an expert knowledge of the underlying business, the results of these models could be misleading.




Leave a comment

Data, Data and More Data

Current computing capabilities and a proliferation of data sources has unearthed new avenues for insurers to do business. For example, the current use of credit scores to underwrite car insurance and the use of social media to glean behavioural insights are the beginning signs of a revolution in the industry. The effectiveness with which new sources of data are explored and the level to which this data are exploited will be the key to success and profits in the future. More importantly, as competition between companies soon shifts to the realm of data mining, the ability to extract useful insights from the data could ensure a company’s survival.

However, to fully utilise these data, organisations themselves will need to change. This article mentions a few interesting ways in which data are currently being collected, proposes how this volume of data can be managed and suggests a framework for company transformation. It also makes some comments on how these data can be incorporated into models.

Leave a comment

Data Mining in Actuarial Science

There is nearly unlimited amounts of data out there these days. Getting through all of it is a major issue all companies in the actuarial profession face. However this is changing. It is becoming increasingly accepted that the volume of data available to a company is less important than the actual models being used to analyse the data. More important than volume of data, is the ability to extract insight and application value form that data in an accurate and timely manner.

Data are being used in all aspects of the actuarial profession. The original uses being risk elimination and the expectation of resource demands. However, data are now also being used for fraud detection, cyber security and analysing the response rates of marketing campaigns.



Leave a comment

Too Big to Fail – When State Benefit Schemes Become Insolvent

Social Security is an insurance programme consisting of Old-Age (pension), Survivor and Disability (OASID) insurance. It was originally signed into law in 1935 by President Franklin Roosevelt as the Social Security Act and has subsequently undergone several amendments to include other social welfare programmes, nonetheless the core benefits remain OASID.

Social Security is funded by federal payroll taxes paid by employees and their employers. The taxes are collected by the Internal Revenue Service (IRS) and entrusted to the Federal Old-Age and Survivors Insurance (OASI) Trust Fund, the Federal Disability Insurance (DI) Trust Fund. The trusts are used as an internal accounting measure for the federal government and as the accumulated holdings of the Social Security programme.

The underlying assets which are purchased with the taxes are special issues federal government securities. Special issue in the sense that they are always redeemable at any time at par and they can be issued at any amount. The tax revenues exchanged for the federal government securities are deposited into the general fund of the U.S. Treasury and are indistinguishable from revenues in the general fund that come from other sources. This highlights the importance of the trusts as an accounting measure as the trust fund balance represents the amount of money owed by the general fund of the U.S. Treasury.

The structure of social security can be considered as a DB fund as the benefits are calculated according to a person’s lifetime earnings (formula can be found in Although Social Security is often viewed as a single programme, its financing comes from two legally distinct funds. As the law currently stands, the OASI and DI funds are legally distinct and do not have the authority to borrow from each other.

Members of Congress and the public have raised concerns over the solvency of the DI fund from which disability benefits are paid. Since 2005, total expenditures have exceeded non-interest income and as of 2009 they have surpassed total income (including interest). According to projections from the Congressional Research Service, the fund will be exhausted by the end of 2016.

This did not cause much concern in the past as Congress approved temporary cash infusions to increase reserves of depleted funds. This was used as a short term measure to give lawmakers time to develop and implement longer-term solutions. However, given the current political and economic state in which the U.S. finds itself, there is genuine concern that this option is no longer viable.

Given that the OASI fund will be solvent until 2036, congress could authorize interfund borrowing to cover the shortfall of the DI fund. However, this is also not a sustainable solution as this is projected to bring down the solvency of the OASI fund significantly. Another proposed short-term intervention is to change the tax allocation to bolster the DI fund, however this strategy will again undermine the solvency of the OASI fund.

The government could try to raise taxes and/or cut benefit payments. Both methods are very unpopular and given that this is an election year both parties are shying away from addressing that topic, given that people who receive benefits or are close to receiving benefits are a significant portion of the electorate. A couple of candidates have proposed raising taxes on the rich but according to the IRS, this will not make a significant difference and taxes will have to be raised on everybody to fund benefits.

The most likely scenario is that the DI fund will have to borrow the money to pay benefits. This again is very problematic given the current federal debt levels and economic growth projections (slow employment growth results in lower taxes collected). Even though the DI can borrow from the federal government itself, through the issue of special securities, future taxpayers will be responsible for both their own future benefits as well as the interest-accumulating debt on past collections. This fundamentally goes against the idea of fairness as future members are unduly burdened.

The whole Social Security structure has to go through radical reform to ensure its existence to current and future beneficiaries. Although the DI fund is the most at strain, all trusts of Social Security are projected to be depleted within the next generation of workers. This means that people starting their careers now and paying taxes will most likely not receive any benefits when they become eligible. Unlike a private benefit scheme where there are a number of avenues the scheme can go into to get rid of the liability, e.g. pay out the benefits in cash lump sums or buy insurance contracts for the individuals, this is not a simple task for Social Security due to funding levels, dependents’ benefits, intergenerational fairness issues and political constraints.

The question of what the proper role of government is in the U.S. has been fought for centuries since the inception of the country. Few other issues have divided the nation as much as Social Security. It has reached a point where the programme has become too big to fail with no possibility of a bail out. However this situation plays out, actuaries should always keep in mind the long term consequences of any programme/scheme and should advise governments appropriately. Keynes did say that in the long run we are all dead but future beneficiaries will most likely still be alive and they shouldn’t have to pay for the mistakes of the past-generations.


History of retirement

Social Security Website

Congressional Research Service Repoert.   Social Security: The Trust Funds

Leave a comment

Some of the possible consequences of using incorrect assumptions

It is extremely important for the provider of any insurance contract to use reasonable assumptions in pricing of the contract because of the many consequences an inadequate premium has on the operations of the company. In this article, one of the consequences of this is explored in the context of long term care (LTC) contracts. This is a relatively new product so when insurers first began pricing LTC policies, they had little comparable experience to use. Over time and with the benefit of hindsight, it has become clear that the LTC industry as a whole missed the mark in their initial pricing assumptions and underestimated factors such as the number of policies that would lapse, the longevity of policyholders, and rising health care costs. As a result, LTC insurers set premiums too low for those older blocks of business and have since sought regulatory approval for rate increases. Not surprisingly, policyholders have fought back.Due to the nature of the benefits that these products offer i.e. costs of care in nursing homes for the elderly, some argue that it is unfair to pass the costs of inaccurate actuarial assumptions to the consumers. What options then, if any, do insurance companies have in such scenarios?


Leave a comment

Possible mass exodus of workers in New Jersey

Recently Governor Chris Christie of New Jersey announced a plan involving retirement and health benefits which decrease the costs to the state by lowering the benefit to public workers.  This comes after 5 years where his previous policy involved requiring members to pay more with the promise that the state would provide additional money. This promise did not last long as there were year on year budget cuts to pension payments.

After admitting that the 2011 reform did not help reduce costs and liabilities enough hence the need for a new plan. This plan entails freezing the old plan i.e. ending the accrual of employee contributions and new benefits with the aim of achieving a more manageable pension payment schedule.

With the freezing of old plans, new retirement plans need to be created. The suggestion is to move active members in to a cash balance plan, a hybrid of a DB and DC fund. It is similar to a DC fund as the employee benefits would show in an individual’s account.

The plan would be more attractive to younger, newer employees than it would be to the more experienced older ones.  A study of a similar plan implemented in another state proved that younger employees would receive more money on the new plan than their previous DB one but older employees would receive far less.  These factors could inspire the mass exodus of important public officials which will impact the effectiveness of operations.  It also runs the risk of failing to attract qualified and talented individuals in the future.

The commission stated that there is more than $80billion in unfunded pensions liabilities and that there is decreasing window to make a change. But how great is the cost of this soloution going to be to the average public employee?

Sources :

Leave a comment

Providing new Life to the German Insurance market.

In a recent report published by the International Monetary Fund (IMF) on global financial stability there was a worrying outlook. The issue was not a further Greek crisis or struggling Chinese growth but rather the now high-risk life insurance sector of Europe and in particular Germany.

Life insurance has formed the backbone of German long term savings for over two centuries, providing guaranteed returns for millions. However the IMF has noted that the failure of a single one of these firms “could trigger an industry-wide loss of confidence” and “could engulf the financial system”.

The industry manages over €900bn in assets and earns more than €90bn in premium income annually. However a major challenge faced by the insurers is the persistence of low interest rates that constrain investment returns. A second challenge faced by the company is the management of their guaranteed long-term policies. The first major concern is that guaranteed rates far outstrip today’s marginal investment returns. The government has tried to introduce new legislation to counteract this imbalance resulting in a cap of 1.25%. However due to the long-tail of the business (typically 30 years) the average guarantees still remain unattainably high at 3.2%. Comparing these guarantees with the 0.14% yield on 10-year Bunds it becomes clear the strain felt by many German Life Insurers. A second issue arises due to a large mismatch between assets which generally have a term of 9 years, and liabilities which have an average term of 20 years.

The ratings agency Moody’s further predicts that investment returns are likely to fall for the next three years. Concurrently attracting new business has proved difficult for the companies as the lower guaranteed returns appear unappealing. A Bundesbank stress test further found that a third of businesses will be short of capital if the current financial climate persists.

However a private equity group Cinven is attempting to purchase many life insurance books in the German industry and thus provide mass relief to the industry. This is a similar move that was seen in the UK by companies like Phoenix and Guardian.

Cinven already offers many unit-linked life insurance contracts but is looking to expand and in the process is buying up traditional guaranteed life insurance contracts to combine these unwanted portfolios and their current holdings.

Cinven understand the considerable risk in taking up many discontinued products but have a steady track record in their dealings. In 2015 the company sold Guardian Financial Services, a closed book of UK life insurance, to Swiss Re for £1.6bn effectively quadrupling their investments. They have also been interested in purchasing Old Mutual’s UK wealth management arm with an initial offer made.