From Data to Decisions: How Actuarial Assumptions Shape Financial Outcomes

Estimated Reading Time: 6 minutes

Actuarial assumptions are at the heart of actuarial science, serving as the foundational estimates used to predict future financial outcomes in insurance, pensions, and other financial domains. These assumptions are not mere guesses; they are carefully crafted based on statistical data, historical trends, economic conditions, and the actuary's professional judgment. They play a pivotal role in determining how insurance premiums are set, how pension liabilities are calculated, and how financial risks are assessed across various sectors. In essence, actuarial assumptions provide a lens through which future uncertainties are viewed and quantified. A few examples inspired from real world situations are:

• Imagine a life insurance company that had been profitable for decades. Their pricing models were built on assumptions that had stood the test of time, and they felt confident about their future. Then, suddenly, a pandemic struck, altering mortality rates overnight. The company's reserves, once deemed sufficient, quickly dwindled as claims poured in at an unprecedented rate. It became clear that their assumptions, based on historical data, no longer applied. This crisis forced the actuaries to revisit their assumptions, adapt to new data, and recalibrate their models to ensure the company’s survival.

• Consider a pension fund that had thrived for years, promising employees a secure retirement. Their assumptions about investment returns were based on a decade of strong market growth. Then came a financial downturn. The fund’s investments plummeted, and the assumed returns were nowhere in sight. Suddenly, the comfortable cushion they had projected vanished, leaving the fund scrambling to meet its obligations. This scenario is a stark reminder of how assumptions, when misaligned with reality, can jeopardize even the most established financial plans

• Another example involves a health insurance company that faced unexpected challenges due to advancements in medical treatments. For years, their assumptions about morbidity rates had been consistent, with claims following predictable patterns. However, with the introduction of expensive, cutting-edge therapies, claim costs surged beyond their projections. This sudden shift forced the company to revise their assumptions and adapt their pricing models to stay afloat. These stories underscore the dynamic nature of actuarial work, where assumptions are not static but must evolve with an ever-changing world.

One of the most critical aspects of actuarial assumptions is that they must be adaptable. For example, life insurance companies rely heavily on mortality assumptions to estimate how long policyholders are expected to live. These assumptions are based on large data sets that analyze life expectancy trends over time. However, factors such as medical advancements, lifestyle changes, or unexpected health crises can influence mortality rates, necessitating adjustments to these assumptions. Actuaries must regularly review and update their assumptions to ensure that they remain accurate and relevant, especially when the potential for deviation can have significant financial implications.

Since so much of our work relies upon data as the starting point, it’s important to realize that assumptions cannot compensate for poor data as garbage in becomes garbage out. Just like in life we should be careful not to be too presumptuous, in actuarial work we should also know when assumptions scope ends and not to assume till the logical extreme. Relying upon data driven decision making is good, but it is not the whole picture. As professionals we have to be holistic instead of only seeing one side of the equation. For many practical outcomes there might be limitations in the data such as:

• Data might be too small to be credible. We cannot evaluate 10 claims, say that law of large numbers has kicked in on 10 claims and construct our whole pricing based on that.

• Data might be too volatile or erratic. That is the main reason why many general insurance lines of business don’t apply chain ladder reserving and rely upon alternative reserving methods then.

• Relying upon the wrong data might be worse than relying upon no data at all. So, your company has corporate health insurance in top tier cities, hospitals and companies. Relying upon that data to arrive at pricing premium for micro-health insurance will render your premium likely 10-20 times higher than the required market price.

• This is one of the very attractive aspects of actuarial training. We have seen non-quantitative professionals relying upon some numbers as either a point of absolute truth or something that can never be relied upon. Actuaries achieve a balance in this bipolar interpretation of number where numbers are not seen as immutable facts and, in our reports, we quantify the numbers but also disclose our assumptions and points of uncertainty that we don’t claim to know it all. An insightful example is by statistician Abraham Wald’s contribution in World War 2. During World War II, statisticians analyzed returning planes and noticed bullet holes were mostly on the wings and fuselage and so wings should be reinforced with more armor. While others thought these were the areas needing reinforcement, the statistician Abraham Wald argued that since planes hit in the engines didn't return, those were the vulnerable spots. This example demonstrates critical statistical thinking: rather than taking data at face value, Wald focused on the unseen data (planes that didn't make it back), highlighting the importance of considering what's missing when analyzing real-world problems.

• There might be no data for emerging risks and insurance products in the first place. Here, benchmark data while acknowledging differences via adjustments is key to arrive at premiums.

• Data alone isn't enough for successful innovation; it requires anticipating consumer needs. For instance, if Henry Ford had solely relied on customer feedback, he might have focused on enhancing horses instead of inventing cars. Similarly, early focus groups rejected products like the iPhone and Crocs, which later became global sensations. This illustrates that while data-driven decision-making is important, the ability to foresee trends and desires, even before consumers recognize them, is crucial for creating revolutionary products that resonate with the market.

Transforming Insurance Industry with Actuarial Automation

Embark on a journey with Visionet to transform your strategies and embrace the limitless possibilities of Actuarial Automation. Dive into our white paper and discover how you can embrace efficiency, innovation and ensure adherence to regulatory standards. Download white paper: The Rise of Actuarial Automation Whitepaper - Visionet

Assumption setting is as much of as an art as it is a science. It should have some basis behind it too and not just classify as a hunch that it becomes a ‘black box’ where reason for setting some assumption becomes ‘professional judgement’ only. The reason doesn’t need to be exact or too concrete but it shouldn’t be just an inexplicable hunch too.

Actuaries should be careful not to be strict about insisting for data in every situation that this insight or work output can come from data otherwise nothing can be done about it. Assumption setting also operates around a spectrum of more assumptions or less depending upon the situation and it’s not the case that there would be no assumption required if data was perfect. There is always a layer of modeling and assumption setting that goes into the data to arrive at insights and projections. The best explanation on this point comes from Stephen Hawking in his book ‘The Grand Design’. He discussed how there is no "theory-independent" reality, suggesting that what we perceive as reality is always filtered through our models and theories. According to Hawking, we interpret the universe based on the frameworks we construct, and these frameworks shape our understanding. This concept aligns with the idea that our knowledge is model-dependent, meaning we can never access reality in a completely objective way without the influence of our theoretical constructs. The same is true for financial and actuarial modeling as well.

Economic assumptions are equally vital in actuarial work. These include assumptions about interest rates, inflation rates, salary growth, and investment returns. For example, in the context of pension plans, an actuary needs to estimate the rate at which employees' salaries will increase over time. This is crucial because pension benefits are often linked to the employee’s final salary. If the assumption about salary growth is too high or too low, it could lead to underfunding or overfunding of the pension plan, affecting the financial health of the fund and its ability to meet future obligations. Similarly, assumptions about inflation are essential, as inflation erodes the purchasing power of money over time. Actuaries must consider how inflation might impact the value of future payments, ensuring that the benefits promised today retain their real value years down the line.

The Process of Developing Actuarial Assumptions

The development of actuarial assumptions is a meticulous process that involves analyzing historical data, understanding trends, and applying statistical methods to make educated forecasts. This process begins with data collection, where actuaries gather information from internal experience studies, industry benchmarks, and external data sources. Experience studies are particularly valuable because they provide insights into how an organization's actual experience compares to the assumptions that have been made in the past. For instance, an insurance company might conduct an experience study to evaluate its mortality rates over the last decade, comparing them with the assumptions used in pricing their life insurance products.

Once the data is collected, actuaries use statistical techniques to identify patterns and trends. This step is crucial because it enables actuaries to make informed predictions about future events. For example, when determining the likelihood of policyholders making a claim, actuaries will analyze past claims data, looking for patterns that indicate how often claims are made and the average claim size. These insights help in establishing assumptions that are reflective of actual experience, thereby improving the accuracy of pricing models and reserve calculations.

Expert judgment plays a significant role in this process as well. While data analysis provides a quantitative foundation, actuaries must also consider qualitative factors, such as changes in legislation, shifts in consumer behaviour, or emerging risks that may not be fully captured in historical data. This combination of quantitative analysis and qualitative insight ensures that assumptions are robust and responsive to a wide range of potential scenarios.

Challenges and Risks Associated with Actuarial Assumptions

One of the major challenges in setting actuarial assumptions is dealing with uncertainty. The future is inherently unpredictable, and even the most carefully constructed assumptions can be wrong. For example, during the global financial crisis of 2008, many actuarial assumptions related to investment returns were rendered inaccurate almost overnight. This led to significant financial losses for organizations that were unprepared for such a drastic change. Such events underscore the importance of regularly reviewing assumptions and being willing to adjust them in response to changing conditions.

Another challenge is the risk of "model risk," which occurs when the models used to develop actuarial assumptions are not properly calibrated or fail to capture certain aspects of reality. For instance, if an actuary uses a model that doesn't adequately consider the impact of rare but severe events (often referred to as "black swan" events), the resulting assumptions might be overly optimistic, leading to inadequate reserves or pricing. This is why stress testing and scenario analysis are essential components of the actuarial process, allowing actuaries to evaluate how their assumptions hold up under various extreme conditions.

There's also the issue of data limitations. Actuarial assumptions are only as good as the data on which they are based. If the data is outdated, incomplete, or inaccurate, the assumptions derived from it will be flawed. For example, in emerging markets where insurance data might be scarce, actuaries face the challenge of making assumptions based on limited information, which increases the risk of error. To mitigate this, actuaries often rely on industry benchmarks or international studies to supplement their analysis, but even these sources have limitations.

The Impact of Actuarial Assumptions on Financial Decision-Making

The influence of actuarial assumptions extends far beyond the actuarial profession itself; they play a crucial role in shaping financial decision-making across various industries. For insurance companies, these assumptions directly affect the pricing of policies, the estimation of reserves, and the assessment of profitability. A small change in an assumption, such as the mortality rate for life insurance policies, can significantly impact the premiums charged to policyholders or the reserves held to pay future claims. Therefore, getting these assumptions right is not just a matter of professional pride—it’s a matter of financial survival.

In the realm of pension funds, actuarial assumptions determine the level of contributions required to meet future benefit obligations. If assumptions about investment returns are overly optimistic, the fund may find itself underfunded, jeopardizing the retirement security of its members. Conversely, overly conservative assumptions could lead to unnecessarily high contributions, placing a financial burden on both employers and employees. This delicate balancing act highlights the critical role actuaries play in ensuring the long-term sustainability of pension plans.

Moreover, actuarial assumptions are integral to the field of enterprise risk management. By providing a framework for evaluating potential future outcomes, these assumptions help organizations identify, measure, and manage risks more effectively. For example, an insurance company might use actuarial assumptions to model the potential impact of a natural disaster on its claims experience, allowing it to establish appropriate reinsurance arrangements and capital reserves. This proactive approach to risk management is essential in an increasingly volatile and uncertain world.

Conclusion

In conclusion, developing accurate and relevant actuarial assumptions is a complex and dynamic process that requires a blend of technical expertise, data analysis, and professional judgment. These assumptions are the backbone of actuarial science, guiding decision-making, financial planning, and risk management across a wide range of industries. As the world continues to evolve, actuaries must remain vigilant, continually updating their assumptions to reflect changing realities and emerging risks. By doing so, they ensure that their work remains not only relevant but also invaluable in helping organizations navigate the uncertainties of the future.

Winner of our 3rd referral giveaway

We’re happy to announce the winner of our 3rd referral giveaway of Apple iPad is Arpit Surana. This is the second time Arpit has won our referral contest. He was also the winner of our last referral contest of Apple MacBook Pro.

We conduct referral giveaway contests every quarter. Look out for the next one sometime in January!

In the meantime, feel free to take advantage of our regular tiered referral reward program (see the below for details)