Calculating Technical Provisions: Prospective and Retrospective Methods

Two Paths to the Same Destination

At any point during the life of a contract, like a pension plan or an insurance policy, we can calculate the necessary reserve or liability. We can think of this valuation date as a point on a timeline. The prospective method calculates the reserve by looking at everything that is expected to happen from this point forward. The retrospective method arrives at the same figure by looking at everything that has happened up to this point.

This key duality is captured perfectly in the following diagram.

As you can see, the prospective method is "forward-looking" and uses present value (PV) formulas to discount future events. The retrospective method is "backward-looking" and uses future value (FV) or accumulation formulas to compound past events.

The Prospective Method: A Forward-Looking View

The prospective method is generally the more intuitive of the two. It answers the question: "As of today, what is the net amount of money we need to have on hand to meet all future obligations?"

The liability is calculated as the present value of all expected future outgo (benefits, expenses) minus the present value of all expected future income (premiums, contributions).

Prospective Liability = PV(Future Benefits + Expenses) - PV(Future Premiums)

This approach is the bedrock of modern valuation, as it directly aligns with the market-consistent principle of discounting expected future cash flows. Your work in CM1 on calculating policy values is a direct application of this method.

A key text on pension mathematics provides the formal definition for the actuarial liability, which is what we are calling the technical provision in this context.

In practice, this is the standard method used for valuing most life insurance and pension liabilities. For example, a paper from the Society of Actuaries on valuing with-profits business notes that even for these complex products, a prospective approach is a valid and recognized method.

The Retrospective Method: A Backward-Looking View

The retrospective method answers a different, but equivalent, question: "What is the accumulated value of all the net contributions made on behalf of this policy or member to date?"

The liability is calculated as the accumulated value of all past income (premiums, contributions) minus the accumulated value of all past outgo (benefits paid, expenses incurred, or the cost of providing insurance cover).

Retrospective Liability = AV(Past Premiums + Expenses) - AV(Past Benefits Paid)

This method can feel less intuitive. A crucial component of the accumulation for a group of policies is the "benefit of survivorship." As some policyholders leave the group (due to death, lapse, or withdrawal), the funds held for them are effectively redistributed among the remaining members. Your background in probability theory will help you see this as a consequence of moving from an unconditional to a conditional probability space—we are calculating the reserve for those who have survived to the valuation date.

The Winklevoss text also provides a formal definition for this approach.

While the prospective method is more common for statutory reporting, the retrospective method is vital in certain contexts. Its most prominent use is in the calculation of an asset share for with-profits policies. An asset share is the policy's notional share of the firm's assets, built up by accumulating its premiums at the actual investment return achieved by the fund, and deducting charges for expenses, costs, and profits.

The SOA paper you looked at earlier provides an excellent description of this.

The Principle of Equivalence

The most important concept to grasp is that, given a consistent set of assumptions, the prospective and retrospective methods must produce the same result. They are simply two different ways of calculating the same quantity.

This equivalence stems from the original equation of value set at the start of the contract:

At any valuation date $t$, we can split both sides into past and future components:

Rearranging this gives:

The left side of this equation is the prospective liability. If we accumulate the right side to the valuation date $t$, we get the retrospective liability. Thus, the two are mathematically equivalent. The Winklevoss text demonstrates this by showing that both the prospective definition in equation (6.3a) and the retrospective definition in (6.4a) simplify to the same underlying liability formula (5.3).

Conclusion

In this lesson, we have demystified the two primary methods for calculating technical provisions. We saw that the prospective method is a forward-looking calculation based on the present value of future net cash flows, while the retrospective method is a backward-looking calculation based on the accumulated value of past net cash flows.

You learned that these are not competing methods but rather two sides of the same coin, which must produce identical results under a consistent set of assumptions. While the prospective method is the default for most modern valuation regimes like Solvency II, the retrospective approach remains indispensable for specific products like with-profits business (via asset shares) and as a powerful tool for cross-checking results.

In our next lesson, we will use the liability values we've learned to calculate today. We will explore how they are compared against a firm's assets to determine its financial health, leading us to define and analyze accounting surplus and economic surplus.

Key Takeaways & Cheatsheet

Here is a comparative framework to help you consolidate these concepts for your exam preparation.