DISSERTATION IMPLEMENTATION GUIDE
Chapters 1–3 and Appendix Additions
Based on Supervisor Feedback — Ropafadzo E Rupende
Shaun B Nyathi | N02219845D | NUST 2026
Quick Reference: Where Everything Goes
Use this table to locate each addition in your dissertation at a glance.
Item Location
Full PVFB derivation (infinite sum to Ch. 2, Section 2.2.4
B × aᴿ)
Life annuity formulation Ch. 2, Section 2.2.4
Survival modelling (ₛpₓ) Ch. 2 §2.2.4 + Ch. 3 §3.7
IFoA subject mapping sentence Ch. 2, Section 2.2.4 (opening line)
Salary growth model Ch. 3, Section 3.7
NRR formula stated explicitly Ch. 3, Section 3.7
Explicit assumption table Ch. 3, Section 3.7
Real vs nominal (Fisher equation) Ch. 3 §3.7 + Appendix E
Actuarial Control Cycle Ch. 3 §3.1 + §3.2
Scheme vs member distinction Ch. 1, Section 1.4
Break-even framework (method) Ch. 3, Section 3.7
Full PVFB working Appendix A
Full numerical example Appendix B
Break-even full calculation Appendix C
Sensitivity analysis tables Appendix D
Real return illustration Appendix E
1. IFoA Subject Mapping Sentence
WHERE: Chapter 2, Section 2.2.4 — Opening line
This single sentence signals to the examiner that your methodology is grounded in
professional actuarial standards. Place it as the very first sentence of Section 2.2.4.
Suggested text:
"This study builds on CM1/CM2 valuation theory — specifically discounting,
cashflow modelling, and life annuity mathematics — extended using CS2
survival modelling and SP4 pension scheme frameworks for defined benefit
adequacy assessment."
2. Formal PVFB Derivation
WHERE (body): Chapter 2, Section 2.2.4 — after existing PUC formulas
WHERE (working): Appendix A
Define the annual pension payment B paid at the end of each year of survival after
retirement age R. The Present Value of Future Benefits is the sum of all future discounted
payments, each weighted by the probability of surviving to receive that payment:
PVFB = Σ B × vᵗ × ₛpᴿ (t = 1 to ∞) ... (2.1)
Since the annual benefit B is constant after retirement, it factors out of the summation:
PVFB = B × Σ vᵗ × ₛpᴿ (t = 1 to ∞) ... (2.2)
The remaining summation is recognised as the whole life annuity valued at retirement age R:
aᴿ = Σ vᵗ × ₛpᴿ (t = 1 to ∞) ... (2.3)
Substituting equation (2.3) into equation (2.2) gives the compact valuation expression:
PVFB = B × aᴿ ... (2.4)
Intuition sentence to add immediately after:
"This expression shows that the total pension obligation is equivalent to the
annual pension multiplied by the expected present value of a unit life annuity
at retirement, capturing both the time value of money and survival uncertainty
simultaneously."
What goes to Appendix A: Full expansion of the infinite sum, assumption of independence
between survival and discounting, and substitution of the NSSA benefit formula B = k × Sᶠᵢⁿᵃˡ ×
T into equation (2.4).
3. Formal Survival Model Definition
WHERE (body): Chapter 2, Section 2.2.4 — immediately before the PVFB derivation
WHERE (working): Appendix A alongside PVFB
Let T(x) be the future lifetime random variable for a life aged x. The survival probability is
defined as:
ₛpₓ = P( T(x) > t ) = S(x + t) / S(x) ... (3.1)
where S(x) is the survival function giving the probability of surviving from birth to age x. The
force of mortality μ relates to the survival function through the integral:
ₛpₓ = exp( − ∫₀ᵗ μₓ₊ₛ ds ) ... (3.2)
Intuition sentence to add immediately after:
"In this study, ₛpᴿ represents the probability that a member who reaches
retirement age R survives a further t years to receive the pension payment
due at time t, and is derived from the IPEC-sanctioned mortality table applied
to the Zimbabwean public service demographic."
What goes to Appendix A: Full derivation of ₛpₓ from first principles using the force of
mortality, and a note on how IPEC table values are used to compute discrete survival
probabilities.
4. Life Annuity Definition
WHERE: Chapter 2, Section 2.2.4 — as a named definition immediately before the PVFB
derivation
This is one definition, two lines. State it, name it, use it in the PVFB derivation that follows.
The whole life annuity valued at retirement age R is defined as:
aᴿ = Σ vᵗ × ₛpᴿ (t = 1 to ∞) ... (4.1)
"This represents the expected present value of a payment of one unit per
year for the remainder of a life aged R, discounted at rate i, where v = 1 / (1 +
i)."
5. Salary Growth Model
WHERE: Chapter 3, Section 3.7 — first equation in the quantitative model description
Let S₀ denote the starting salary at entry age and g the assumed annual salary growth rate.
The projected salary at time t is:
Sₜ = S₀ (1 + g)ᵗ ... (5.1)
The final salary at retirement after T years of service is therefore:
Sᶠᵢⁿᵃˡ = S₀ (1 + g)ᵀ ... (5.2)
"This final salary feeds directly into the NSSA benefit formula to project the
pension benefit at retirement."
6. NRR Formula Stated Explicitly
WHERE: Chapter 3, Section 3.7 — immediately after the salary model
The Net Replacement Ratio is the ratio of the annual pension benefit at retirement to the
final pre-retirement salary:
NRR = B / Sᶠᵢⁿᵃˡ = k × T ... (6.1)
"Under the NSSA formula for service not exceeding 30 years, k = 4/300. For
service exceeding 30 years, the two-tier formula applies as defined in Section
2.2.1. This shows that NRR is directly determined by the accrual rate and
years of service, and is independent of the salary level."
7. Explicit Assumption Table
WHERE: Chapter 3, Section 3.7 — after the model equations
This is the single most important addition to Chapter 3. It is the direct link between
assumptions and results that your supervisor flags as missing. Fill in the shaded cells with
your chosen values before submitting.
Assumption Symbol Value Source
Entry age x 30 Cohort definition
NRA Pre-reform R₁ 65 SI 197 of 2024
NRA Post-reform R₂ 70 SI 197 of 2024
Accrual rate (≤30 k 4/300 ≈ 1.33% NSSA formula
yrs)
Accrual rate (>30 k₂ 1% NSSA formula
yrs)
Salary growth rate g State assumed Ministry of Finance / IPEC
value
Discount rate i State assumed IPEC recommended rate
value
Mortality table — IPEC-sanctioned IPEC
table
Starting salary S₀ State assumed Public service salary scales
value
"These assumptions are held constant across both scenarios to isolate the
effect of the NRA change. Sensitivity analysis in Chapter 4 will test how
results change when key assumptions are varied."
8. Real vs Nominal Consideration
WHERE: Chapter 3, Section 3.7 — after the assumption table, as a modelling risk note
ALSO: Appendix E for full numerical illustration
All projections in this study are conducted in nominal terms. In Zimbabwe's economic
environment, where inflation π has historically exceeded the nominal discount rate i, the real
return on pension assets may be negative. The Fisher approximation gives:
(1 + r) ≈ (1 + i) / (1 + π) ... (8.1)
where r is the real return. If π > i, then r < 0.
"This means that even if the pension fund grows nominally, its purchasing
power declines. This study acknowledges that nominal adequacy results may
overstate real adequacy in a high-inflation environment. This limitation is
carried forward into the sensitivity analysis in Chapter 4."
9. Actuarial Control Cycle
WHERE: Chapter 3, Section 3.1 Research Philosophy — closing paragraph
ALSO: Brief reference in Section 3.2
Suggested paragraph:
"The methodology is further guided by the Actuarial Control Cycle, the
professional framework underpinning actuarial practice. The first stage,
specifying the problem, is addressed in Chapter 1 through the identification of
the pension adequacy gap and the research objectives. The second stage,
developing the solution, is operationalised in this chapter through the
construction of the PUC-based actuarial model, the assumption-setting
process, and the survey instrument. The third stage, monitoring the
experience, is reflected in the sensitivity analysis, which tests how results
change under varying economic and demographic conditions. This framework
ensures that the study moves beyond mechanical calculation toward critical
actuarial judgement, particularly important in Zimbabwe's unstable
macroeconomic environment."
10. Scheme vs Member Distinction
WHERE: Chapter 1, Section 1.4 Conceptual Framework — end of second paragraph
Suggested sentence:
"A central analytical distinction guiding this study is that a policy reform which
improves pension fund sustainability does not necessarily improve individual
pension adequacy — the interests of the scheme as an institution and the
interests of its contributing members may diverge significantly under an NRA
increase."
11. Break-Even Analysis Framework
WHERE (method): Chapter 3, Section 3.7 — final model component before real vs nominal
note
WHERE (calculation): Appendix C
Let B₆₅ and B₇₀ denote the annual pensions under each NRA scenario. The value of
pension foregone during the five delayed years is:
Lost Pension = 5 × B₆₅ ... (11.1)
The annual pension gain from retiring at 70 rather than 65 is:
Annual Gain = B₇₀ − B₆₅ ... (11.2)
The break-even period n, measured in years of survival after age 70, is therefore:
n = Lost Pension / Annual Gain = (5 × B₆₅) / (B₇₀ − B₆₅) ... (11.3)
"A member must survive at least n years beyond age 70 to benefit financially
from the NRA increase. This break-even result is compared against the
survival probabilities derived from the IPEC mortality table to assess the
likelihood of benefit."
What goes to Appendix C: Full numerical calculation with assumed values substituted, all
intermediate steps, and the resulting break-even figure with interpretation.
12. Appendix Structure
WHERE: New section after Chapter 5 and References
Every appendix must be cross-referenced from the relevant body section using a bracketed
note, for example: (see Appendix A for full derivation).
Appendix A: Full First-Principles PVFB Derivation
• Survival function S(x) defined from first principles
• Force of mortality μ and derivation of ₛpₓ via the integral formula
• Full infinite sum expansion showing each step from PVFB = Σ B × vᵗ × ₛpᴿ to PVFB =
B × aᴿ
• Assumption of independence between survival and discounting stated explicitly
• Substitution of NSSA benefit formula B = k × Sᶠᵢⁿᵃˡ × T into the final expression
Appendix B: Full Numerical Example
• Salary projection at NRA 65 and NRA 70 using the assumed g and S₀
• Pension benefit under both NRA scenarios using the two-tier NSSA formula (split at
30 years)
• NRR computed and compared at each NRA
• PVFB under both NRAs using assumed mortality table and discount rate
• All intermediate steps shown
Appendix C: Break-Even Analysis Full Working
• Lost pension calculated: 5 × B₆₅ with numerical values substituted
• Annual gain calculated: B₇₀ − B₆₅ with numerical values substituted
• Break-even years n calculated and interpreted against IPEC survival probabilities
Appendix D: Sensitivity Analysis Tables
• PVFB across at least three values of i (discount rate)
• PVFB across at least three values of g (salary growth rate)
• PVFB under adjusted mortality assumptions
• Directional interpretation for each: higher i → lower PVFB; improved survival →
higher PVFB; higher g → higher PVFB
Appendix E: Real Return Illustration
• Numerical example showing (1 + r) = (1 + i) / (1 + π) under different inflation
scenarios
• Demonstration that when π > i, real pension value falls even as nominal figure grows
• Zimbabwe-specific inflation figures used to contextualise the modelling risk
13. Things to Remove or Change in Existing Chapters
Chapter 1
Section 1.2 Background — Trim
The second paragraph loses focus on the member-level question. From the sentence
beginning “France and China have also increased normal retirement ages” onward, trim this
to one sentence. The global comparison belongs in Chapter 2, not the background.
Section 1.12 Assumptions — Remove two assumptions
Remove the following two lines entirely — they are not research assumptions:
• "The respondents will be independent in answering any questions."
• "The questions asked will be clear and easy to understand."
Replace with:
• One assumption about the stability of the NSSA benefit formula over the full career
horizon.
• One assumption that the IPEC mortality table is representative of the public service
demographic.
Chapter 2
Section 2.2.1 — Remove Duplicate Introduction
The paragraph beginning “The purpose of this chapter is to review the literature relevant to
this study” appears in both Section 2.1 and Section 2.2.1. Delete the duplicate in Section
2.2.1 entirely.
Section 2.2.4 — Reframe Opening Tone
The section currently reads as a literature comparison of valuation methods. Reframe it as
your actuarial framework. Change the opening sentence from a review statement to a
framework statement. For example, replace an opening like “The valuation of Defined
Benefit schemes requires a robust actuarial cost method…” with:
"This section establishes the actuarial mathematical framework underpinning
this study, grounded in CM1/CM2 valuation theory, CS2 survival modelling,
and SP4 defined benefit scheme principles."
Section 2.3.1 — Remove Duplicate Benchmark Discussion
The paragraph on OECD adequacy benchmarks in Section 2.3.1 repeats material already
covered in Section 2.2.2. Cut the benchmark discussion from 2.3.1 and keep only the
empirical findings on actual replacement rates achieved globally.
Chapter 3
Section 3.5 Sampling — Expand or Acknowledge
Currently one paragraph with no sample size stated and no justification beyond naming
random sampling. Do one of the following:
1. State the sample size and justify it using a standard formula such as Slovin’s formula
or Cochran’s formula.
2. If the survey is not yet complete, acknowledge the sample size as a delimitation
explicitly.
As written, this section will draw a question from the examiner.
Section 3.7 Data Analysis — Expand Significantly
This is currently the weakest section in Chapter 3 and the one that needs the most work. All
additions listed in Sections 5 through 11 of this document belong here. It should become
your longest methodology section because it contains your full model specification: the
salary model, NRR formula, PVFB equation, assumption table, break-even framework, and
real vs nominal note.
Section 3.8 Validity and Reliability — Expand
Currently one line. Add two sentences:
3. One on how using IPEC-sanctioned mortality tables and IPEC-recommended
discount rates strengthens the validity of the actuarial component.
4. One on how the PUC method’s alignment with IAS 19 (Employee Benefits) ensures
methodological reliability and consistency with international best practice.
END OF IMPLEMENTATION GUIDE
All additions are grounded in supervisor feedback and IFoA subject material. No content has been
hallucinated.