Kasuku Pension Scheme: Chasing an 11% Discount Rate with a 70/30 Portfolio
The Kasuku Pension Scheme holds KES 10 billion of assets against actuarial liabilities discounted at 11% p.a., giving a funding ratio of 96%. The current policy is 70% equities and 30% bonds. The investment committee is debating a move to 50% equities/50% bonds to cut volatility after a 14% drawdown, and the consultant has proposed instead 55% equities, 35% bonds and a new 10% allocation to unlisted infrastructure with a three-year lock-up and no secondary market. Expected returns: equities 13%, bonds 8%, infrastructure 10.5%. Volatilities: 18%, 7%, 9%. Equity-bond correlation 0.25, equity-infrastructure 0.30, bond-infrastructure 0.10. Contributions are KES 500m a year and benefit payments KES 800m a year.
1. Compute expected return and volatility for all three allocations and test each against the 11% required return.
Read the worked answer
70/30: return = 0.7(13) + 0.3(8) = 11.50% variance = 0.49(324) + 0.09(49) + 2(0.7)(0.3)(18)(7)(0.25) = 158.76 + 4.41 + 13.23 = 176.40, so sigma = 13.28%
50/50: return = 6.5 + 4.0 = 10.50% variance = 81 + 12.25 + 15.75 = 109.00, so sigma = 10.44%
55/35/10: return = 7.15 + 2.80 + 1.05 = 11.00% variance = 98.01 + 6.00 + 0.81 + 12.13 + 5.35 + 0.44 = 122.74, so sigma = 11.08%
Test against 11%:
- 70/30 clears it with 50bp to spare but at 13.28% volatility.
- 50/50 misses it by 50bp - roughly KES 50m a year of expected return, compounding against a 96% funding ratio. That is the option the committee is actually debating and it does not work.
- 55/35/10 hits exactly 11.00% with volatility 220bp below the current policy - and reachable only because infrastructure has a low correlation with both asset classes.
Note the 50/50 fix: to reach 11% it would need equities to return 14%, above the 13% assumption.
2. Address the liquidity objection: can the scheme hold a KES 1 billion locked-up sleeve?
Read the worked answer
Cash flow arithmetic
- Net outflow = 800 - 500 = KES 300m a year
- Liquid assets after the shift = 10bn - 1bn infrastructure = KES 9bn
- Coverage = 9,000/300 = 30 years of net outflow from liquid assets alone.
So the lock-up does not threaten benefit payments in any normal year. The real risks are different:
- Rebalancing under stress. If equities fall 30%, the equity sleeve drops from 5.5bn to 3.85bn while the fund is 9.63bn - target equity is 5.30bn. Closing that gap needs 1.45bn from bonds, and 1bn of the bond sleeve is now illiquid. Workable, but only if the liquid bond book is large enough; stress-test it before signing.
- Valuation and governance. Unlisted assets are marked on models, not prices. Ask who values them, how often, and what the fee load is net of the 10.5% gross assumption.
- Drawdown risk on commitments. If the allocation is by capital commitment rather than funded now, undrawn commitments are a claim on liquid assets that must be pre-funded.
- Concentration and vintage. A single fund, a single vintage and a single manager is not a diversifier; it is a bet on one GP.
Conclusion: liquidity is not the binding constraint; the binding constraints are valuation independence and the credibility of the 10.5% net assumption.
3. Recommend an allocation to the committee, with the conditions attached.
Read the worked answer
Recommendation: adopt 55/35/10, subject to conditions.
Why not the alternatives: 50/50 is the intuitive response to a drawdown and it is the one option that fails the actuarial test - it cuts volatility by 284bp but creates a structural 50bp shortfall. 70/30 meets the return but needs 13.28% volatility to do it, which is what produced the 14% drawdown in the first place.
Why 55/35/10: it delivers 11.00% with 11.08% volatility, a materially better risk-return trade than either alternative, because the diversifying asset has low correlation with both sleeves. It cuts the required equity risk rather than the required return.
Conditions before funding:
- Independent valuation policy and quarterly reporting, with fees netted from the 10.5% assumption - re-run the maths if net expected return falls below 10.5%.
- Cap the sleeve at 10% and no single manager above that; require a co-investment or secondaries route so the lock-up is not absolute.
- Stress test: equities -30% and bonds +200bp simultaneously, then confirm rebalancing is possible with liquid assets only.
- Update the investment policy statement and the actuarial discount rate assumption together; if the 11% is itself unsupportable, the conversation moves from asset allocation to contribution policy, and the committee should say so to the sponsor rather than quietly de-risking.