ALMF March 2021 - Solutions
ALMF March 2021 - Solutions
ALMF March 2021 - Solutions
Code: E_FIN_IIALM
Graphical calculator
allowed: No
Number of questions: 15
1
Please type in formulas in the following way: 𝑥 ⋅ (1+𝑦)𝑛 is x * 1 / (1+y)^n
Good luck!
School of Business and Economics
USE YOUR TIME EFFICIENTLY: IT IS NOT NECESSARY TO WRITE DOWN LONG STORIES, I.E. PLEASE BE
CONCISE WITHOUT LOSING CLARITY.
The financial position of a defined benefit pension fund is often measured by the nominal funding ratio.
The nominal funding ratio FR at time t is calculated as:
𝐴𝑡
𝐹𝑅𝑡 = ,
𝐿𝑡
where 𝐴𝑡 represents the market value of the assets and 𝐿𝑡 the present value of the liabilities. There are
many factors that drive changes in the nominal funding ratio. Longevity is one of those factors.
(a) Mention two other factors that drive changes in the nominal funding ratio. (6 pts)
• Asset returns
• Indexation
• Contributions
(b) Explain what longevity risk means in the context of pension funds. (4 pts)
1
The price of entitlements is based on expected payments which depends on mortality rates. If people
live longer than expected then the contribution paid for entitlements was too low. The consequence is
that the funding ratio will fall.
In a defined benefit pension fund the present value of future cash-flows is calculated by using risk free
zero rates.
(c) Explain why the valuation method for liabilities does not contain a risk premium for longevity risk. (4
pts)
In a pure defined benefit pension plan pension rights are guaranteed, i.e. the participant has
possession without any downside risk. This asset should earn the risk free rate. Longevity risk is an
issue but the idea of defined benefit is that the (increased) cost of longevity is covered by future
generations.
Suppose that pension fund Blue Sky has just two cash-flows to be paid in the future: EUR 100,000,000 in
10 years and EUR 100,000,000 in 20 years. Risk free zero rates (per annum with annual compounding)
are 0.00% for ten years and 0.40% for 20 years. The present value of these cash-flows is EUR
192,326,366.
Assume now that the Board of pension fund Blue Sky hires an asset manager to hedge the interest rate
risk of the liabilities. The Board has a preference for a cash-flow matching approach. The asset manager
receives EUR 192,326,366 (the value of the liabilities) to execute this assignment. The market offers risk-
free (i.e. non-defaultable) zero-coupon bonds for all possible maturities. The asset manager does not
have access to the swap market.
(a) Give the details of the strategy that will lead to a perfect match of the cash-flows. (5 pts)
(b) If the market would offer high-yield corporate bonds (i.e. corporate bonds with low credit rating) for
all possible maturities, would you consider to include these in your hedge portfolio if you were the
asset manager? Please motivate your answer. (3 pts)
No, the objective is to hedge the risk of the liabilities. This can be achieved (perfectly) by the strategy
of (a), so there is not a single reason to consider any other asset.
Unfortunately, the assumption that the market offers zero-coupon bonds for all possible maturities is
too optimistic. At the moment the asset manager wants to bring the trade to the market, there is only
liquidity in the 15-years zero-coupon bond. The other investment possibility the asset manager has is
2
cash. Cash generates a return of 0% and has a duration of 0. The 15-years zero rate is 0.35% (per annum
and with annual compounding). The asset manager proposes to employ duration-matching.
(c) Give the full details of the replication strategy, i.e. provide the amounts of money the asset manager
wants to invest in the 15-years zero-coupon bond and cash. (7 pts)
𝟏𝟓
= 𝟏𝟒. 𝟗𝟓.
(𝟏 + 𝟎. 𝟑𝟓%)
𝟏𝟒. 𝟕𝟔
× 𝟏𝟗𝟐, 𝟑𝟐𝟔, 𝟔𝟔𝟔 = 𝟏𝟖𝟗, 𝟗𝟒𝟎, 𝟓𝟏𝟔.
𝟏𝟒. 𝟗𝟓
The modified convexity of a zero coupon bond is given by the following formula:
(𝑇 − 𝑡)2 + (𝑇 − 𝑡)
Mod Conv (ZCB) = .
(1 + 𝑦𝑡:𝑇 )2
The portfolio that was constructed in (c) is “short” convexity versus the liabilities, i.e. the convexity of
the hedge portfolio is lower than the convexity of the liabilities.
(d) Give two economic scenarios under which the hedge portfolio will underperform the liabilities as a
result of the convexity position. (6 pts)
• Large parallel upward movements of the yield curve (in environments of inflation uncertainty)
• Large parallel downward movements of the yield curve (in crisis situations)
Suppose pension fund TechNovation has a liability profile of which the modified DV01 properties are
given in the first table below in column Mod DV01 L. The details of the first bucket are given in the
second table. The pension fund has a nominal funding ratio equal to 105%. The present value of the
3
liabilities equals EUR 800,000,000. The fund has the aim to hedge the interest rate sensitivity of the
funding ratio.
The fund decides to enter into a 2-years nominal interest rate swap to hedge the funding ratio for
changes in the interest rates that fall into the first bucket. Zero rates (per annum and with annual
compounding), discount factors and swap rates are given in the following table. You can assume that we
live in the text book world without credit risk, i.e. counterparty credit risk is not relevant for swap
pricing, and that the reset frequency of the floating rate is once per year.
(a) Calculate the notional amount that is needed in the 2-years nominal interest rate swap. (10 pts)
Now, calculate the mod DV01 of EUR 1 notional in the 2-yrs interest rate swap.
𝟎. 𝟏% (𝟏 + 𝟎. 𝟏%)
𝟏⋅ 𝟐⋅
(𝟏 + 𝟎. 𝟎𝟎%) (𝟏 + 𝟎. 𝟏𝟎%)𝟐
𝑴𝒐𝒅 𝑫𝑽𝟎𝟏 𝒇𝒊𝒙𝒆𝒅 𝒍𝒆𝒈 = + = 𝟎. 𝟎𝟎𝟎𝟏𝟗𝟗𝟕.
𝟏𝟎, 𝟎𝟎𝟎 ⋅ (𝟏 + 𝟎. 𝟎𝟎%) 𝟏𝟎, 𝟎𝟎𝟎 ⋅ (𝟏 + 𝟎. 𝟏𝟎%)
And
4
(𝟏 + 𝟎. 𝟎𝟎)%
𝟏⋅
(𝟏 + 𝟎. 𝟎𝟎%)
𝑴𝒐𝒅 𝑫𝑽𝟎𝟏 𝒇𝒍𝒐𝒂𝒕𝒊𝒏𝒈 𝒍𝒆𝒈 = = 𝟎. 𝟎𝟎𝟎𝟏.
𝟏𝟎, 𝟎𝟎𝟎 ⋅ (𝟏 + 𝟎. 𝟎𝟎%)
Then
𝟕𝟑, 𝟓𝟎𝟎
𝑵= = 𝟕𝟑𝟕, 𝟐𝟎𝟕, 𝟐𝟎𝟒.
𝟎. 𝟎𝟎𝟎𝟎𝟗𝟗𝟕𝟎
The hedge will not be fully accurate in all possible interest rate scenarios. Assume that all interest rates
move up with 10bps. Assume also that the value change of the liabilities and all fixed income
instruments can be very accurately approximated by using only modified DV01.
(b) Calculate the funding ratio after the change in interest rates. For the avoidance of doubt: it is not
necessary to calculate exact values of the liabilities and the fixed income assets: an approximation
using modified DV01 is sufficient. The only exception is the 2-years swap: you need to calculate the
value of this swap exactly. If you were not able to solve question (a) you can use a notional amount
of EUR 700,000,000. (10 pts)
𝟎. 𝟏% 𝟏 + 𝟎. 𝟏%
𝑽(𝒇𝒊𝒙𝒆𝒅 𝒍𝒆𝒈) = 𝑵 ⋅ ( + ) = 𝟕𝟑𝟓, 𝟕𝟑𝟕, 𝟗𝟑𝟔
(𝟏 + 𝟎. 𝟏𝟎%) (𝟏 + 𝟎. 𝟐𝟎%)𝟐
𝟏 + 𝟎%
𝑽(𝒇𝒍𝒐𝒂𝒕𝒊𝒏𝒈 𝒍𝒆𝒈) = 𝑵 ⋅ ( ) = 𝟕𝟑𝟔, 𝟒𝟕𝟎, 𝟕𝟑𝟒
(𝟏 + 𝟎. 𝟏%)
𝑽(𝒂𝒔𝒔𝒆𝒕𝒔) = 𝟏𝟎𝟓% ⋅ 𝟖𝟎𝟎, 𝟎𝟎𝟎, 𝟎𝟎𝟎 − 𝟕𝟑𝟕, 𝟐𝟗𝟕 − 𝟐, 𝟏𝟐𝟏, 𝟎𝟎𝟎 ⋅ 𝟏𝟎 = 𝟖𝟏𝟖, 𝟎𝟓𝟕, 𝟐𝟎𝟑.
5
This delivers a funding ratio of:
The hedge is very accurate. This might be a bit surprising given that the notional amount in the 2-years
swap is very large (especially if we compare this notional amount to the total value of the liabilities).
(c) Give one explanation for the fact that the hedge is very accurate in this situation. (3 pts)
We analysed the consequences of a small and parallel change of the yield curve. Besides, the first
bucket covers only a small part of total interest rate sensitivity.
The second part of the course was about allocating a large amount of money sensibly across asset
classes and within asset classes. One tool that is often used in both academia and industry is the mean-
variance approach. Under certain conditions the mean-variance approach delivers efficient portfolios.
One of these conditions is that the risk-return preferences of investors are described by quadratic utility.
Suppose pension fund Bankia decides to apply the mean-variance approach. The Board of the pension
fund recognizes that the fund is a long-term investor. However, they are also aware of the drawbacks of
solving a long term optimization problem. Therefore, the pension fund takes a myopic approach, where
the fund reconsiders the strategic investment policy each year.
(a) Explain carefully what a myopic investor is and mention one (theoretical) disadvantage of the
myopic approach for long term investing (5 pts)
In the context of long horizon portfolio optimization, the myopic investor treats the long term
portfolio optimization problem as a series of short term optimization problems. The myopic investor
incorporates new information into the problem at each date the portfolio is optimized.
The nominal funding ratio of pension fund Bankia is 110%. As a consequence, the fund sets the expected
return target for the portfolio equal to the return on liabilities plus 2%. The investment opportunity set
of pension fund Bankia consists of the nominal matching portfolio (that tracks the liabilities perfectly),
equity developed markets and equity emerging markets. The expected return on equity developed
markets equals the return on liabilities + 3% and the expected return on equity emerging markets equals
the return on liabilities + 6%. The standard deviation of returns on equity developed markets (in excess
of the return on liabilities) equals 15% while the standard deviation of returns on equity emerging
markets (in excess of the return on liabilities) equals 30%. Finally, the correlation between (excess)
returns on equity developed markets and (excess) returns on equity emerging markets is 0.7.
6
As said the pension fund applies the mean-variance approach, meaning that in general the weight in
risky assets is given by the following formula:
1
𝑤𝑡∗ = Σ𝑡−1 𝜇𝑡 ,
𝛾
Under the assumptions above, the inverse of the variance-covariance is matrix given by:
87.14597 −30.5011
Σ𝑡−1 = [ ].
−30.5011 21.78649
(b) Calculate the portfolio that minimizes risk given the expected return target of the fund. (9 pts).
Gamma is unknown but that is not relevant for this problem because the expected return target is
given. First we calculate the optimal portfolio for an investor with risk aversion equal to 1.
Scaling this portfolio back to weights that sum to 100% gives the tangent portfolio:
𝑡𝑔𝑡 66.67%
𝑤𝑡 = [ ]
33.33%
This portfolio delivers an excess returns of liabilities equal to 4%. We only need an excess return of
2%. Hence we invest for 50% in the tangent portfolio and 50% in the nominal matching portfolio.
A pension fund wants to have an investment policy such that the characteristics of the policy are in line
with the short term and long term objectives of the fund. In other words, when the Board of a pension
fund decides on the investment policy it should take the interests of all stakeholders of the fund into
account, i.e. it should make a trade-off between risk and ambition that is accepted by all stakeholders.
(a) Mention two important risks a pension fund should focus on when deciding on the strategic
investment policy (4 pts).
The table below shows some outcomes of an ALM-study. This ALM-study was conducted at a starting
nominal funding ratio of 95%. The table shows the ALM-results of five different strategies:
• Strategy 1 (Strat 1): an investment of 100% in the nominal matching portfolio
7
• Strategy 2 (Strat 2): an investment of 100% in the real matching portfolio
• Strategy 3 (Strat 3): 75% nominal matching portfolio and 20% return portfolio
• Strategy 4 (Strat 4): 75% nominal matching portfolio and 30% return portfolio
• Strategy 5 (Strat 5): 75% matching portfolio and 40% return portfolio.
Please note that this pension fund has a policy that benefits are cut if the regulatory nominal funding
ratio falls below 97% (in any year).
A couple of remarks concerning the table:
• The long term results (bottom part of the table) are calculated under the assumption that the
investment policy doesn’t change during the coming 15 years (for the sake of completeness:
portfolio weights are rebalanced to the strategic weights each year).
• The long term indexation result of the strategy (criterion 8) shows the expected portion of the
realized inflation that can be given as indexation over the full payment of all current cash-
flows. This criterion is shown under the short term results because the 1-year average
investment return (in excess of the nominal liabilities) is used to calculate this figure.
• Purchasing power (criteria 12 and 13) shows the real value of EUR 1 pension.
8
Short term (1 year) Strat 1 Strat 2 Strat 3 Strat 4 Strat 5
Funding ratio
1 Average nominal funding ratio after 1 year 94.8% 94.4% 95.4% 95.8% 96.2%
Lower band 95% confidence interval nominal funding ratio
2
after 1 year 90.7% 85.5% 88.8% 88.2% 87.5%
Upper band 95% confidence interval nominal funding ratio
3
after 1 year 99.0% 105.7% 103.4% 104.1% 104.8%
4 Probability nominal funding ratio lower than 104% after 1 year 100% 90% 98% 96% 94%
Costs for the employer
5 Average pension premium as % of salary for year 2 32% 33% 33% 33% 33%
Indexation
6 Average investment return in year 1 3.1% 2.8% 3.5% 3.8% 4.2%
7 Probability of a benefit cut after year 1 0% 0% 0% 0% 0%
8 Long term indexation result of the strategy 0% 0% 20% 40% 60%
(b) The table also shows that the average nominal funding ratio of the strategy that invests for 100% in
the real matching portfolio (strategy 2) is lower than the average nominal funding ratio of strategy
1. Please provide an explanation for this. (4 pts)
The nominal matching portfolio has a higher expected return than the real matching portfolio because
of the inflation risk premium.
(c) Which strategy would you choose? Please motivate your answer by taking into account the interests
of the following stakeholders: the young active participants, the sponsor, the older pensioners (age
appr. 80-85 years) and the younger pensioners (age appr. 67-70 years). (10 pts)
The strategic asset allocation of a pension fund should balance the interests of all stakeholders. The
pension fund is (in most cases) not able to optimize the utility function of all stakeholders. Therefore,
the pension fund chooses a policy in which pains and gains are spread across stakeholders in a way
that is considered to be fair.
9
• Young active participants: have a long term horizon (much longer than 15 years actually).
Hence, they are not so interested in criteria 1-7 but will predominantly focus on indexation
and growth in the long term (criteria 8, 9 and 12)
• Sponsor: is interested in costs (criteria 5 and 11) but also have an ambition for the pension
arrangement they have with their employers. Therefore, they will also look at criteria 8 and
12.
• Older pensioners: are afraid of nominal cuts. They like to have certainty about the money they
are getting paid from the fund. They will look at criteria 2 and 4. Criterion 7 is irrelevant
because there is no difference across strategies.
• Younger pensioners: are afraid of nominal cuts but also of zero indexation (because they want
to maintain purchasing power if the short term risks are acceptable). They will look at criteria
2, 4, 8 and 12.
Looking at the strategies and assessing whether these would be acceptable for the stakeholders:
• Strategy 1: has the best outcome for short term funding ratio risk (criterion 2). This strategy
seems in the interest of older pensioners. Notice, however, that the funding ratio is way
below 100%. A strategy that invests for 100% in the nominal matching portfolio won’t lead to
funding ratio growth and will therefore ultimately lead to a cut in benefits (negative
indexation). Conclusion: strategy 1 is not acceptable
• Strategy 2: is dominated by strategies 1 and 3. Conclusion: strategy 2 is not acceptable.
• Strategy 3: is not dominated by another strategy and has some upside potential and
indexation capacity. Of the remaining strategies this one scores best on criterion 2. The
downside is that the indexation results is rather low. If the fund has many older pensioners
this strategy is defendable. Conclusion: strategy 3 is acceptable
• Strategy 4: is not dominated by another strategy and has some upside potential and
indexation capacity. The upside potential and indexation capacity of this strategy is higher
than of strategy 3. The cost of this is visible in criteria 2 and 13. The extra risk for 20%-points
extra long term indexation result seems small though. Conclusion: strategy 4 is acceptable
• Strategy 5: is not dominated by another strategy and has upside potential and indexation
capacity. The strategy carries more short term and long term risk than strategies 3 and 4. It
depends on the average age of the fund what the ambition is and whether the resulting risk is
acceptable. All in all, we can say that strategy 5 is acceptable.
My personal preference would be strategy 4 as it seems to be the middle ground between strategies 3
and 5. An indexation capacity of 40% in the current situation of a low funding ratio is explainable to
the younger participants while the resulting risk level would be defendable to the older pensioners.
10