You have to approach this from an AO risk-return optimization perspective for a surplus efficient portfolio. Most of you are right in disagreeing but The right justification is: T-bills should not be expected to form part of any optimized surplus efficient frontier for a DB plan. Holding T-bills in a surplus is riskier than holding bonds which are more effective in neutralizing surplus risk relative to the liabilities of the plan. Although T-bills themselves as an asset class are less risky than bonds they are a worse match for the liabilities and hence lie below the surplus efficient frontier.
Paraguay, you are spot on!!!
Paraguay Wrote: ------------------------------------------------------- > me.tega Wrote: > -------------------------------------------------- > ----- > > Paraguay Wrote: > > > -------------------------------------------------- > > > ----- > > > me.tega Wrote: > > > > > > -------------------------------------------------- > > > > > > ----- > > > > I agree that T-bills are appropriate > assets. > > > > > > > > One can argue both ways but from the level > of > > > > minimizing surplus volatility, T-bills will > > go > > > a > > > > long way to help in achieving this. If > > interest > > > > rates are high, bonds, equity investments > and > > > real > > > > estate are all going to be negatively > > affected > > > but > > > > t-bills will provide a huge boost for > surplus > > > > portfolio return. A well diversified > > portfolio > > > > should include t-bills in order to improve > > the > > > > risk adjusted return and this is the reason > > we > > > > have a surplus efficient frontier. > > > > > > OK well I am just going to tell you now. > This > > is > > > wrong and against what the book says. > > > > > > You will get this wrong. There is no > argument. > > > > > Interest rate effects are irrelevant due to > the > > > fact the surplus efficient frontier is > matched > > > against a liability with a duration > than a > > > T-Bill. By purchasing the T-Bill you are > > adding > > > risk in not meeting the liability not > > subtracting > > > it. > > > > > > This is the basis of ALM. The thought > process > > you > > > are using is AO. > > > > Are you sure about your argument? Here is what > the > > book says: > > > > Surplus = Asset - Liability > > > > The question CFASniper asked was if t-bills are > > appropriate for an optimized surplus portfolio. > A > > surplus portfolio has no liability to match > > anymore and the only way to look at it is > through > > the lenses of AO. > > Absolutely. To quote Example 32 “The proposed > 60/40 allocation includes a 10 percent weighting > in T-Bills. As Exhibit 32 shows, US T-bills do > not enter into ANY SURPLUS EFFICIENT PORTFOLIO; > including T-bills in the policy mix accounts at > least in part for the 60/40 portfolio not > appearing on the surplus efficient frontier.” > > “By itself, holding long-term bonds is riskier > than holding T-bills, but relative to the pension > liability, T-bills are riskier. The MSV portfolio > is 100 percent long-term bonds. If we want to > move from the MSV portfolio to a high-expected > surplus portfolios, we logically require equities > with a 10% expected return NOT T-BILLS.” > > I would say that is definitive. > > Take position in long-term bonds. Invest surplus > in equities/real estate or higher yielding assets > to generate great surplus. Definitely don’t > invest them in T-bills which have the lowest > return. Okay. Now that was based on the example. The characteristics of that portfolio was based on the risk, return and correlation attributes of that particular example and this does not apply all the time. Run another optimization using Chinese assets with their peculiar correlation characteristics and you will get another efficient frontier that may or may not include Chinese t-bills. The example is not a hard and fast rule. It was just there to help expand your thought process. Let’s look at the concept… a DB plan that has a long term horizon and a large surplus has the ability to take risk but that risk has to be one that will have the best return per unit of risk taken. We are no longer concerned about the underlying liability because it is already adequately covered by the surplus. It is quite obvious that the expected return from using an ALM strategy is generally lesser than using an AO strategy. Now that you have all your bases covered by using why not just increase incremental return per risk taken? You may be right in your argument from another perspective but I can categorically tell you that you are wrong if you say I am wrong.
I guess I am wrong!! 
me.tega Wrote: ------------------------------------------------------- > I guess I am wrong!! > > That is why we have a community. Good debate though. Just from an ALM perspective. It has to be wrong. If you have already met the pension liability and have a surplus to begin with growth of surplus is paramount. The surplus hypothetically should not shrink barring some changes to actuarial assumptions. That is the assumption they are using growth is priority A.
Great discussion - Is there a specific page number for these exhibits? For some reason, I’m having a hard time finding them…
u guys are seriously awesome …
To summarize, T-bill is NOT appropriate for optimized surplus portfolio. However, in MVO, NOT MVS, according to the text (Vol 3. p. 258), it (T-bills) “can be included as a risky asset class with positive standard deviation and nonzero correlation with other asset classes”
I have to bump this thread. I went through CFAI Curriculum Volume 3 Pg 280 - 286 with a fine comb last night and I believe I have to retake my stand that t-bills can form part of an optimized surplus portfolio for a DB plan. The case in point that was used in the argument showing Exhibit 32 was based on a hypothetical optimization using four asset classes. This example was also based on certain assumptions that may not hold in the real world. For example, here in Nigeria, the current one year risk free rate is 9%… this figure was just 2% less than 16 months ago. If I had a surplus 16 months ago, I will not invest in t-bills because of the lower returns. Assuming I had a 20 year liability based on a discounted rate of 7% based on the 20-year bond yields at that time, there is nothing stopping the optimization from throwing t-bills into the mix now that the yields are higher than the liability discount value. Example 17 on Page 305 of the same volume chose a portfolio that has t-bills in a bid to reduce “surplus volatility”. If t-bills have no place in a surplus portfolio then this example is wrong and it is counterintuitive and I don’t know why the curriculum will mislead us this way. And there is nowhere in the text where it says as much. We may have different interpretation about this issue but if I see this kind of question in the exam, I will gladly tick the option that says t-bills can form part of an optimized surplus portfolio. If I get it wrong, I’ll give the three points to CFAI and seek my alpha somewhere else.
xelak73 Wrote: ------------------------------------------------------- > To summarize, T-bill is NOT appropriate for > optimized surplus portfolio. > However, in MVO, NOT MVS, according to the text > (Vol 3. p. 258), it (T-bills) “can be included as > a risky asset class with positive standard > deviation and nonzero correlation with other asset > classes” The optimized portfolio is dependent on the expected return, standard deviation and correlation of the assets that are under consideration. There is nothing that says t-bills will not from part of an optimized portfolio if commodities, hedge funds, private equities and real estate are thrown into the mix. I may be alone on this but I have to dissociate myself from this summary…
MVS is long term capital market expectations. This would be a short term expectation for higher returns in bills. I would never make a long term allocation based on tbills outperforming long term nominals as it has never happened over the long term. Again you are looking at this asset only. This needs to be ALM. To have a theoretical no vol surplus the entire portfolios duration of assets surplus included must be equal to PV liabilities. You are still thinking of this as allocating the portfolio of surplus secondary to the pension assets to meet the liability, which is not correct as tbills have a weak correlation with the pension liability and we are looking for high correlations to pension liabilities. I also completely agree that tbills are included as a risky asset class, just not in surplus maximization. me.tega Wrote: ------------------------------------------------------- > I have to bump this thread. > > I went through CFAI Curriculum Volume 3 Pg 280 - > 286 with a fine comb last night and I believe I > have to retake my stand that t-bills can form part > of an optimized surplus portfolio for a DB plan. > The case in point that was used in the argument > showing Exhibit 32 was based on a hypothetical > optimization using four asset classes. This > example was also based on certain assumptions that > may not hold in the real world. > > For example, here in Nigeria, the current one year > risk free rate is 9%… this figure was just 2% > less than 16 months ago. If I had a surplus 16 > months ago, I will not invest in t-bills because > of the lower returns. Assuming I had a 20 year > liability based on a discounted rate of 7% based > on the 20-year bond yields at that time, there is > nothing stopping the optimization from throwing > t-bills into the mix now that the yields are > higher than the liability discount value. > > Example 17 on Page 305 of the same volume chose a > portfolio that has t-bills in a bid to reduce > “surplus volatility”. If t-bills have no place in > a surplus portfolio then this example is wrong and > it is counterintuitive and I don’t know why the > curriculum will mislead us this way. And there is > nowhere in the text where it says as much. > > We may have different interpretation about this > issue but if I see this kind of question in the > exam, I will gladly tick the option that says > t-bills can form part of an optimized surplus > portfolio. If I get it wrong, I’ll give the three > points to CFAI and seek my alpha somewhere else.
T-bills shouldn’t be included, Stalla guides even have a nice graph explaining this. Commit to memory and move on.