Q: Describe the primary characteristics of pension investments that would be considered low risk under AO
My answer : AO - A low risk pension investment plan in an AO approach would have a small standard deviation portfolio of assets.
CFAI (2012): AO-.primary characteristics of low-risk investments would be low correlation with the Plan’s assets. Under this approach , the focus is on creating efficient frontire portfoliosl therefore low-risk investments are those that have low correlations with plan assets.
I understand CFAI’s point, that the goal is to creating an efficient frontier…but I feel as if my answer holds ground just as much as theirs. How could a Pension investment NOT be low risk when it has a small(er) standard deviation. In an MVO / AO framework risk defined by Standard deviation?
The question doesn’t mention anything about having the highest Risk/Return profile which would give me a clue that they are looking for an optimal portfolio…they don’tmention that this “investment” would be an add on.
Am I missing something here as I feel like my answer wouldn’t get credit during D-day as they would be looking for correlations…
even if you had low standard deviation of the plan assets - that portfolio could have a high correlation with plan assets. you could have a situation where the plan assets tank and so does your portfolio.
A low correlation with plan assets ensures that your pension plan does not suffer if your plan assets do…
Yes, it’s arguably more important than standard deviation alone.
Diversification maximizes return for risk as per MPT, even if you have low standard deviations (with let’s say, one asset, or a whole bunch of assets that are perfectly correlated), you may have limited risk exposure, but your retrurns are not optimal given the risk profile, compared to another diversified portfolio with the same total portfolio standard deviation.
Another way to put it is, low risk plans can have high individual standard deviations, that alone makes your statement questionable and should knock off some points.
So is what you are saying is that if I had a risk-free asset as an investment, and all my plan assets were also all risk-free assets, well then I have a risky portfolio?
I think your answer highlights my point that standard deviation is the primary characteristic that defines low-risk…not correlation as in my example, correlation was perfectly +1 while standard deviation was zero…and I am sure we can both agree that an investment that is risk-free coupled with pension assets that are risk-free, would also be risk-free despite having perfectly +1 correlation…
Yes, but a risk-free portfolio is not exactly a ‘plan’ now is it?
You’d likely have 5-20% as treasuries. What about the non-zero standard deviations?
By the way, there is no such thing as a zero standard deviation security in practice. You still take the standard deviation and the correlation of treasuries with the rest of the portfolio, although they should exhibit the least of both. So it doesn’t apply in any case.
Just understand that standard deviation alone does not represent risk, and the correlation between assets determines the final risk exposure, stating a low standard deviation is only half the answer.
I agree, but the problem is, that’s actually a CFA question (2012). Not to belabor the point, but this is taken straight from the book (SS6):
From an ALM perspective, the characterization of risk in the IPS needs to be stated in relative terms. The emphasis shifts from the expected volatility of pension assets to the expected volatility of pension surplus and to probabilities concerning expected levels of funded status over appropriate time frames.
Having this new information, I would say that the answer should be “a low risk investment is one that decreases the volatility of pension surplus.” It’s clear how correlation plays a vital role, but what I think is troubling is that the answer key does not mention volatility when asked for risk. I wonder if I put this exact word for word answer from the book if I would get this wrong because I did not say “correlation.”