Portfolio payments and multiple liability immunization

S

sfad

New member
Joined
May 28, 2013
Messages
0
Reaction score
0
From a past paper, case says:
Client- the defined benefit pension plan for this client has an economic surplus of zero. In order to meet the liabilities for this plan, I will construct the portfolio duration to be equal that of the liabilities. In addition, I will have the portfolio payments be less dispersed in time than the liabilities.
2 questions
1. What does it mean by “I will have the portfolio payments be less dispersed in time than the liabilities”? Does this mean that payments are closer to the horizon date, as opposed to liabilities which are more spread out?
2. The case later says that the client will meet the necessary conditions for a multiple-liability immunization in the case of a non-parallel rate shift.
How can multiple liability immunisation meet conditions in the case of a non-parallel rate shift, don’t we assume parallel rate shifts?
 
Can anyone help with the above please?

Summary:
1. What does portfolio payments less dispersed in time than the liabilities mean?
2. How can multiple liability immunization meet conditions in the case of a non-parallel rate shift, don’t we assume parallel rate shifts for immunization?
 
for multiple liability immunization
a) PV(Assets) should be equal to PV(Liabilities)
b) Composite Duration of the Portfolio (Assets) should be equal to Composite Duration of the liabilities
c) The distribution of durations of individual portfolio assets must have a wider range than the distribution of the liabilities. The third condition requires portfolio payments to bracket (be more dispersed in time than) the liabilities. That is, the portfolio must have an asset with a duration equal to or less than the duration of the shortest-duration liability in order to have funds to pay the liability when it is due. And the portfolio must have an asset with a duration equal to or greater than the longest-duration liability in order to avoid the reinvestment rate risk that might jeopardize payment of the longest duration. This bracketing of shortest- and longest-duration liabilities with even shorter- and longer- duration assets balances changes in portfolio value with changes in reinvestment return.
 
cpk123 wrote:
c) The distribution of durations of individual portfolio assets must have a wider range than the distribution of the liabilities.
Click to expand...
But if “portfolio payments are less dispersed in time than the liabilities” doesnt this mean the assets have less of a range than the liabilities?
 
So does “portfolio payments are less dispersed in time than the liabilities” mean the assets have less of a payment range than the liabilities?
 
The assets should have wider dispersion, not less, around the liabilities. Longest asset durations should be longer than the longest liability duration, and the shortest asset durations should be shorter than the shortest liability duration. Picture a scatter plot.
Was this a true/false exercise in that item set?
 
mellex wrote:
The assets should have wider dispersion, not less, around the liabilities. Longest asset durations should be longer than the longest liability duration, and the shortest asset durations should be shorter than the shortest liability duration. Picture a scatter plot.
Was this a true/false exercise in that item set?
Click to expand...
Yeah this was a true/false in the item set.
Well i’m more comfortable with knowing that assets should have wider dispersion than liabilities, so will disregard the part above.
 
sfad wrote:
Can anyone help with the above please?

Summary:
1. What does portfolio payments less dispersed in time than the liabilities mean?
2. How can multiple liability immunization meet conditions in the case of a non-parallel rate shift, don’t we assume parallel rate shifts for immunization?
Click to expand...
Can anyone answer question 2? How about immunization for nonparallel shifts? Should we just simply use zero-coupon bonds and portfolios that have cashflows concentrated around the horizon? …
Thanks,
 
A blend between cash-flow matching and duration matching.
Which is something along the lines of, cashflow matching on the short term, while meeting the effective duration overall.
That should provide a good comprimise between cost savings on classical immunization, and interest rate risk.
 
I add that curvature of the yield curve is usually larger/ more significant during the first years (when we use the CF matching…).
 
MrSmart wrote:
A blend between cash-flow matching and duration matching.
Which is something along the lines of, cashflow matching on the short term, while meeting the effective duration overall.
That should provide a good comprimise between cost savings on classical immunization, and interest rate risk.
Click to expand...
Great, I’m learning something here, thanks everyone.
 
mellex wrote:
The assets should have wider dispersion, not less, around the liabilities. Longest asset durations should be longer than the longest liability duration, and the shortest asset durations should be shorter than the shortest liability duration. Picture a scatter plot.
Click to expand...
Almost..
The idea that they’re getting at with dispersion is that the closer your assets mimic the liabilities in all facets the less overall risk you have in an Asset-Liability Matching framework. In terms of cash flows, yes the longest duration should be longer than the longest duration of the liability, however, on the other side you want the first cash flow to be as close to the time when you need it to match the liability as possible. Otherwise the longer length of time leads to higher cash flow reinvestment risk.
The reason/clue in the question that says you should immediately focus on ALM is the fact that the economic surplus of the fund is zero. The curriculum extensively covers this concept in the institutional ivnesting IPS section.
 
You must log in or register to reply here.
Back
Top