This does indeed reduce the net coverage + gap cash to a much smaller number for the case of no yield change.You wrote this question while I was writing the reply I just posted, and I think the answer contained therein is "perhaps".Is it somehow wrong to say that if you buy an certain amount of excess 2054 holdings, and then find that you have 100K left over after the swap, that maybe you bought too much to begin with? Maybe we need to revisit the criteria for buying the excess holdings?
Not too long ago I went down a rabbit hole by tweaking the 2035 duration match weights so that there was exactly 3M total ARAs, and then when I got to the third rung it got too tedious to try an optimize the multipliers. But perhaps we could revisit tweaking the initial multipliers, and then just use linear interpolation to decrease the amounts so they all end up at 1, or something like that. Would be easier using a single holding to do this, like the 2054 coverage case.
Through trial and error I found that an initial multiplier of 4.92004 for the 2054 got total ARAs to 3,000,000. Reducing that by 0.78 per year gets it to 1 when all gaps are filled. Here is the table with all gaps filled:
So the magnitude of the delta is much smaller with no yield change. On the other hand, we need some extra cash to do just the coverage/gap swaps, and a small amount after all swaps.
However, the deltas (negative cash) are larger for the case where yields increase from 2% to 4%:
Also, the positive cash deltas are much smaller for a yield decrease from 2% to 0% than if we start with multiplier 6 for the 2054:
So starting with an excess of 2054s seems to have generally more favorable outcomes for the yield change scenarios. Is there a limit to this?
Statistics: Posted by Kevin M — Sat Jun 22, 2024 11:28 pm — Replies 189 — Views 15224