Secondary menu

Sensitivity Analysis For Inflation Rates

Submitted by jdupton on 24 August 2009 - 9:52pm

Anyone,

I have been running a bunch of simulations of various inflation rate assumptions to see the effects on five different retirement strategies. I have come across something I cannot really explain and wanted to see if others may have seen similar results.

I have five scenarios I run with different assumptions regarding pension vs partial lump sum distribution, possible buyout packages, and social security options. I had been doing most base planning at an assumed inflation rate of 2.5% to 3.5% but decided I should see how various inflation rates affect each of these strategies.

I ran simulations against each strategy for inflation rates of 2.5% up to 7.0% by 0.5% increments. I later went back and added a few simulations in between at 0.25% increments. The results were both informative, insightful, and also very puzzling.

Two of my strategies followed one slope as inflation increased and three followed another. These differences I have found were due to the effects of a fixed (non-indexed) pension. The scenarios which relied less on the fixed pension fared better in the face of inflation as would be expected.

Now for the puzzling part.

There is a large and very sharp discontinuity in the curves between the inflation rates of 5.0% and 5.25% that I am at a loss to explain. In general my base recommended consumption declined at a rate (slope) of about 3%/percent of inflation from 2.5% to 5.0% inflation rates. From 5.5% to 7.0% inflation, the slopes were again at about 3%/%inflation. However the curves between 5.0% and 5.25% drop at a rate (slope) of more than 50%/%inflation. (Think of a curve with a gentle downward slope up to the 5.0% mark, dropping precipitously, and then resuming the gentle downward slope after the 5.25% mark.)

I am left wondering what might be happening at around 5.0% inflation. At first I thought this was a side effect of having a substantial portion of my assets in lower yielding fixed income accounts and this was the point at which inflation bypassed the earnings rate for that portfolio. I don't know that this would cause such a large effect at a particular point, though.

Any other places to look to explain what's happening?

John

‹ 2011 Consumption Consumption in Today's Dollars ›
Submitted by Laurence Kotlikoff on 25 August 2009 - 10:34am.
1

Hi John,

We have a very non-linear tax system, which is anything but neutral to inflation. So you could be getting pushed into higher tax brackets because of higher nominal asset income (Uncle Sam taxes nominal, not real, asset income). So I'm not surprised that you are seeing this type of thing. I'd be surprised if you weren't.

best, Larry

Submitted by jdupton on 25 August 2009 - 11:03am.
2

Hi Larry,

Thanks for the additional pointer. It makes sense.

I may be able to test this by running a couple of simulations with Assets and Pension values set 10% to 20% lower than they really are. If the effect is that of getting whacked with extra taxes, it should move the discontinuity out to a higher inflation rate than it is now. In other words, with a lesser starting standard of living due to reduced assets, it will take a higher inflation rate to trigger the increased tax bill.

I'll run a few more simulations and report back.

The graph I did last night is shown at this link:
http://www.atm-workshop.com/misc/Inflation_Effects.gif

John

Submitted by jdupton on 25 August 2009 - 8:43pm.
3

Larry,

I am still a bit confused on what's happening. I went into my profile and reduced my assets by about 20%. I also reduced the pension income by about the same amount. Upon rerunning simulations for inflation rates of 4.75%, 5.0%, and 5.25%, I still see the same precipitous drop in living standard at the same 5.25% inflation rate point as if there is something magic about 5.25%. If this were caused by higher tax rates kicking in, I would have expected the drop point to move upward by a bit. I would also not have expected such a great drop in a very narrow range of inflation rates.

I guess my next step is to output a couple of these as Excel worksheets and see if anything jumps out with respect to taxes.

Do you have any other suggestions for areas to investigate the discontinuity of the curves?

John

Submitted by jdupton on 25 August 2009 - 9:18pm.
4

Hi again, Larry,

Sorry for the multiple posts here but this doesn't appear to be tax related. I cannot definitively figure it out. I ran more simulations and found that the difference in living standard is coming from my retirement assets income.

I narrowed the range of inflation that caused the sharp drop in living standard. Between the inflation rates of 5.15% and 5.20%, the annual income from my retirement assets drops by about 40%. This is what is causing the drop in living standard. Taxes appear to be in line with the tiny inflation rate change and follow the income drop downward rather than increasing.

I will add this to my support ticket for the other inflation rate change error I found. The two may be closely related.

John

Submitted by AnalisysGeek on 24 February 2011 - 10:05am.
5

Tax system was developed that way, that no one, except for professionals could understand what's going on. From some point of view it makes sense. Anyway, if you're in the dead end - get algorithms assignment help. This guys are really good and will help you.

Share this

Navigation