Question on the Otar calculator

[Ken11]

Using historical/"look back" versus "plan ahead" mode in the portfolio section produces very different results. The user guide states the following:

"Plan Ahead" mode should be used in all retirement planning calculations. Why? Because, prior to1990, the average dividend yield was about 4.5%. Nowadays it is around 2%. If you use the "Look Back" mode you would be building-in unrealistic return rates for your retirement planning.

So the "plan ahead" model ignores many decades of performance data and assumes that the recent poor performance will apply forever? Hmmm. Was this modeled before or after the 2008 crash? Given the crash are more aggressive future returns rational to assume? Transparency to the detailed assumptions is not provided. In the look forward model exactly what total returns are assumed?

Without a better understanding of these very key assumptions I'm not sure how to interpret the results from the plan ahead modeling..

Has anyone gotten into this?

 

[RonBoyd]

Quantext Portfolio Planner is a "plan ahead" device that is useful toward answering your type of question. The creator, Geoff Considine, explains his method in the eBook, "Survival Guide for a Post-Pension World."​

Topic 1: Looking Forward vs. Looking Back
Quantext Portfolio Planner (QPP) takes historical data on the performance of risky
assets and an assumption for the equity risk premium, and uses these to calculate
expected return and expected risk for individual assets and for the total portfolio. We
will be going into detail later, but for now we can simply think of QPP as an engine that
calculates expected return and expected risk for a portfolio. QPP also calculates how the
volatility of a portfolio, combined with other aspects of our plans, determines how much
income we can plan to safely draw in income from our portfolios.​

There are many ways that we might come up with estimates for the expected future return
and risk of a portfolio. The most simple is to look at the historical risk and return. What
is the historical average annual return for a portfolio? What is the historical risk level?
These are useful statistics to know, but they do not provide a good basis for planning.
We need something more. The solution to this problem is to come up with what are
referred to as forward-looking models. These are statistical models that are specifically
designed to generate reasonable estimates of the risk and return for the various assets that
we might want to include in our portfolio. These estimates are the expected risk and
return. An example will help to motivate why this is important. QPP is a forward-looking
model for portfolio planning.​

Using QPP to analyze the 60/40 portfolio vs. the most recent ten years of data is a useful
way to illustrate how expectations differ from trailing performance.​

Historical vs. Expected Risk and Return (10 years of Data Through 2009)​

When we look at these data, there is obviously a massive contrast between the trailing
and expected future returns. For the last ten years, bonds have soundly thrashed stocks.
Looking forward, stocks are expected to beat bonds. If we were making plans on the
basis of the last ten years, we would just buy bonds. Over this period, the bond fund
(VBMFX) has generated more return, with less volatility, than stocks or a 60/40 mix of
stocks and bonds. Why bother with stocks at all? The reason, of course, is that we
expect that stocks will ultimately return more than bonds.​

The first question that a reader will naturally ask is why they would believe the modelgenerated
expected values. On the other hand, why would we believe that the past ten
years (or five years, etc.) is a reasonable basis for predicting future returns? Using a
forward-looking model forces us to think about where our expected future returns are
coming from---we all use a model, even if that model is simply to assume that the future​

will be like the past.

(Page 27)

 

[W2R]

Using historical/"look back" versus "plan ahead" mode in the portfolio section produces very different results. The user guide states the following:

"Plan Ahead" mode should be used in all retirement planning calculations. Why? Because, prior to1990, the average dividend yield was about 4.5%. Nowadays it is around 2%. If you use the "Look Back" mode you would be building-in unrealistic return rates for your retirement planning.

So the "plan ahead" model ignores many decades of performance data and assumes that the recent poor performance will apply forever? Hmmm. Was this modeled before or after the 2008 crash? Given the crash are more aggressive future returns rational to assume? Transparency to the detailed assumptions is not provided. In the look forward model exactly what total returns are assumed?

Without a better understanding of these very key assumptions I'm not sure how to interpret the results from the plan ahead modeling..

Has anyone gotten into this?

"All models are wrong. Some models are useful." - - George E.P. Box

I take the above quote to heart. As a modeler in my former career, I would never have based my ER only on the results of any one model. I preferred to look at the results of a lot of models, mainly to make sure that my ER plan was nearly always considered to be a successful plan.

Given the changes in investment returns that you mention, economic changes, and other changes just in our lifetimes, I think we need to keep in mind that past performance is no guarantee. Historical market performance gives us a frame of reference and these model results can be useful from that perspective, but I would tend to (and did) approach ER from a more pessimistic point of view than model output might provide. We need to remain as flexible as possible in ER, as we see how things progress.

 

[Ken11]

Thanks for the replies.

What I found frustrating was that the specific assumptions underlying the look forward model are not provided. Exactly what return(s) are used by Otar?

Of course I don't rely on any one model...if you only knew how many numbers I crunched before ESR'ing.. I'm glad we only make this decision once in a lifetime.

While near term returns may not look like the recent past it could be argued that, in the long run, the past is the best guide to the future. Why apply conservative (i.e below long term average)assumptions that may be appropriate for the next few years to a model that goes out 30+ years?

Also... what is done about inflation in the more conservative scenario?

I guess I'll need to go check some formulas to see whats going on in the spreadsheet.

 

[walkinwood]

Ask Otar. jim@retirementoptimizer.com
Please share what you learn. Thank you.

 

[Ken11]

Ask Otar. jim@retirementoptimizer.com
Please share what you learn. Thank you.

Tx for the contact.

I just purchased his book,
I suppose I'll read it before bugging him.

Cheers

 

[walkinwood]

I recall reading that he uses historical data for the index values, but the less generous 2% dividend for forward looking projections. But, I can't seem to find where he says that in the book. But looking through it, I feel that I should read it again - it is very dense with information.

 

[Ken11]

I recall reading that he uses historical data for the index values, but the less generous 2% dividend for forward looking projections. But, I can't seem to find where he says that in the book. But looking through it, I feel that I should read it again - it is very dense with information.

It tuns out that this is correct. (Jim confirmed this to me in an e-mail).
Taking more than 2% of dividends out of the equation and not replacing it with any incremental capital appreciation is certainly a very conservative assumption.
The historical data already includes extended periods of no/low returns.

However, this roughly corresponds to adjustments my FA made.
Prior to the 08/09 crash he assumed 7.5% returns in the long term.
He now assumes 5.5%.

 

[haha]

Ask Otar. jim@retirementoptimizer.com
Please share what you learn. Thank you.

In his articles, Otar states that if you have 33* your needed annual withdrawal, you are in the green zone. I believe he places the zone betwen 20* and 33* as the gray zone, and lower than 20* the red zone where failure risk is very high. This is for a normal mid 60s retirement age.

He says his calculator is based on past market performance. As is Firecalc

Why do his survival rates seem considerably more pessimistic, and his recommended maximum withdrawal rate quite a bit lower than Firecalc's? (.03333 vs. .04).

I notice that he does have a business giving speeches, etc. mostly to insurance agents, so this is likely a most welcome message to these folks who are the gateway to his recommended products- CPI indexed annuiities. Still, it seems very unlikely that this played a big part in his formulation.


Ha

 

[walkinwood]

In his articles, Otar states that if you have 33* your needed annual withdrawal, you are in the green zone. I believe he places the zone betwen 20* and 33* as the gray zone, and lower than 20* the red zone where failure risk is very high. This is for a normal mid 60s retirement age.

He says his calculator is based on past market performance. As is Firecalc

Why do his survival rates seem considerably more pessimistic, and his recommended maximum withdrawal rate quite a bit lower than Firecalc's? (.03333 vs. .04).

I notice that he does have a business giving speeches, etc. mostly to insurance agents, so this is likely a most welcome message to these folks who are the gateway to his recommended products- CPI indexed annuiities. Still, it seems very unlikely that this played a big part in his formulation.


Ha

Otar uses a 2% dividend, but historical index values in determining is "aftercast". His take is that dividends were a significant part of total return in the past and they are minuscule today and likely to stay that way in the future.That's why it is so pessimistic.

 

[haha]

Otar uses a 2% dividend, but historical index values in determining is "aftercast". His take is that dividends were a significant part of total return in the past and they are minuscule today and likely to stay that way in the future.That's why it is so pessimistic.

Thank you. Perhaps that is his back-door way to take valuations into account.

Ha