ASA Connect

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.  

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Dear All and Brandy:

Ah TIAA CREF (Now TIAA) is a very complex operation.  TIAA traditional is actually an annuity and insurance plan.  It is very difficult to understand what its underlying value is, though it has one calculated.  If you hold TIAA traditional in a regualr retirement plan, generally one cannot get at it except through 10 payments, which generally takes nine years plus one day.  (There are some exceptions.)  It is very difficult to figure out what it is.  

Assuming any rate will hold is fraught with problems.  Inflation is now roughly zero, but stocks are probably quite overvalued according to Shiller and other commentators.  Shiller PE Ratio

Multpl		remove preview
Shiller PE Ratio Shiller PE Ratio chart, historic, and current data. Current Shiller PE Ratio is 32.56, a change of -0.01 from previous market close. View this on Multpl >

 
But this may be an artifact of the financial crisis, which made earnings lower, etc.  Also the very low interest rates.  So there is no simple way to figure out how to allocate under the current circumstances.  You should also look at the expense ratios of TIAA and other companies, if you have the ability to move funds to the other companies in your plan.

There are a few financial planners (use those who charge by the hour not by the % of your portfolio) who understand TIAA CREF, you might reach out to one of them.

TIAA pays their reps based upon keeping money in TIAA and/or brining in new money.  They no longer claim to be disinterested, but they do not charge commission exactly, though sort of.

I along with my wife are about to retire.  We were lucky in when we moved much of our funds into Stocks and then into Bonds.  

We are still mostly in TIAA, but they have been cutting their expense ratios recently.

You should read a Random Walk Down Wall Street!!!  

Andy

------------------------------
Andrew Beveridge
Professor of Sociology
Queens and Grad Center CUNY
------------------------------

Original Message:
Sent: 12-26-2017 16:49
From: Brandy Sinco
Subject: Retirement Account Mystery Parameter

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Read these about TIAA first before you make a decision.  I would suggest that you get a second, independent opinion. THere have been some rather tough articles about TIAA cref from the NYTimes

https://www.nytimes.com/2017/10/21/business/the-finger-pointing-at-the-finance-firm-tiaa.html?mtrref=www.google.com&gwh=1102A936DB541B25CAEA98148F375D5E&gwt=pay

 

ttps://www.nytimes.com/2017/11/13/your-money/tiaa-403b.html?mtrref=www.google.com&gwh=4865413BF98F146328BFDD6E3C8D7E94&gwt=pay

 

https://www.nytimes.com/2017/11/09/business/tiaa-subpoena.html?mtrref=www.google.com&gwh=945B57645F3B0785B27B903DC049BC9F&gwt=pay

 

TIAA Training Material - The New York Times

https://www.nytimes.com/interactive/2017/11/09/business/document-tiaa3.html

 

https://www.chronicle.com/forums/index.php?topic=253215.0

------------------------------
Ronald LaPorte, Ph.D.
Professor Emeritus Epidemiology
University of Pittsburgh
ronaldlaporte@gmail.com
724 934 9023
------------------------------

Original Message:
Sent: 12-27-2017 08:33
From: Andrew Beveridge
Subject: Retirement Account Mystery Parameter

Dear All and Brandy:

Ah TIAA CREF (Now TIAA) is a very complex operation.  TIAA traditional is actually an annuity and insurance plan.  It is very difficult to understand what its underlying value is, though it has one calculated.  If you hold TIAA traditional in a regualr retirement plan, generally one cannot get at it except through 10 payments, which generally takes nine years plus one day.  (There are some exceptions.)  It is very difficult to figure out what it is.

Assuming any rate will hold is fraught with problems.  Inflation is now roughly zero, but stocks are probably quite overvalued according to Shiller and other commentators.  Shiller PE Ratio

Multpl		remove preview
Shiller PE Ratio Shiller PE Ratio chart, historic, and current data. Current Shiller PE Ratio is 32.56, a change of -0.01 from previous market close. View this on Multpl >

But this may be an artifact of the financial crisis, which made earnings lower, etc.  Also the very low interest rates.  So there is no simple way to figure out how to allocate under the current circumstances.  You should also look at the expense ratios of TIAA and other companies, if you have the ability to move funds to the other companies in your plan.

There are a few financial planners (use those who charge by the hour not by the % of your portfolio) who understand TIAA CREF, you might reach out to one of them.

TIAA pays their reps based upon keeping money in TIAA and/or brining in new money.  They no longer claim to be disinterested, but they do not charge commission exactly, though sort of.

I along with my wife are about to retire.  We were lucky in when we moved much of our funds into Stocks and then into Bonds.

We are still mostly in TIAA, but they have been cutting their expense ratios recently.

You should read a Random Walk Down Wall Street!!!

Andy

------------------------------
Andrew Beveridge
Professor of Sociology
Queens and Grad Center CUNY

Original Message:
Sent: 12-26-2017 16:49
From: Brandy Sinco
Subject: Retirement Account Mystery Parameter

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Great modeling topic!

This is something that I have given lots of thought to over the years as I have now guided our parents, our children, and ourselves on these matters. I’m not a professional financial advisor but I just want to point out the many random aspects that we all need to consider in making our personal decisions.

I think the modeling problem is much more complex than a linear equation.

There are multiple statistical aspects. First, the key unstated point regarding any retirement plan is the power of compounding, where the rate of stock appreciation (or depreciation) is random. Second, future employer contributions are never guaranteed; back in the 70’s, my employers contributed 10%+ which if allocated 100% to stock allowed the maximum appreciation potential. Third, our current and future lifestyle needs are all deterministically different; the reduced 3-5% employer annual contributions forces us to save the personalized gap out of taxable income which means thinking about long term retirement needs with inflation being random; it has varied widely from 2% to 10% over the last 40 years. Fourth, remember that your individual life expectancy must also be considered so you don’t run out of money; we are living longer than 40 years ago and that could further increase when you are ready to retire. Last, while stocks are riskier than annuities, the expected long term returns on stocks have usually been better than annuities, but we all have different appetites for risk; there are so many investment options as well to consider as well as when to switch investments.

So there are multiple factors to model, best accomplished after setting and revisiting personal financial goals and getting access to a financial advisor who you trust. My hunch is that there is no single model that will serve us all.

Ideally, if you save enough, there will be less pressure to reallocate to fixed income securities as you near retirement.

In closing, as I recently shared with our children, it’s never too early to think about your retirement so try to revisit your goals and evaluate your progress towards those goals at least once a year.

Happy new year.

Phil Lavin

Sent from my iPhone


Original Message------

Read these about TIAA first before you make a decision.  I would suggest that you get a second, independent opinion. THere have been some rather tough articles about TIAA cref from the NYTimes

https://www.nytimes.com/2017/10/21/business/the-finger-pointing-at-the-finance-firm-tiaa.html?mtrref=www.google.com&gwh=1102A936DB541B25CAEA98148F375D5E&gwt=pay

 

ttps://www.nytimes.com/2017/11/13/your-money/tiaa-403b.html?mtrref=www.google.com&gwh=4865413BF98F146328BFDD6E3C8D7E94&gwt=pay

 

https://www.nytimes.com/2017/11/09/business/tiaa-subpoena.html?mtrref=www.google.com&gwh=945B57645F3B0785B27B903DC049BC9F&gwt=pay

 

TIAA Training Material - The New York Times

https://www.nytimes.com/interactive/2017/11/09/business/document-tiaa3.html

 

https://www.chronicle.com/forums/index.php?topic=253215.0

------------------------------
Ronald LaPorte, Ph.D.
Professor Emeritus Epidemiology
University of Pittsburgh
ronaldlaporte@gmail.com
724 934 9023
------------------------------

Hi Brandy:

Your approach makes sense if you look only at the mean return, but it
doesn't take into account the volatility. If you Google "mean-variance
theory" or "Modern Portfolio Theory" you'll see what I mean. If you
take the market volatility into account and make some parametric
assumptions (normal returns) you can obtain the entire distribution of
future return as opposed to considering just the worst case scenario.
For instance, if the market crash probability is 1E-10, it certainly
makes sense to allocate 100% to the market, but your solution is
insensitive to that.

Regards,
Nik


Original Message------

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.  

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Original Message------

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.  

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Brandy

There is nothing wrong with your math but I think you may be doing yourself a disservice
by in effect using a static model for a very dynamic underlying process. I think you can probably get away
with using static assumptions for the guaranteed rate  "T" and even for the inflation rate "I" but not
the rate of return "M" pegged to the market. In your post you computed a 2.8% if the market
crashed. This assumes  that whatever proportion you allocate to the non-guaranteed portion
(1-lambda) literally goes to 0 and stays at  0. But this is a highly highly unlikely outcome. Yes markets crash
but they also rebound and then correct/crash again. If as is probably the most prudent course you invested the non-guaranteed
portion in ETFs or index funds it would take a fairly cataclysmic event to bring those funds to 0 and then to stay at 0.

This problem cries out for a simulation approach allowing for at least your M parameter to vary by assigning
some appropriately chosen PDF. Of course if anyone knows how to choose the "correct" PDF to reflect the
expected variation in the market indexes over the next say 30 years then that person is very smart and presumably 
 poised to become very wealthy. Yet even admitting how difficult it is to choose a PDF I think choosing a simulation
model over the next 30 years is a better decision approach than using a static model that assumes falsely that the market can crash completely.

Also your Lambda model does not take into account risk tolerance which is a highly personal matter. Also it does not take into account other sources of potential liquidity for retirement years such as an equity in a personal residence. But this is more of a quibble.




------------------------------
Michael Sack Elmaleh
Principal
Michael Sack Elmaleh CPA, CVA
------------------------------

Original Message:
Sent: 12-26-2017 16:49
From: Brandy Sinco
Subject: Retirement Account Mystery Parameter

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Hello Brandy Sinco, let me tell you my experience many years ago with
TIAA and CREF.?? I did a 50:50 split.?? When it came to "claim" money it
was easy with CREF, in fact I moved it to an IRA for various reasons.

For TIAA?? I had no control or flexibility with the money,?? There were
annual payments for a fixed number of years until the contract was
fulfilled. ?? What I did not appreciate at the time is that TIAA is
Teachers INSURANCE ANNUITY Association.?? With a TIAA you buy an
insurance annuity contract and then you are locked in to the terms of
the contract.

You may wish to find yourself a FEE based financial advisor to factor in
your multi variables specific to YOUR situation.

John Bartko, retired Bio Statistician,


Original Message------

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.  

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

If you are concerned about protecting against a market crash, you might be better off investigating a parameter, like the Shiller 10-year inflation-adjusted S&P500 price earnings ratio that was previously suggested, to predict the likelihood of a crash occurring and adjust your proportion of stocks and bonds accordingly as that parameter changes.

The Shiller SP500 is currently about 32.6, indicating the market is overvalued and subject to a correction. It has been over 30 only twice before in the last century, just before the crash of 1929 and during the dot com boom in the late 1990s. An approach like this would decrease the proportion of stocks when the market is overvalued and increase it when it is overvalued.

Of course such an indicator does not predict the market completely accurately. For example, it stayed above 30 for several years during the 1990s and reached 40 before the crash. But it has had approximate accuracy so far. If one is concerned with the risk of retiring right when a crash occurs, it would seem a better way to adjust for this risk than using a single fixed lifetime ratio that doesn’t permit any adjustment based for overall market value.

Jonathan Siegel
Associate Director Clinical Statistics

Sent from my iPhone


Original Message------

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.  

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Thanks to everyone who took the time to read my post and offer their insights.

One person replied to me privately and said that λ = (Age in years)/100 is a rule of thumb for
proportional allocation to a fixed annuity versus market-based funds.

As suggested, I will explore the Shiller index and experiment with Monte Carlo simulations.  Although, my spreadsheet with various values of T, M, and I generated values for λ that looked reasonable to a financial advisor.

Personally, I am happy with TIAA and am glad that my employer offers this option.  Some have commented that taking money out of TIAA-traditional is difficult.  That's  because TIAA-traditional is intended to be a lifetime annuity, not a savings account.
https://www.tiaa.org/public/offer/products/annuities/retirement-plan-annuities/tiaa-traditional-annuity

I also appreciate the social choice (socially responsible) mutual funds and life cycle funds that TIAA offers.  Life cycle funds are designed to become more conservative as the client ages. In 2018, TIAA will celebrate its 100th anniversary.

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------

Original Message:
Sent: 12-26-2017 16:49
From: Brandy Sinco
Subject: Retirement Account Mystery Parameter

Dear ASA Colleagues:

Some day, I hope it will be financially possible to retire.  My retirement plan at work is TIAA-CREF.  One option is TIAA-Traditional, which has a guaranteed rate of return of 3%, although it's usually between 3% and 4%.  With TIAA-Traditional, it's impossible to lose money unless TIAA-CREF goes out of business.  Other options are various mutual funds, which usually has higher rates of return, but are vulnerable to downturns in the market.  According to the TIAA-CREF literature, they do not recommend 100% in TIAA-Traditional because that makes a person vulnerable to the inflation risk; nor do they recommend 100% in mutual funds because that risks losing money in an economic downturn.  So, how does a person decide what proportion to allocate to TIAA-Traditional, the investment option with a lower bound on the rate of return?

Here's my procedure.

Let λ = Proportion of funds invested in the guaranteed option, aka the Retirement Account Mystery Parameter.
Let T = Rate of return of guaranteed option.
Let M = Rate of return from mutual funds.
Let I = Inflation rate.

As age approaches retirement, choose L such that

λT + (1 - λ )M ≥ I.

==> λ  ≤ (M - I)/(M - T)., provided that M > T and M > I.

For example, if T = 3.5, M = 6, and I = 4, then λ = .8.  So, 80% in the guaranteed fund will protect the retiree from inflation loss, provided that inflation is 4% or lower, the weighted average of the guaranteed rate ≥ 3.5%, and the market rate ≥ 6%.  If the market crashes, λ = .8 has a lower bound of .8*3.5% = 2.8%.

I'm thinking that for someone ~65.  λ  ∈ [.7, .8].  For younger person, choose a smaller  λ, assuming an increase of .01 for each year of age, until the person approaches retirement age.

Does anyone use a procedure similar to why I have described above?  Does anyone have an alternative algorithm to calculate λ as a function of age?

------------------------------
Brandy Sinco, BS, MA, MS
Statistician and Programmer/Analyst
------------------------------