Letters to the Editor
Updated: October 15, 2005.
I Bonds versus TIPS
Greg asks: "How do you feel about using I bonds in place of tips to develop a retirement portfolio?"
This can be a good choice. Most likely, you would be better off by investing in both.
Initial Baseline
I have extracted this baseline information from Always insist on a baseline. TIPS currently yield 2% (plus inflation). I Bonds yield 1.2% (plus inflation). These numbers provide you with a worst case comparison based on interest rates alone.
Always Insist on a Baseline
You can use the mortgage formula (or mortgage tables) for actual interest rates. Here are some numbers. They are in terms of real dollars:
1) With an interest rate of 1.0%, you can withdraw 3.87% for 30 years or 3.40% for 35 years or 3.05% for 40 years with total safety.
2) With an interest rate of 1.5%, you can withdraw 4.16% for 30 years or 3.69% for 35 years or 3.34% for 40 years with total safety.
3) With an interest rate of 2.0%, you can withdraw 4.467% for 30 years or 4.00% for 35 years or 3.66% for 40 years with total safety.
These percentages are large because you are withdrawing principal as well as interest.
Expanded Baseline
I have used the information in Equations for Design to construct the following tables related to interest rates of 1.0% and 1.2%.
Equations for Design
The TESWR (TIPS Equivalent Safe Withdrawal Rate) tells you how much you can withdraw from a TIPS and/or I Bond portfolio for exactly a specified number of years. You can use this information directly. You can also use TESWRs to calculate how much money remains after making withdrawals for a specified number of years.
TESWRS for 1.0% TIPS and I Bonds
10-Year TESWR: 10.56%
15-Year TESWR: 7.21%
20-Year TESWR: 5.54%
25-Year TESWR: 4.54%
30-Year TESWR: 3.87%
35-Year TESWR: 3.40%
40-Year TESWR: 3.05%
45-Year TESWR: 2.77%
50-Year TESWR: 2.55%
Remaining fractions for 1.0% TIPS and I Bonds
With 1.0% interest, when withdrawing 3% for 10 years: 79.1%.
With 1.0% interest, when withdrawing 3% for 15 years: 67.8%.
With 1.0% interest, when withdrawing 3% for 20 years: 55.9%.
With 1.0% interest, when withdrawing 4% for 10 years: 68.6%.
With 1.0% interest, when withdrawing 4% for 15 years: 51.7%.
With 1.0% interest, when withdrawing 4% for 20 years: 33.9%.
TESWRS for 1.2% TIPS and I Bonds
10-Year TESWR: 10.67%
15-Year TESWR: 7.32%
20-Year TESWR: 5.65%
25-Year TESWR: 4.65%
30-Year TESWR: 3.99%
35-Year TESWR: 3.52%
40-Year TESWR: 3.16%
45-Year TESWR: 2.89%
50-Year TESWR: 2.67%
Remaining fractions for 1.2% TIPS and I Bonds
With 1.2% interest, when withdrawing 3% for 10 years: 81.0%.
With 1.2% interest, when withdrawing 3% for 15 years: 70.6%.
With 1.2% interest, when withdrawing 3% for 20 years: 59.6%.
With 1.2% interest, when withdrawing 4% for 10 years: 70.4%.
With 1.2% interest, when withdrawing 4% for 15 years: 54.2%.
With 1.2% interest, when withdrawing 4% for 20 years: 37.1%.
Augmented Baseline
I augment this baseline further by applying TIPS Ladders for Today. Simply replace the 10-year TIPS ladder with I Bonds.
TIPS Ladders for Today
The basic strategy behind TIPS Ladders for Today is to vary allocations of TIPS and stocks according to valuations (i.e., P/E10). The algorithm normally has a TIPS allocation of 75% or 100%. Only if stocks were to become really attractive (with P/E10 levels below 9) would the TIPS allocation fall as low as 50%.
This approach would raise the Safe Withdrawal Rate (i.e., with a 95% probability of success) to 4.0% over 40 years. This was based upon TIPS interest rates of 1.8%. (Today, TIPS 10, 20 and 30-year rates are very close to 2.0% on the secondary market.)
The approach has you invest 50% of your money in long-term (20 and 30-year) TIPS. At least one-half of the remainder goes into a 10-year TIPS ladder. You invest whatever is left over in 10 or 20-year TIPS or high dividend stocks.
Simply replace the 10-year TIPS ladder with I Bonds. This is your augmented baseline.
There should be a slight adjustment to withdrawal rates. It consists of an increase (since today’s long-term TIPS now yield 2.0% instead of 1.8%) and a decrease (since you would replace 1.8% 10-year TIPS with 1.2% I Bonds).
A comparison of 10-year remaining fractions using 1.2% and 1.8% interest rates suggests lowering the I Bond withdrawal percentage from 4.0% to 3.7%. But only 25% (or 50% at the most) of portfolio would be in I Bonds. Even without making the adjustment to increase the 1.8% TIPS interest rate to today’s 2.0%, you could withdraw 3.9% of your total portfolio’s initial balance (plus inflation) for 40 years (with a 95% probability of success).
Other Considerations
You are limited in how much you can invest each year into I Bonds. The maximum is $60000 per year (with $30000 electronic and $30000 paper) per social security number.
You may not sell I Bonds for one year. You may sell at any time after one year, similar to a traditional bank savings account. (There is a minor interest penalty during the first five years.) You can buy and sell I Bonds only with the Federal Government. There is no secondary market. You will never have any capital gains or losses.
The Federal Government sets I Bond interest rates. You must take them or leave them. Nobody has the ability to negotiate rates. There is no auction.
I Bond interest rates (that is, the fixed part of the interest rates) stay in place for 30 years. I Bonds have interest rates similar to those of 5-year TIPS, but you may extend their maturity indefinitely. If (the fixed part of) I Bond rates go up, you can replace your old I Bonds with new I Bonds at the new rates (subject to your annual purchase limit).
I Bonds should be held in taxable accounts. They have a built in tax deferral (unless you choose otherwise). This leaves the tax shelter features of your standard retirement accounts (IRAs, 401ks, etc.) in full force.
I Bonds are a great choice for an emergency fund (after the one-year waiting period).
I Bonds can be a good choice if you expect to pay post-secondary educational expenses. This is not limited to colleges. However, there are limitations, including income limitations. To the extent that you qualify and use I Bonds to pay tuition, you do not have to pay taxes on the interest.
I Bonds and educational expenses:
US Savings Bonds: General Information
Summary
Using I Bonds can make sense.
There are lots of factors that are unique to individual circumstances. For example, limitations on purchases may limit your I Bond total at the time of retirement. There are limitations in how much that you can invest in tax sheltered accounts. This may cause you to limit your purchases of TIPS. Speculative factors may cause you to shift funds to lock in a favorable TIPS interest rate.
Should You Buy an Annuity?
Mike asks: “The Vanguard site says $493.85 monthly, fixed annuity, 100k invested.
Do Gummy's numbers mean that I would need about $700/month to make the annuity worth while?”
Mike’s question is related to the article: Should You Buy an Annuity?
Should You Buy an Annuity?
Missing Information
Gummy’s equations showed that inflation should not directly influence your decision whether to buy an annuity. It does enter into decision making indirectly.
The basic tradeoff is how the annuity withdrawal rate compares with your Safe Withdrawal Rate. The power of statistics favors the annuity. An insurance company is able to base its payments on a narrow amount of time around life expectancies. You have to plan for your BEST case, which is that you live longer than your life expectancy. Typically, the insurance company makes its calculations based on 20 years (or so) while you must be able to finance 30 or more years.
Almost all of our Safe Withdrawal Rate information is based on REAL dollar amounts. That is, we adjust our numbers for inflation. Sometimes we vary the dollars that we withdraw, but we report our results in terms of today’s buying power.
Gummy’s equations tell us to use Safe Withdrawal Rate numbers that are based on NOMINAL dollar amounts. (I am assuming that Mike’s numbers are in constant NOMINAL dollar amounts.)
I used my Deluxe Calculator Version V1.1A08 to collect data for some simple comparisons. I determined the REAL and NOMINAL withdrawal rates for the first failure and the sixth failure for 30-year historical sequences beginning in 1921-1980 (sixty sequences). I used 50% stocks and 80% stocks. I used commercial paper and 2.0% I Bonds for the component other than stocks. I set expenses at 0.20%. I varied withdrawal rates in increments of 0.1%.
With 50% stocks and 50% commercial paper, the 30-year REAL rates were 4.0% for no failures, 4.1% for two failures and 4.3% for 6 failures. The 30-year NOMINAL rates were 4.9% for no failures, 5.0% for one failure and 6.7% for a minimum of 6 (actually, 7) failures.
With 80% stocks and 20% commercial paper, the 30-year REAL rates were 3.8% for no failures, 3.9% for one failure and 4.4% for 6 failures. The 30-year NOMINAL rates were 4.4% for no failures, 4.5% for one failure and 7.1% for 6 failures.
With 50% stocks and 50% I Bonds (at 2% interest), the 30-year REAL rates were 4.2% for no failures, 4.3% for two failures and 4.6% for a minimum of 6 (actually, 7) failures. The 30-year NOMINAL rates were 5.0% for no failures, 5.1% for one failure and 6.5% for 6 failures.
With 80% stocks and 20% I Bonds (at 2% interest), the 30-year REAL rates were 3.9% for no failures, 4.0% for one failure and 4.5% for 6 failures. The 30-year NOMINAL rates were 4.4% for no failures, 4.5% for one failure and 7.0% for 6 failures.
Taking ratios, we find that the NOMINAL withdrawal rate at first failure ranges between 1.13 and 1.25 times the REAL withdrawal rate. The NOMINAL withdrawal rate at the sixth failure (out of sixty sequences) ranges between 1.41 to 1.61 times the REAL withdrawal rate.
Now let’s adjust Vanguard’s numbers. The annual amount is $5926.20 or 5.9262% of $100K. If we divide this percentage by 1.41 to 1.61, the adjusted rate is 4.2% to 3.7%. If we divide by 1.13 to 1.25, the adjusted rate is 5.2% to 4.7%.
I presented the 2% TIPS-only baseline from Always insist on a baseline earlier. These rates include adjustments for inflation:
“3) With an interest rate of 2.0%, you can withdraw 4.467% for 30 years or 4.00% for 35 years or 3.66% for 40 years with total safety.”
Vanguard’s numbers are competitive when you make six-failure comparisons. In terms of the first-failure comparisons, Vanguard’s rates are very attractive. If I understand the numbers correctly, Vanguard’s rates do NOT include adjustments for inflation.
In terms of Gummy’s $700 per month or $8400 per year, which is 8.4% of $100K: the relevant comparisons are 6.0% to 5.2% using the six-failure comparisons and 7.4% to 6.7%. These numbers are more than competitive.
Answering the question
Mike asked: “Do Gummy's numbers mean that I would need about $700/month to make the annuity worth while?”
Clearly, the answer is NO. Vanguard’s numbers are attractive.
Now let me get back to the original question, which is the title of the article: “Should You Buy an Annuity?”
Most likely, it makes sense for Mike to put SOME of his money into the annuity. The reason that Gummy’s number is so large is that bond interest rates were much higher then. In terms of today’s markets, Vanguard’s numbers are very attractive for amounts that require a guarantee.
We have developed attractive alternatives, especially for younger retirees, based upon what is likely, but not guaranteed. It is highly likely that there will be a great buying opportunity in stocks within the next 10 to 15 years. If you are able to preserve your capital until then, you should receive a handsome reward. But 10 to 15 years is a long time for a traditional retiree to wait.
We know what to look for: low valuations as measured by P/E10 and/or much higher dividends that are sustainable.
We are likely to encounter at least one of these conditions within a few years. We have no guarantee. We could end up waiting for a long time.
New Letter Inspires a New Area of Research
A reader named Mike found something that he thought that I might find interesting. I found it very interesting. It was so interesting that I have built another calculator and started a brand new area of current research.
Thanks, Mike. Great job!
Current Research D: Expanded Switching Algorithms
Earlier Letters
How Long Do You Have to Wait for 7%?
I finally answer Rob Bennett’s October 6, 2005 question. Well, I come close enough.
How Long Do You Have to Wait for 7%?
Rob Bennett's October 6, 2005 Request
This is what Rob asked for in the first place. I provided information, but I did not answer his question. He wanted A New Take on the 7% Rule.
October 6, 2005 Letters to the Editor
Glossary for Beginners
I am amenable to writing a Glossary for Beginners.
Glossary for Beginners
From a Novice in Investing
Read this letter and take advantage of its links to earlier letters.
From a Novice in Investing
Have fun.
John Walter Russell
Updated: October 15, 2005