Vanguard Diehards

Number 28 Correction

bulldog bond — Sat, 02 Jun 2007 00:15:00 GMT

In Post #28, I was examining what would be the effect of choosing between a hypotherical 10Yr TIPS bond and a 10 Yr Treasury in June of 1966. I was being too clever with my spreadsheet, trying to accrue OID Tax liability semi-anually and pay it annually. Well I screwed that up and managed to tax a major portion of the MM account very 6 months! Sorry about that.

I made the needed corrections and recalculated the returns for the two options. Not surprisingly, they were better than I had indicated in my earlier post. The 10Yr Treasury had been more significantly affected by the error, owing to the larger MM account accumulation in Treasury option.

The corrected Real After-tax Total Return comparison: the 10 Yr TIPS real return was 0.47% and the 10 Yr Treasury, in real dollars, lost -1.81% anually.

My conclusion was unchanged. The TIPS option (if it had been offered at that time) would have been the superior choice. This example of an unexpected, long-term, inflationary era in our history shows TIPS could preserve your real wealth & Treasury Bonds would consume it.

Originally posted in thread: 9792

Doc -- Explanations

Taylor Larimore — Fri, 01 Jun 2007 18:00:00 GMT

"Taylor -- When we talk about risk and return being correlated we mean higher return implies higher risk."

Thank you.

"Your two statements while maybe "correct" are in conflict with each other and that is what needs explanation."

Chris explained it for you better than I can:

"Surely you can appreciate the difference between a concept that will hold true over your investing life versus a specific report of 5 year returns."

Best wishes.
Taylor

Originally posted in thread: 9792

Chris,

mjromney — Fri, 01 Jun 2007 16:47:00 GMT

"Surely you can appreciate the difference between a concept that will hold true over your investing life versus a specific report of 5 year returns."

I am not the one who is advocating a mix of TBM and TIPS based on short term data. So I think you are directing your comment at the wrong person.

Regards,

Doc

Originally posted in thread: 9792

Doc - risk "showing up"

CMErick — Fri, 01 Jun 2007 14:59:00 GMT

The higher risk with TIPS so far has shown itself as higher volatility than other Treasuries.

Both despite TIPS higher volatility and because of it, at the moment TIPS are still ahead of conventional bonds in the time period mentioned.

The concept of correlated risk and return is just a concept and will not be true for every investment in every time period.

Surely you can appreciate the difference between a concept that will hold true over your investing life versus a specific report of 5 year returns.

Also, to paraphrase Swedroe, when the risk shows up, you don't get a high return.

Chris

Originally posted in thread: 9792

Taylor #29

mjromney — Fri, 01 Jun 2007 14:32:00 GMT

When we talk about risk and return being correlated we mean higher return implies higher risk. Yet you say the the TIPS fund had the higest return and the lowest risk.

Your two statements while maybe "correct" are in conflict with each other and that is what needs explanation.

Regards,

Doc

Originally posted in thread: 9792

Doc -- Explanation

Taylor Larimore — Thu, 31 May 2007 22:34:00 GMT

""Vanguard's Inflation-Protected Securities (TIPS) fund provided the highest return and lowest risk (of the 4 funds listed)."

"Risk and return are highly correlated."

How do you reconcile these two statements?"

They don't need to reconcile. I believe both statements are correct.

Best wishes.
Taylor

Originally posted in thread: 9792

TIPS in a long-term inflationary spike

bulldog bond — Thu, 31 May 2007 22:21:00 GMT

There is lot of discussion about how TIPS might fare in an inflationary period, and I have done a lot of spreadsheet work examining this question. I agree with Taylor that a mix TIPS and Treasury Bonds make sense. Because there are so many possible assumptions as to how rates might play out in the future, there is no easy answer as to which is better.

There is a simple analysis based on real data that illustrates the power of TIPS as insurance in an inflationary era.

Assume it is June 1966 and you have a choice between buying a 10 Yr TIPS and a 10Yr Treasury. The conditions then were somewhat like today: Yr/Yr Inflation was 2.53%, the 10 Yr Treasury Yield was 4.81%. Inflation and Treasury rates had been rising slightly for some time. There was no clue as to the inflation spike to come. In the next 10 years inflation went above 12%, in the next 20 it went thru a 13% peak before the inflationary cycle was broken.

Let's assume a TIPS instrument was availible in June of 1966 and its Coupon Rate equalled that of the Nominal 10 Yr bond less the Yr/Yr inflation on that date (probably a reaonable assumption).

If you bought the 10Yr TIPS with a 2.278% coupon, accumulated interest in a MM account paying the prevailing historical rate, payed 25% tax on interest and OID gains, you would have an after-tax total return of 6.15%.

If you opted for the 10 Yr Treasury, accumulated interest payments in a MM Acc't and payed taxes as per the TIPS assumptions, you wound up with an after-tax total return of 0.95% (pretty shabby, but it gets worse).

The annualized inflation ran at 5.60% in that 10 year span. The TIPS option (had it been offered at that time) netted you a real annualized gain of 0.56%. Not too hot, but your principal was preserved. The 10 Yr Treasury left you with a real loss of -4.64% per annum.

Who knows what the future holds. This example from a long-term inflationary era in our history shows TIPS could preserve your real wealth & Treasury Bonds would consume it. Other analyses show, at worst, TIPS offer low-cost insurance against unexpected inflation.

Bulldog

Originally posted in thread: 9792

Huhhh?

mjromney — Thu, 31 May 2007 21:41:00 GMT

Taylor wrote:

"Vanguard's Inflation-Protected Securities (TIPS) fund provided the highest return and lowest risk."

and

"Risk and return are highly correlated."

How do you reconcile these two statements?

Regards,

Doc

Originally posted in thread: 9792

Schellhase -- New Recommendation?

Taylor Larimore — Thu, 31 May 2007 21:30:00 GMT

"Taylor, are you still maintaining this 50/50 TIPS/TBM allocation with today's lower yields?"

Before answering, let's look at 5-year annualized returns ending 12-31-06 to see what happened with these Vanguard funds since I posted this Conversation. Figures in (parenthesis) are for 2002--the worst bear market year:

3.30%....Short-Term Bond Index Fund (+6.1%)
4.61%....Total Bond Market Index Fund (+8.3%)
7.04%....Inflation-Protected Securities (+16.6%)
6.07%....S&P 500 Index Fund (-22.1%)

OBSERVATION:
1. Vanguard's Inflation-Protected Securities (TIPS) fund provided the highest return and lowest risk.

2. TBM outperformed S.T. Bond Index (return and risk).

3. Owning bonds (any kind) during the 1992 bear market provided significant portfolio protection from decline.
--------------------------------------------------------------------
A 50% Total Bond Market and 50% Inflation-Protected Securities was an excellent bond combination over the short period studied. What will happen in the future is anyone's guess.

I do not think it is possible to forecast future bond fund performance. I also think market timing in the highly efficient bond world is futile. Risk and return are highly correlated.

So, to answer your question: Yes, I am "still maintaining this 50/50 TIPS/TBM allocation with today's lower yields."

Having said that, I don't feel strongly about my bond combo or any other good quality short- or intermediate-term bond fund(s). . All should do the job of providing income and safety for our portfolios.

Best wishes.
Taylor

Originally posted in thread: 9792

Middle ground

John3 — Thu, 31 May 2007 20:26:00 GMT

I have been following the ongoing discussions on bonds for a while now and have come up with a middle ground for my bond holdings. I use 1/3 Total Bond, 1/3 TIPS and 1/3 Short Term Index. Another way to Dublin? John

Originally posted in thread: 9792

A look...

TrevH — Thu, 31 May 2007 15:36:00 GMT

.
Below is a look at Short Term Treasury vs Total Bond mixed with the 500 Index 1970-2006 including 10K Growth and Volatility details.

I think what Larry often suggest is that tilting to "Value" is a better bet than going longer on bond duration.

Trev H
======

60% 500 Index
40% Total Bond
380,974.34
10.91

50% 500 Index
10% Small Value
40% Short Term Treasury
368,385.44
10.23

45% 500 Index
15% Small Value
40% Short Term Treasury
393,249.05
10.17

40% 500 Index
20% Small Value
40% Short Term Treasury
418,979.68
10.17

35% 500 Index
25% Small Value
40% Short Term Treasury
445,538.26
10.23

30% 500 Index
30% Small Value
40% Short Term Treasury
472,878.84
10.35

Short Term Treasury = lower correlation to 500 Index than Total Bond.

Small Value is also quite disconnected from the 500 Index.

The combination of 500 Index & Small Value combined with ST Treasury did not excede the volatility of 500 Index & Total Bond until you reached around 43% exposure to Small Value.

17.5% 500 Index
42.5% Small Value
40% Short Term Treasury
544,286.23
10.89

Listing 500 Index and Total Bond again below...

60% 500 Index
40% Total Bond
380,974.34
10.91

In this backtest Short Term Treasury return is calculated using a 50/50 mix of 5 Year T-Notes & T-Bills. Total Bond is Lehman Agg & VBMFX.

Anyway what Larry usually stresses is that you should not take risk on the fixed income side... better off doing that on the equity side with value tilts.

Trev H

Originally posted in thread: 9792

NOTE

pkcrafter — Thu, 31 May 2007 13:59:00 GMT

Readers should note if they missed it that the original post is #9792, made in 2001. It was updated in O2, 03, 04 and now 07. Good history of bond AA and tips.

Paul

Originally posted in thread: 9792

Strategy Still Working?

schellhase — Thu, 31 May 2007 05:35:00 GMT

Taylor, are you still maintaining this 50/50 TIPS/TBM allocation with today's lower yields?

Thanks

Originally posted in thread: 9792

2004 Update about TIPS

Taylor Larimore — Sun, 22 Feb 2004 01:33:00 GMT

Hi Diehards:
Annette Thou, the highly regarded author of "The Bond Book," has written an article about TIPS in the February 2004 issue of the AAII Journal.

I posted excerpts from her article in Conversation 33336. I hope you find this link helpful.

Best wishes.
Taylor

Originally posted in thread: 9792

Alec - Yes, the same article

jlnxxx — Mon, 05 May 2003 18:38:00 GMT

Alec wrote:

For those that don't have access to the American Economic Review, I found this article on Campbell's Harvard website. John, does this look like the same article?

Yes, that's the one. I actually got a chance to listen to Campbell present this paper at a seminar here at NU a few years ago.

He has an interesting and I think valid argument - basically, MPT makes long bonds look bad because MPT (mean-variance portfolio analysis) focuses only on the short term, using short-term return and variance data. For long-term investors we need to use more appropriate (and significantly more complicated) analytic techniques, and when we do, long bonds are quite attractive - much more so than short bonds under many quite reasonable scenarios.

As a simple example he uses to make his basic point, most people think of "cash" (e.g., short US Treasuries, savings accounts, or money market funds) as the "safest" investments. But for a long-horizon investor there's huge reinvestment risk, and in fact for such a long-horizon investor long-term inflation-protected bonds are actually the "riskless" asset. In low inflation environments he also shows using his statistical and mathematical models that long-term nominal bonds are better for long-horizon investors than short-term bonds are.

It's certainly food for thought. Unfortunately, the math is really hard to understand even if the conclusions based on the math are not so difficult.

The book I mentioned goes into this in much more detail and covers many other interesting topics where "strategic" or "long-horizon" investing is often different and more complicated than the "tactical" or "short-horizon" investing modeled by simple MPT.

John Norstad

Originally posted in thread: 9792