Larry -- TIPS studies

[SmallHi]

Larry,

I often notice you will quote papers on a particular asset class you have read and are fond of.

Do you mind sharing what a few of your favorite studies on TIPS are?

Thanks, SH

[larryswedroe]

1.Robert J. Shiller, “The Invention of Inflation-Indexed Bonds in Early America,” Yale ICF Working Paper No. 04-09, Cowles Foundation Discussion Paper No. 142, October 2003.
2. The Handbook of Inflation Indexed Bonds, edited by John Brynjolfsson and Frank Fabozzi, Wiley (1999).
3.S.P. Kothari and Jay Shanken, “Asset Allocation with Inflation-Protected Bonds,” Financial Analysts Journal (January/February 2004).
4.Abdullah Mamun and Nuttawat Visaltanachoti, “Diversification Benefits of Treasury Inflation Protected Securities: An Empirical Puzzle” (February 15, 2006).

[alec]

Asset Allocation with Inflation-Protected Bonds

Diversification Benefits of Treasury Inflation Protected Securities: An Empirical Puzzle

[SmallHi]

a quick question...

In the 70s and early 80s, as inflation continued to rise unexpectedly, would the real yield of new issue TIPS contiuned to fall upon issuance? Say, if real yield was 4.5% in early 70s, falling to 3% in late 70s and 1.5% or so in early 80s? The exact #s aren't important...do I have the trend correct?

So, in a sense, the temporary return to an existing TIPS at that time would have been (real yield @ inception + inflation + price increase to offset newer issue TIPS offered with lower real IRs)?

And, vice versa for a period where inflation declines unexpectedly?

Thanks,

SH

[BlueEars]

SmallHi, I don't really understand your question, but I'd like to offer some data. Don't know if this will help. This data is for 10 yr treasuries minus the previous 12 mo inflation:

1971-1974 0 to 3.3%
1974-1976 0 to -4.0%
1976-1979 0 to 1.8%
1979-1981 0 to -3.5%
1981-1986 0 to 9.8%
1986-1990 6.0 to 3.0%

I have this in Excel chart form but haven't been successful at posting charts. Also have data for 10 yr treasuries minus forward 10 yr inflation rate (what those things actually returned, sort of).

[larryswedroe]

small hi

IMO the real yield on say ten year TIPS should have been approximately the same as the real yield on nominal bonds at the time--which could be approximated by taking the yield on the first of the year say and subtracting prior year inflation--at least reasonable estimate as indicated by the prior poster.

But I seriously doubt you would ever see real yields anywhere near 5 or 6%. That kind or real yield with economic growth of say 3-4 would kill economy.

[bobcat2]

Here's a link to the Shiller paper.
http://diehards.org/forum/viewtopic.php ... highlight=

Bob K

[LH]

Swedroe's TIPs favorties

1.Robert J. Shiller, “The Invention of Inflation-Indexed Bonds in Early America,” Yale ICF Working Paper No. 04-09, Cowles Foundation Discussion Paper No. 142, October 2003.
2. The Handbook of Inflation Indexed Bonds, edited by John Brynjolfsson and Frank Fabozzi, Wiley (1999).
3.S.P. Kothari and Jay Shanken, “Asset Allocation with Inflation-Protected Bonds,” Financial Analysts Journal (January/February 2004).
4.Abdullah Mamun and Nuttawat Visaltanachoti, “Diversification Benefits of Treasury Inflation Protected Securities: An Empirical Puzzle” (February 15, 2006).

Shamelessly copied links to above(thanks bobcat and alec):

1. The Invention of Inflation-Indexed Bonds in Early America

2. library or buy: Handbook of inflation indexed Bonds

3. Asset Allocation with Inflation-Protected Bonds

4. Diversification Benefits of Treasury Inflation Protected Securities: An Empirical Puzzle

[SmallHi]

OK,

Let me try a hypothetical example...

NOW
Yield on 10YR T-Note = 6%
Yield on 10YR TIPS = 2.75%

spread or "difference" = (expected 10YR inflation = 3.0% + 0.25% for NOMINAL inflation risk premium)

Now, 5 years goes by, and inflation has been persistently higher. Over the last 5 years, it has averaged 4.5%.

5YRs LATER
New issue 10YR T-Notes now yield = 7.5%

I guess, from comments above, we would assume new 10YR TIPS sill have a real yield of 2.75%?

Therefore, the nominal price change (not including inflation credit) of the 10YR TIPS issued 5 years ago is still PAR?

OR,

would it be logical to assume there should also be an inflation risk premium adjustment (say 0.5%, up from 0.25%), therefore increasing the spread between new issue 10YR T-Notes and TIPS to 5.0% (4.5% inflation + 0.5% risk premium)?

for example:

10YR T-Note = 7.5%
10YR TIPS = 2.5% (5% spread = 4.5% inflation exp + 0.5% risk premium)

If this is the case (or is reasonably possible), then the current price on a 10YR TIPS with 5YRs left till maturity (with a 2.75% real yield) should also RISE...

Is my thinking right? Or is this too simplistic?

SH

[BlueEars]

Doesn't the current rate on 10 yr TIPS have to do with expected inflation over the next 10 yrs plus current supply/demand and only marginally on the previous inflation? So in the 5 yrs after scenario why should TIPS yield 2.75%?. Seems too complex to hazard a guess on these dynamics.

Though much of the 1970's bond investors were repeatedly burned by inflation but I guess they had few choices back then. Why did they not go on strike? Perhaps they had no place to hide. Real 3mo treasuries were negative in the 1970's. The government did not provide inflation protected bonds. Also taxes were not indexed, a back door way to increase revenue.

Then after a decade and major political changes bond investors were able to get high real rates. Perhaps someone who was a bond investor back then can explain why in the early 1980's bond investors were able to get decent real returns and not in the 1970's. Were they just stupid back then?

[Valuethinker]

Doesn't the current rate on 10 yr TIPS have to do with expected inflation over the next 10 yrs plus current supply/demand and only marginally on the previous inflation? So in the 5 yrs after scenario why should TIPS yield 2.75%?. Seems too complex to hazard a guess on these dynamics.

Though much of the 1970's bond investors were repeatedly burned by inflation but I guess they had few choices back then. Why did they not go on strike? Perhaps they had no place to hide. Real 3mo treasuries were negative in the 1970's. The government did not provide inflation protected bonds. Also taxes were not indexed, a back door way to increase revenue.

People don't learn immediately from events. No one at the start of the 70s knew how bad inflation would get.

In addition, investors who needed bond characteristics (certain income and return) had no choice. Equities of course were risky, and total dogs.

There was huge speculation in commodities and property, not to mention gold and silver.

Then after a decade and major political changes bond investors were able to get high real rates. Perhaps someone who was a bond investor back then can explain why in the early 1980's bond investors were able to get decent real returns and not in the 1970's. Were they just stupid back then?

That was a function of the Fed's policy to kill inflation. Of course in retrospect it was the deal of the century *but*

- oil prices in today's money were $100 and some forecasts said they would go to $200
- inflation was c. 12%, and there were forecasts of 20% -- Britain certainly got there, as well as other countries- 13% long bonds didn't look so attractive -- no one at the time knew inflation would come plummeting down (and indeed it was a long slow painful ride, the worst recession we had experienced since the 1930s)
- most bonds that were issued, even US Treasury Bonds, were callable

Real interest rates are only truly known in retrospect, since the real rate is equal to the interest rate minus the expected inflation rate over the same period.

So those 13% bonds didn't look like they were paying a high real interest rates. Your T-Bills were paying 21% and most of us loaded up on T-Bills and other short term instruments, which the 70s had taught us were the only safe investment.

[larryswedroe]

small hi

I don't think you have this correct. The real return on TIPS IMO will be much more reflective of what is happening in the real economy--which is what drives real interest rates.

It is the nominal bonds that have the big impact from changes in UNEXPECTED inflation.

Remember the yield on a nominal bond is made up of three things
a) real return (reflective of economic growth)
b) expected inflation
c) risk premium for unexpected inflation

In rising inflation, where it was unexpected, the risk premium will likely rise--reflecting greater risks and vice versa.

[BlueEars]

People don't learn immediately from events. No one at the start of the 70s knew how bad inflation would get.

In addition, investors who needed bond characteristics (certain income and return) had no choice. Equities of course were risky, and total dogs.

There was huge speculation in commodities and property, not to mention gold and silver.

That was a function of the Fed's policy to kill inflation. Of course in retrospect it was the deal of the century *but*
....
So those 13% bonds didn't look like they were paying a high real interest rates. Your T-Bills were paying 21% and most of us loaded up on T-Bills and other short term instruments, which the 70s had taught us were the only safe investment.

I was in Silicon Valley at the time and bought my first house in 1975. There certainly was the beginning of real estate speculation at that time as most of us knew, despite the inflationary pay raises, we were not going to easily stay ahead of this thing. I recall reading an article or two in business magazines suggesting inflation would be a useful tool in getting out of RE debt. During the Carter years 1976-1980 it was clear fiscal policy was not addressing inflation and that the government was growing without having to raise taxes. Probably Carter lost a few votes because of this (mine included). After the elections rates went into the stratosphere going to above 15% in Oct. 1981.

I was not a bond buyer back then but did open a MM account at Vanguard. For much of the 1970's Tbills did not even keep up with inflation so I guess, as you said, there was no place to hide -- except in RE, gold, collectables. I pitty the poor retiree back then. Still it is hard to envision today's market's being so sluggish in responding to inflationary pressures.

[BlueEars]

Larry, if you are still following this, what was your response to inflation in the 1970's? How did you invest back then?

[Doc]

SmallHi wrote:

would it be logical to assume there should also be an inflation risk premium adjustment (say 0.5%, up from 0.25%), therefore increasing the spread between new issue 10YR T-Notes and TIPS to 5.0% (4.5% inflation + 0.5% risk premium)?

I think the risk premium would increase. People are more concerned about inflation when it is high and therefore are more likely to buy TIPS rather than nominal bonds. But what happens on the supply side. Will the treasury supply a lot more TIPS and fewer nominal bonds? We don't know and we can't measure the insurance premium anyway since it is combined with the "expected inflation".

Larry thinks that the premium will increase also (see prior) and his opinion is worth a lot more than mine but it is still only an opinion.

For planning purposes I would assume that there will be no change in premium and any change in spread is due to only the change in the perceived inflation itself.

[larryswedroe]

Les
I was 100% equity at the time. Young, aggressive risk taker, long horizon, etc.

[SmallHi]

thanks for the comments.

I am beginning to think a TIPS mutual fund may not be the best vehicle for targeting this asset class, even if you can do so at 0.11% as you can with Vanguard.

Here is my thinking...

For the 10 years ended 2/2007, the Lehman 1-30YR TIPS Index compounded at 6.8% a year. The annualized level of CPI over this period was only 2.5%, and the original issue TIPS back in 1997 had a yield of only 3.375%.

An investor who bought and held the 10YR TIPS would have earned about 5.9% (less than the 10YR Treasury Note holder of the same duration who earned 6.25% -- as inflation came in less than expected).

The difference between the TIPS security investor return (+5.9%) and the TIPS Fund Investor (6.8%) return was due to the lower real rate of interest offered on future TIPS securities (causing existing TIPS to rise).

Today, for example, the nominal/TIPS spread on 10YR issues is only 2.3% (vs. 2.9% 10YRs ago).

It appeared that lower realized inflation caused the REAL YIELD on future TIPS (as well as nominal bonds) to fall. This price decline increased the price of existing TIPS issues.

So, wouldn't we assume the same pattern would happen in reverse if inflation headed higher over the next 10 years?

Lets say realized inflation climbed back to the 2.9% level, and in 10 years, TIPS were back to a real yield of 3.375, and 10YR T-Notes were back to 6.25%. (as they were in 1997)

Wouldn't we then assume this 0.9% TIPS "gain" (above buying the TIPS security) that we saw on LEH TIPS Index over the last decade would turn into a future "loss", as newly issued TIPS at higher real yields would push DOWN the price of existing TIPS (temporarily), negatively impacting a TIPS mutual fund?

If this is the case, TIPS mutual funds don't appear to be as inflation protecting as the individual security...

SH

[larryswedroe]

small hi

few thoughts

1) the majority of the literature especially the newer stuff finds that TIPS should dominate portfolios. some say 100%.

2) You are taking period when inflation was less than expected. Obviously nominal bonds will outperform. that is a given. But that is not what you buy TIPS for.

3) the alternative to TIPS is not similar maturity nominal bonds but shorter term bonds due to lower correlation to equities and similar inflation hedge. So would be better IMO to look at say one year returns during that period

4)You are also making mistake of thinking of asset class in isolation. TIPS should have greater diversification benefits due to negative correlation to stocks vs. low for fixed income nominal bonds.

Only reason I don't like funds is you cannot shift maturity. at low real yields I prefer shorter term bonds like FHLB for tax advantaged accounts. When real yields high then I prefer TIPS. And longer TIPS.

[Doc]

SmallHI wrote:

It appeared that lower realized inflation caused the REAL YIELD on future TIPS (as well as nominal bonds) to fall.

This "appears" to be impossible. The notes promised a fixed coupon that is the real yield. It then continued to pay out the same real yield and also paid out an amount that was equal to the amount of inflation whether or not that inflation rose or fell over the period. so I don't undersatnd your statement.

Going back to the early history of TIPS and trying to draw numeric conclusions ftom the dta is futile because the TIPS returns were so distorted when they were new.

The treasury does not even include the TIPS yield before 2003. http://www.treas.gov/offices/domestic-f ... ical.shtml

[SmallHi]

The notes promised a fixed coupon that is the real yield. It then continued to pay out the same real yield and also paid out an amount that was equal to the amount of inflation whether or not that inflation rose or fell over the period. so I don't undersatnd your statement.

I didn't explain that well.

Lets take 1/1/2000 for example. 10YR TIPS were offered at 4.25%, 10YR T-Notes offered at 6.5%. Moving forward, the nominal yields on new issue T-Notes and the real yield on new issue TIPS fell. (4.75% and 2.45% earlier this year)

That would have meant the price of existing TIPS and T-Notes that were issued in 2000 increased temporarily.

I was simply making the point that, if nominal yields on T-Notes and real yields on new issue TIPS had insteadincreased, that would have exerted downward price pressure on existing Notes and TIPS...and, in a mutual fund strucutre, this downturn in price would partially offset the inflation credit and real yield payment on the existing TIPS in the portfolio.

Going back to the early history of TIPS and trying to draw numeric conclusions ftom the dta is futile because the TIPS returns were so distorted when they were new.

http://www.treasurydirect.gov/instit/an ... re_iis.htm is the link for all historical TIPS auctions.

SH

[SmallHi]

Larry,

I should add, your points are well taken.

At this point, I think I am trying to figure out how the pricing of TIPS works given different inflation/interest rate enviroments. Real Yield + Inflation is pretty straight forward, I am trying to understand how interm price movements of existing TIPS are affected by new issues at higher or lower real yields, and, overall, how this impacts the fund/security.

I apologize to all for such long, rambling commentary. Its the best way for me to learn...thanks for the helpful comments. I am certainly no Frank Fabbozzi!

SH

[Doc]

SmallHI wrote:

Lets take 1/1/2000 for example. 10YR TIPS were offered at 4.25%,

Now I am not explaining things well. The early TIPS data is not representative because the TIPS were 1) new, 2) unproven, 3) limited in amount, 4)misunderstood, 5) etc. etc.

The January 2000 auction was in the "early" period. The Treasury's "Daily Real Yield Chart" which I linked above started in 2003. I believe this is where the Treasury thinks that the reliable data begins. Earlier data is unreliable because it is atypical. You can argue where the reliable data starts but it is certainly later than Jan 2000. A real yield of 4.25% is just way out of line.

Unfortunately we have less than five years of reliable data and the analysis that you are trying to do has limited value at least until there is more data.

I was simply making the point that, if nominal yields on T-Notes and real yields on new issue TIPS had insteadincreased, that would have exerted downward price pressure on existing Notes and TIPS

Of course that's the way bonds work. You don't have to go through this "data mining" to prove it. Trying to use this early TIPS data to "prove" this concept is just confusing. (At least it confused me but that's not hard to do. )