Tips breakevens and illiquidity premia

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Tips breakevens and illiquidity premia

Post by grok87 » Thu Dec 06, 2018 9:10 am

The bond market is sending some interesting signals
Right now:

Nominal treasuries;
5 yr: 2.75
10 yr: 2.88
30 year: 3.14

Tips
5 yr: 1.04
10 yr: 0.96
30 yr: 1.15

Breakeven inflation
5 year: 1.71
10 yr: 1.92
30 yr: 1.99
5x5: 2.13
20x10: 2.025

I think long run fair value breakeven inflation would be about 2.25%. Because the fed targets PCE inflation at 2% and CPI usually runs 25 bps higher. Since 20x10 is at 2.02 that implies. 23 bp illiquidity Premium for long dated tips.

Interesting.
Keep calm and Boglehead on. KCBO.

User avatar
Svensk Anga
Posts: 444
Joined: Sun Dec 23, 2012 5:16 pm

Re: Tips breakevens and illiquidity premia

Post by Svensk Anga » Thu Dec 06, 2018 10:23 am

I think what is happening is that there is increasing doubt that the Fed can keep inflation up to its target of 2% (PCE). The five year break-even inflation rate has dropped from 2.00% on 10/17 to 1.74% on 12/4. It would appear that concern is rising about the onset and severity of the next recession.

grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Re: Tips breakevens and illiquidity premia

Post by grok87 » Thu Dec 06, 2018 10:39 am

Svensk Anga wrote:
Thu Dec 06, 2018 10:23 am
I think what is happening is that there is increasing doubt that the Fed can keep inflation up to its target of 2% (PCE). The five year break-even inflation rate has dropped from 2.00% on 10/17 to 1.74% on 12/4. It would appear that concern is rising about the onset and severity of the next recession.
Agree. That’s why I tried to look past that by looking at 5x5 bei and 20x10 bei
Keep calm and Boglehead on. KCBO.

User avatar
Ice-9
Posts: 1338
Joined: Wed Oct 15, 2008 12:40 pm
Location: Rockville, MD

Re: Tips breakevens and illiquidity premia

Post by Ice-9 » Thu Dec 06, 2018 10:44 am

Sorry for the novice question: What is meant by "5x5" and "20x10" breakeven inflation?

grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Re: Tips breakevens and illiquidity premia

Post by grok87 » Thu Dec 06, 2018 10:50 am

Ice-9 wrote:
Thu Dec 06, 2018 10:44 am
Sorry for the novice question: What is meant by "5x5" and "20x10" breakeven inflation?
5x5 is inflation for the second 5 years of the next 10 years. So 2023-2028

20x10 is inflation for the last 20 years of the next 30 years or 2028-2048.
Keep calm and Boglehead on. KCBO.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Thu Dec 06, 2018 11:34 am

grok87 wrote:
Thu Dec 06, 2018 9:10 am
I think long run fair value breakeven inflation would be about 2.25%. Because the fed targets PCE inflation at 2% and CPI usually runs 25 bps higher. Since 20x10 is at 2.02 that implies. 23 bp illiquidity Premium for long dated tips
Larry Swedroe wrote:The 0.25 percent lower expected return can be thought of as the risk premium investors are willing to pay to guarantee a real return.
The only Guide to a Winning Bond Strategy You Will Ever Need, First Edition March 2006, p223.
grok87 wrote:
Thu Dec 06, 2018 9:10 am
The bond market is sending some interesting signals
Right now:
Bogleheads Forum December 2018

Right now? :D
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

stlutz
Posts: 4852
Joined: Fri Jan 02, 2009 1:08 am

Re: Tips breakevens and illiquidity premia

Post by stlutz » Thu Dec 06, 2018 1:08 pm

The 0.25 percent lower expected return can be thought of as the risk premium investors are willing to pay to guarantee a real return.
This would go the other way. If investors are paying more to guarantee a real return, the breakeven rate would be 2.5%, not 2% (assuming grok's formulation of 2.25% being "neutral").

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Thu Dec 06, 2018 1:54 pm

stlutz wrote:
Thu Dec 06, 2018 1:08 pm
The 0.25 percent lower expected return can be thought of as the risk premium investors are willing to pay to guarantee a real return.
This would go the other way. If investors are paying more to guarantee a real return, the breakeven rate would be 2.5%, not 2% (assuming grok's formulation of 2.25% being "neutral").
The "problem" is with the term breakeven inflation. The "breakeven inflation" includes inflation, a risk factor and a liquidity factor. So it is not an estimate of actual inflation predictions unless the other two factors are very small.

Aside: With the size of the current TIPS market the liquidity factor is probably small. But in the fall of 2008 that was certainly not the case. If we get another Lehman type market disruption the liquidity factor could again become quite large. Caveat emptor. (I didn't pay attention to Grok's math. I often get the sign wrong on the adjustment wording. That's why I used the word "factor". Is it a price factor or a return factor? I don't know and don't really care. :D )
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

ThrustVectoring
Posts: 588
Joined: Wed Jul 12, 2017 2:51 pm

Re: Tips breakevens and illiquidity premia

Post by ThrustVectoring » Thu Dec 06, 2018 2:07 pm

Svensk Anga wrote:
Thu Dec 06, 2018 10:23 am
I think what is happening is that there is increasing doubt that the Fed can keep inflation up to its target of 2% (PCE). The five year break-even inflation rate has dropped from 2.00% on 10/17 to 1.74% on 12/4. It would appear that concern is rising about the onset and severity of the next recession.
Yeah, the fed does not have the ability to credibly commit to a long-term inflation rate. If inflation lags under their target, they keep the same target and don't adjust things by saying "we have undershot our inflation target, so we're printing more money now until we reach price levels that indicate 2% annualized inflation since the start of this program".

If you've had any exposure to control theory, this is immediately obvious. If you model the fed as a PID controller on prices, there's no integral term, so long-term deficiencies from the target can and will persist.

There's also some really interesting theoretical work being done on algorithmic monetary policy, especially around NGDP targeting (which generally has this integral portion). But that's much more of a digression.
Current portfolio: 60% VTI / 40% VXUS

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Thu Dec 06, 2018 4:55 pm

The issue that I always run into in discussions about the inflation breakeven points is that they don't fundamentally address the main reason I have a lot of TIPS in my portfolio, which is to protect myself against inflation that runs a lot higher than the Fed's target rate. I've always thought TIPS offer a lot of value to people like myself, whose main fear isn't that TIPS might under-perform nominal bond by 50 basis points (or alternatively outperform nominals by 50 basis points – big deal), but that inflation could really take off in an unexpected way. To me, that's their real value. I've often wondered how much weight is given to that scenario with regard to how the market prices TIPS. I suspect not very much, which is why I think they offer a lot of value as "insurance."

grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Re: Tips breakevens and illiquidity premia

Post by grok87 » Thu Dec 06, 2018 5:03 pm

gmaynardkrebs wrote:
Thu Dec 06, 2018 4:55 pm
The issue that I always run into in discussions about the inflation breakeven points is that they don't fundamentally address the main reason I have a lot of TIPS in my portfolio, which is to protect myself against inflation that runs a lot higher than the Fed's target rate. I've always thought TIPS offer a lot of value to people like myself, whose main fear isn't that TIPS might under-perform nominal bond by 50 basis points (or alternatively outperform nominals by 50 basis points – big deal), but that inflation could really take off in an unexpected way. To me, that's their real value. I've often wondered how much weight is given to that scenario with regard to how the market prices TIPS. I suspect not very much, which is why I think they offer a lot of value as "insurance."
That’s the way I feel too.

I think in general the market underprices regime change and black swans. Except the swan isn’t really black here because it has happened before- ie the high inflation of the 70s and early 80s.

Another source of expected inflation estimates is the philly fed survey
https://www.philadelphiafed.org/researc ... 8/survq418

Right now they have expected CPI over the next 10 years at 2.21% and expected pce inflation at 2.01%.
Last edited by grok87 on Thu Dec 06, 2018 5:08 pm, edited 1 time in total.
Keep calm and Boglehead on. KCBO.

grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Re: Tips breakevens and illiquidity premia

Post by grok87 » Thu Dec 06, 2018 5:08 pm

stlutz wrote:
Thu Dec 06, 2018 1:08 pm
The 0.25 percent lower expected return can be thought of as the risk premium investors are willing to pay to guarantee a real return.
This would go the other way. If investors are paying more to guarantee a real return, the breakeven rate would be 2.5%, not 2% (assuming grok's formulation of 2.25% being "neutral").
Yep
Keep calm and Boglehead on. KCBO.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Thu Dec 06, 2018 5:17 pm

grok87 wrote:
Thu Dec 06, 2018 5:03 pm
I think in general the market underprices regime change and black swans. Except the swan isn’t really black here because it has happened before- ie the high inflation of the 70s and early 80s.
It's been a long time but ...

I belive that in 70's and early 80's the Fed's policy was to control the money supply M1 & M2. That changed with Paul Volcker.

See for example "How Former Fed Chairman Paul Volcker Tamed Inflation — Maybe For Good" https://www.npr.org/2015/12/15/45987100 ... e-for-good
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Re: Tips breakevens and illiquidity premia

Post by grok87 » Thu Dec 06, 2018 5:37 pm

Doc wrote:
Thu Dec 06, 2018 5:17 pm
grok87 wrote:
Thu Dec 06, 2018 5:03 pm
I think in general the market underprices regime change and black swans. Except the swan isn’t really black here because it has happened before- ie the high inflation of the 70s and early 80s.
It's been a long time but ...

I belive that in 70's and early 80's the Fed's policy was to control the money supply M1 & M2. That changed with Paul Volcker.

See for example "How Former Fed Chairman Paul Volcker Tamed Inflation — Maybe For Good" https://www.npr.org/2015/12/15/45987100 ... e-for-good
Thanks.
It’s a good article and a good story.
My own view is a bit different. If you look back over the past 100 years or so there were actually 3 periods when there were bad bouts of high inflation. The late 1910s. The 1940s. And the 70s. I think the common linkage is that these were all periods of wartime deficit spending.
Keep calm and Boglehead on. KCBO.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Thu Dec 06, 2018 7:17 pm

grok87 wrote:
Thu Dec 06, 2018 5:37 pm
If you look back over the past 100 years or so there were actually 3 periods when there were bad bouts of high inflation. The late 1910s. The 1940s. And the 70s. I think the common linkage is that these were all periods of wartime deficit spending.
In the Volker period the Fed changed their objective from controlling the money supply to one that used inflation "control" as one of their major objectives. They were very successful. Their policy changed from the historic goal. Part of the reason was the wide spread of money market accounts which changed the whole metric used in the past. Since that time inflation has been under control. Will it still be so in the future? Time will tell. But using data prior to the Volker period to formulate an inflation prediction is 'ing on the wind.

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.

With today's yield curve it makes more sense to me to get any inflation protection from short term nominals.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Thu Dec 06, 2018 10:50 pm

Doc wrote:
Thu Dec 06, 2018 7:17 pm
grok87 wrote:
Thu Dec 06, 2018 5:37 pm
If you look back over the past 100 years or so there were actually 3 periods when there were bad bouts of high inflation. The late 1910s. The 1940s. And the 70s. I think the common linkage is that these were all periods of wartime deficit spending.
In the Volker period the Fed changed their objective from controlling the money supply to one that used inflation "control" as one of their major objectives. They were very successful. Their policy changed from the historic goal. Part of the reason was the wide spread of money market accounts which changed the whole metric used in the past. Since that time inflation has been under control. Will it still be so in the future? Time will tell. But using data prior to the Volker period to formulate an inflation prediction is 'ing on the wind.

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.

With today's yield curve it makes more sense to me to get any inflation protection from short term nominals.
I’d be curious why you have a low personal risk for unexpected inflation. Perhaps you have Social Security or other inflation adjusted pension? For most people, short term nominalls won’t protect them from a central bank that pegs short term rates below inflation. Eg., most of the last decade. Or to put it another way, when you need them most, ie., when the central bank is working to increase inflation, short nominals won’t be very helpful. Short nominals worked great during the Volcker era, but only because we had a central bank that wanted to lower inflation. In the opposite situation, which has been far more common historically, real short rates will be negative. I don’t consider that inflation protection.

fennewaldaj
Posts: 402
Joined: Sun Oct 22, 2017 11:30 pm

Re: Tips breakevens and illiquidity premia

Post by fennewaldaj » Thu Dec 06, 2018 11:28 pm

Doc wrote:
Thu Dec 06, 2018 7:17 pm

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.
But it doesn't seem like you are paying much for the protection right? It be a different story if say the ten year nominal yielded 2.9% and the ten year TIP real yield was 0%

Angst
Posts: 1874
Joined: Sat Jun 09, 2007 11:31 am

Re: Tips breakevens and illiquidity premia

Post by Angst » Thu Dec 06, 2018 11:31 pm

Doc wrote:
Thu Dec 06, 2018 7:17 pm
grok87 wrote:
Thu Dec 06, 2018 5:37 pm
If you look back over the past 100 years or so there were actually 3 periods when there were bad bouts of high inflation. The late 1910s. The 1940s. And the 70s. I think the common linkage is that these were all periods of wartime deficit spending.
In the Volker period the Fed changed their objective from controlling the money supply to one that used inflation "control" as one of their major objectives. They were very successful. Their policy changed from the historic goal. Part of the reason was the wide spread of money market accounts which changed the whole metric used in the past. Since that time inflation has been under control. Will it still be so in the future? Time will tell. But using data prior to the Volker period to formulate an inflation prediction is 'ing on the wind.

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.

With today's yield curve it makes more sense to me to get any inflation protection from short term nominals.
I'm not sure I accept (or maybe understand?) your logic - at least your conclusion seems somewhat reckless. Isn't that kinda like someone who finally grasps the science behind global warming deciding to give away all their winter jackets?

JackoC
Posts: 360
Joined: Sun Aug 12, 2018 11:14 am

Re: Tips breakevens and illiquidity premia

Post by JackoC » Fri Dec 07, 2018 10:09 am

fennewaldaj wrote:
Thu Dec 06, 2018 11:28 pm
Doc wrote:
Thu Dec 06, 2018 7:17 pm

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.
But it doesn't seem like you are paying much for the protection right? It be a different story if say the ten year nominal yielded 2.9% and the ten year TIP real yield was 0%
Yes, the whole point of the numbers in OP is that right now you are *being paid* to hold TIPS v nominal treasuries, because the liquidity/other yield premium being paid to you is bigger than the theoretical inflation insurance premium you are paying for the inflation protection. You make more now with 10yr TIPS than nominal 10yr treasuries at an inflation rate slightly below 2%, but there's no reason to think the market's actual inflation expectation is any lower than 2%.

Above someone linked the Philly Fed's survey based inflation expectation number. The Cleveland Fed publishes a model based inflation expectation, 2.15% in 10 yrs on the most recent (mid November) run of their model.
https://www.clevelandfed.org/our-resear ... tions.aspx

This has sometimes been the case with TIPS though not always, see Cleveland Fed's graph. Often the TIPS inflation break even is higher than the calculated expectation, as 'should' be the case theoretically with no liquidity/other. Without considering liquidity, TIPS should have a lower expected return than nominal by an 'insurance' premium, as per the standard talking points against TIPS. But usually a liquidity/other premium cancels out part, all or more than all of the 'insurance' premium and TIPS can actually have higher expected return than nominal treasuries, as now.

Some people extrapolate from the generally correct idea that it's a crap shoot at best to guess what the stock market is going to do next to a more general but dubious pseudo-axiom that market pricing should never influence investing decisions. The latter idea is nonsensical, however. If there is a portion of your fixed income allocation that does not depend on being able to suddenly sell those bonds at favorable prices during a crisis, but rather can be held to maturity (and this is some, generally most, of most people's fixed income) then a liquidity premium in yield based on *other* investors' needs to suddenly sell is money in your pocket. And there's no rational argument it couldn't really be there because the market is efficient. The treasury market *is* largely efficient. Other investors are paying a fair liquidity premium for *their* liquidity need. For portions of your fixed income portfolio with a lesser liquidity need than the 'market clearing' investor, you should accept that premium. But it depends on TIPS pricing at a given time.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Fri Dec 07, 2018 4:43 pm

gmaynardkrebs wrote:
Thu Dec 06, 2018 10:50 pm
I’d be curious why you have a low personal risk for unexpected inflation.
1) We have a significant position in investment real estate which should keep up with inflation at least in the long run.

2) Our personal expenses are not highly correlated with inflation.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Fri Dec 07, 2018 4:51 pm

Angst wrote:
Thu Dec 06, 2018 11:31 pm
Doc wrote:
Thu Dec 06, 2018 7:17 pm
grok87 wrote:
Thu Dec 06, 2018 5:37 pm
If you look back over the past 100 years or so there were actually 3 periods when there were bad bouts of high inflation. The late 1910s. The 1940s. And the 70s. I think the common linkage is that these were all periods of wartime deficit spending.
In the Volker period the Fed changed their objective from controlling the money supply to one that used inflation "control" as one of their major objectives. They were very successful. Their policy changed from the historic goal. Part of the reason was the wide spread of money market accounts which changed the whole metric used in the past. Since that time inflation has been under control. Will it still be so in the future? Time will tell. But using data prior to the Volker period to formulate an inflation prediction is 'ing on the wind.

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.

With today's yield curve it makes more sense to me to get any inflation protection from short term nominals.
I'm not sure I accept (or maybe understand?) your logic - at least your conclusion seems somewhat reckless. Isn't that kinda like someone who finally grasps the science behind global warming deciding to give away all their winter jackets?
Given today's yield curve I see no reason to hold long term of maybe even intermediate term bonds. While I don't completely agree with Kevin's tBill idea certainly Treasuries of 2 or 3 years seem reasonable. I am not willing to pay any premium at all for inflation coverage for that short a period of time. If I was building a long term fixed income portfolio like Grok talks about it would be a different storey. But even then I wouldn't be buying thirties in this market. I would be buyin tens or maybe even fives and rolling them when they mature.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

stlutz
Posts: 4852
Joined: Fri Jan 02, 2009 1:08 am

Re: Tips breakevens and illiquidity premia

Post by stlutz » Fri Dec 07, 2018 5:23 pm

One other factor to possibly consider is that during the time that TIPS have actually existed, nominal bonds have actually been a more optimal diversifier to stocks. That is, when stocks have gone down, nominals have outperformed TIPS.

If this behavior is intrinsic to the investment as opposed to random noise, then it could be priced which would help explain why TIPS yields seem "too high" .

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Fri Dec 07, 2018 5:41 pm

stlutz wrote:
Fri Dec 07, 2018 5:23 pm
One other factor to possibly consider is that during the time that TIPS have actually existed, nominal bonds have actually been a more optimal diversifier to stocks. That is, when stocks have gone down, nominals have outperformed TIPS.

If this behavior is intrinsic to the investment as opposed to random noise, then it could be priced which would help explain why TIPS yields seem "too high" .
Certainly this difference in TIPS and nominal performance in the '08 was obvious. I got burned by the mantra "TIPS are just like nominals except for the inflation protection". Obviously not true. The question becomes was the poor TIPS performance in '08 only a liquidity issue that may not longer exist? I don't know if we have any way to tell unless we get another Lehman. But stlutz's observation should definitely be part ones TIPS buying calculation. For me it overcomes any unexpected inflation protection that TIPS might provide regardless of any insurance premium.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Kevin M
Posts: 10306
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Tips breakevens and illiquidity premia

Post by Kevin M » Fri Dec 07, 2018 5:55 pm

Doc wrote:
Thu Dec 06, 2018 1:54 pm
The "problem" is with the term breakeven inflation. The "breakeven inflation" includes inflation, a risk factor and a liquidity factor. So it is not an estimate of actual inflation predictions unless the other two factors are very small.
Or if they are about the same size. The TIPS illiquidity yield premium increases TIPS yield, and the nominal Treasury unexpected inflation yield premium increases nominal yield, so they tend to cancel each other out. Neither premium is directly observable, so it seems to me that anything said here about either is speculation.

My speculation is that the extremely flat nominal Treasury yield curve seems to imply not much of an unexpected inflation yield premium.

I don't have any sense of the current TIPS illiquidity yield premium, but since the TIPS yield curve also is extremely flat, one is taking a lot of real term risk going long. Apparently that's not a concern for those with long-term TIPS ladders, as it's all about the insurance, not about the return. I am a little puzzled about statements that the insurance cost is low, as it seems to me that there is significant potential opportunity cost in locking in historically low, long-term, real rates.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Fri Dec 07, 2018 10:39 pm

Kevin M wrote:
Fri Dec 07, 2018 5:55 pm
Doc wrote:
Thu Dec 06, 2018 1:54 pm
The "problem" is with the term breakeven inflation. The "breakeven inflation" includes inflation, a risk factor and a liquidity factor. So it is not an estimate of actual inflation predictions unless the other two factors are very small.
Or if they are about the same size. The TIPS illiquidity yield premium increases TIPS yield, and the nominal Treasury unexpected inflation yield premium increases nominal yield, so they tend to cancel each other out. Neither premium is directly observable, so it seems to me that anything said here about either is speculation.

My speculation is that the extremely flat nominal Treasury yield curve seems to imply not much of an unexpected inflation yield premium.

I don't have any sense of the current TIPS illiquidity yield premium, but since the TIPS yield curve also is extremely flat, one is taking a lot of real term risk going long. Apparently that's not a concern for those with long-term TIPS ladders, as it's all about the insurance, not about the return. I am a little puzzled about statements that the insurance cost is low, as it seems to me that there is significant potential opportunity cost in locking in historically low, long-term, real rates.

Kevin
The reason I buy long TIPS is due to my rather cynical but probably realistic belief that the first thing the government is going to do if serious inflation breaks out is to discontinue TIPS. A rather extreme form of reinvestment risk. The cost is low. I’ve always thought that long TIPS should cost more, not less than shorter durations, because the government is taking on much greater tail risk as the term increases; somewhat analogous to why 30 year term life costs more than 10 year term life – a greater likelihood that they might actually have to pay up over the guarantee period. Instead the government charges less.

JackoC
Posts: 360
Joined: Sun Aug 12, 2018 11:14 am

Re: Tips breakevens and illiquidity premia

Post by JackoC » Fri Dec 07, 2018 11:47 pm

Kevin M wrote:
Fri Dec 07, 2018 5:55 pm

Or if they are about the same size. The TIPS illiquidity yield premium increases TIPS yield, and the nominal Treasury unexpected inflation yield premium increases nominal yield, so they tend to cancel each other out. Neither premium is directly observable, so it seems to me that anything said here about either is speculation.

My speculation is that the extremely flat nominal Treasury yield curve seems to imply not much of an unexpected inflation yield premium.

I don't have any sense of the current TIPS illiquidity yield premium, but since the TIPS yield curve also is extremely flat, one is taking a lot of real term risk going long. Apparently that's not a concern for those with long-term TIPS ladders, as it's all about the insurance, not about the return. I am a little puzzled about statements that the insurance cost is low, as it seems to me that there is significant potential opportunity cost in locking in historically low, long-term, real rates.

Kevin
I'd again recommend the Cleveland Fed site linked above. There are model outputs and papers attempting to separately quantify the insurance and liquidity premiums of TIPS. Although sure they are not directly observable separately. However what's a step easier to observe is inflation expectations by measures other than TIPS v the 'TIPS breakeven'. Again the Philly Fed publishes an ongoing survey-based measure of market inflation expectations, the Cleveland Fed a model based one. Still not as simple as just looking at TIPS and saying 'yep, the net of the two premiums must be X' and nobody can disagree, but when the TIPS breakeven is below 2% in a pretty good economy, that tends to agree, seat of the pants wise, with the Cleveland Fed model saying inflation expectations are still a 'normal' slightly 2% plus, and the net of the insurance and liquidity premiums is probably a small number in the TIPS investor's favor right now: you're getting paid slightly more for the illiquidity than you are paying for 'insurance'.

Separate from your post, the point made above about 'TIPS didn't do well in 2008' is the same fallacy that often comes up about CD's. Like CD's, TIPS are for the portion of your fixed income *not* likely to be subject to use in rebalancing. And as has been gone over a bunch of times, for that portion of fixed income market to market performance is of no actual relevance, it's just psychological.

Back to your post, choosing a term and nominal v TIPS are somewhat different topics. The comparison of inflation expectations and TIPS breakeven is to evaluate TIPS v nominal in a given maturity rather than choose an appropriate maturity. However the NY Fed publishes results of a model which estimates the term premium*, and it basically agrees with what you say. The term premium now is unusually negative per the model. Not specifically because the curve is flat, it's more negative than past times when the curve was flat. It's ~98.5%-tile lowest (more negative) term premium in model outputs since June 1961 in 10yrs, one can see downloading their spreadsheet of results. The model says the equilibrium expected 10 yr nominal rate is ~3.47% v ~2.87% actual, 60 bp negative term premium. Long duration doesn't suit my other investments to begin with, but now the term premium is actually negative, at least per this model. Still a bit different topic than TIPS v nominal, though I'd grant they are related in that inflation protection for short periods is not as big a deal.
https://www.newyorkfed.org/research/dat ... remia.html

*basically an estimate of the expected return of investing in the short rate and rolling over to term, as compared to the term rate. Usually the term rate is higher than the expected return on rolling over short term treasury investments to that term, the model says now it's lower.

grok87
Posts: 8511
Joined: Tue Feb 27, 2007 9:00 pm

Re: Tips breakevens and illiquidity premia

Post by grok87 » Sat Dec 08, 2018 8:00 am

JackoC wrote:
Fri Dec 07, 2018 10:09 am
fennewaldaj wrote:
Thu Dec 06, 2018 11:28 pm
Doc wrote:
Thu Dec 06, 2018 7:17 pm

Point being, regarding TIPS. It's possible (likely?) that it is a solution to a non-existent problem. I wouldn't spend a whole lot on insurance for a non-existent problem. Especially if there is also a liquidity penalty that goes along with the insurance premium. But that conclusion is partly based on having low personal risk for higher than expected inflation.
But it doesn't seem like you are paying much for the protection right? It be a different story if say the ten year nominal yielded 2.9% and the ten year TIP real yield was 0%
Yes, the whole point of the numbers in OP is that right now you are *being paid* to hold TIPS v nominal treasuries, because the liquidity/other yield premium being paid to you is bigger than the theoretical inflation insurance premium you are paying for the inflation protection. You make more now with 10yr TIPS than nominal 10yr treasuries at an inflation rate slightly below 2%, but there's no reason to think the market's actual inflation expectation is any lower than 2%.

Above someone linked the Philly Fed's survey based inflation expectation number. The Cleveland Fed publishes a model based inflation expectation, 2.15% in 10 yrs on the most recent (mid November) run of their model.
https://www.clevelandfed.org/our-resear ... tions.aspx

This has sometimes been the case with TIPS though not always, see Cleveland Fed's graph. Often the TIPS inflation break even is higher than the calculated expectation, as 'should' be the case theoretically with no liquidity/other. Without considering liquidity, TIPS should have a lower expected return than nominal by an 'insurance' premium, as per the standard talking points against TIPS. But usually a liquidity/other premium cancels out part, all or more than all of the 'insurance' premium and TIPS can actually have higher expected return than nominal treasuries, as now.

Some people extrapolate from the generally correct idea that it's a crap shoot at best to guess what the stock market is going to do next to a more general but dubious pseudo-axiom that market pricing should never influence investing decisions. The latter idea is nonsensical, however. If there is a portion of your fixed income allocation that does not depend on being able to suddenly sell those bonds at favorable prices during a crisis, but rather can be held to maturity (and this is some, generally most, of most people's fixed income) then a liquidity premium in yield based on *other* investors' needs to suddenly sell is money in your pocket. And there's no rational argument it couldn't really be there because the market is efficient. The treasury market *is* largely efficient. Other investors are paying a fair liquidity premium for *their* liquidity need. For portions of your fixed income portfolio with a lesser liquidity need than the 'market clearing' investor, you should accept that premium. But it depends on TIPS pricing at a given time.
thanks for posting the cleveland data. that's really interesting.
this chart shows there term structure of expected inflation
https://www.clevelandfed.org/our-resear ... tions.aspx
the model was run on 11/14/18.
the figures were
Expected inflaton (cleveland fed model @11/14/18)
5 year: 2.10%
10 year: 2.15%
30 year: 2.36%

comparing to the BEIs i posted above which are
5 year: 1.71%
10 year: 1.92%
30 year: 1.99%

would result in TIPs il-liquidity premia of
5 year: 0.39%
10 year: 0.23%
30 year: 0.37%

cheers,
grok
Keep calm and Boglehead on. KCBO.

User avatar
Kevin M
Posts: 10306
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Tips breakevens and illiquidity premia

Post by Kevin M » Sat Dec 08, 2018 6:02 pm

grok87 wrote:
Sat Dec 08, 2018 8:00 am
thanks for posting the cleveland data. that's really interesting.
this chart shows there term structure of expected inflation
https://www.clevelandfed.org/our-resear ... tions.aspx
the model was run on 11/14/18.
the figures were
Expected inflaton (cleveland fed model @11/14/18)
5 year: 2.10%
10 year: 2.15%
30 year: 2.36%

comparing to the BEIs i posted above which are
5 year: 1.71%
10 year: 1.92%
30 year: 1.99%

would result in TIPs il-liquidity premia of
5 year: 0.39%
10 year: 0.23%
30 year: 0.37%

cheers,
grok
I don't think so--only true if you assume that the inflation risk yield premium of nominal Treasuries is 0%, and if you ignore differences between the nominal term risk and real term risk premia. You are chalking up all the difference to just the TIPS illiquidity premium, and ignoring the rest.

The Cleveland Fed model puts the 10-year inflation risk premium at 0.44%, so not the 0% that your conclusion assumes.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Sat Dec 08, 2018 7:12 pm

Kevin M wrote:
Sat Dec 08, 2018 6:02 pm
grok87 wrote:
Sat Dec 08, 2018 8:00 am
thanks for posting the cleveland data. that's really interesting.
this chart shows there term structure of expected inflation
https://www.clevelandfed.org/our-resear ... tions.aspx
the model was run on 11/14/18.
the figures were
Expected inflaton (cleveland fed model @11/14/18)
5 year: 2.10%
10 year: 2.15%
30 year: 2.36%

comparing to the BEIs i posted above which are
5 year: 1.71%
10 year: 1.92%
30 year: 1.99%

would result in TIPs il-liquidity premia of
5 year: 0.39%
10 year: 0.23%
30 year: 0.37%

cheers,
grok
I don't think so--only true if you assume that the inflation risk yield premium of nominal Treasuries is 0%, and if you ignore differences between the nominal term risk and real term risk premia. You are chalking up all the difference to just the TIPS illiquidity premium, and ignoring the rest.

The Cleveland Fed model puts the 10-year inflation risk premium at 0.44%, so not the 0% that your conclusion assumes.

Kevin
Does that mean that the TIPS buyer is paying .44% for inflation protection?

User avatar
Kevin M
Posts: 10306
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Tips breakevens and illiquidity premia

Post by Kevin M » Sat Dec 08, 2018 8:00 pm

gmaynardkrebs wrote:
Sat Dec 08, 2018 7:12 pm
Kevin M wrote:
Sat Dec 08, 2018 6:02 pm
The Cleveland Fed model puts the 10-year inflation risk premium at 0.44%, so not the 0% that your conclusion assumes.
Does that mean that the TIPS buyer is paying .44% for inflation protection?
The point I'm trying to make is that it's difficult to separate out the different premia. Here's what the Cleveland Fed articles says:
The inflation risk premium is a measure of the premium investors require for the possibility that inflation may rise or fall more than they expect over the period in which they hold a bond.
It's not explicitly stated here (although it is in at least one of the papers linked from the brief article), but this is a yield premium for nominal bonds (since TIPS returns adjust for unexpected inflation). The implication is that 0.44% of the 10-year nominal Treasury yield is compensating investors for unexpected inflation risk over the next 10 years.

The unexpected inflation yield premium pushes the nominal Treasury yield up. The TIPS illiquidity yield premium pushes the TIPS yield up. If these two premia were equal, and there were no other factors influencing only one or the other yield, then the breakeven inflation rate would be an accurate estimate of expected inflation.

Unfortunately the Cleveland model does not include the TIPS illiquidity premium, so we don't have a presentation of both premia in this one model.

In other words, the Cleveland model ignores the TIPS illiquidity premium, focusing mainly on the unexpected inflation risk premium in nominal Treasuries as an adjustment to BEI for estimating expected inflation, while grok87 ignores the latter and considers only the former.

To further complicate things, the Cleveland model indicates that the 10-year real term premium is +1.27%, while another model that's been mentioned indicates that the 10-year nominal term premium is negative. I haven't even tried to wrap my head around this yet.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

stlutz
Posts: 4852
Joined: Fri Jan 02, 2009 1:08 am

Re: Tips breakevens and illiquidity premia

Post by stlutz » Sat Dec 08, 2018 9:22 pm

I don't find the idea that TIPS offer a 1/4 or 1/2% in extra interest from illiquidity to be credible. TIPS trade with quite narrow spreads (< 20 bps). More than nominals, but much less than brokered CDs. Brokered CDs tend to offer about .6% or so of extra interest over Treasuries, but spreads for them are 2-3% (or more).

I would peg the illiquidity premium on TIPS as <10 bps and therefore probably not worth thinking all that much about.

Also, I don't know that you can say that economists' expectations of inflation really relate to market expectations at all. A number of folks on this forum have attempted in the past to construct synthetic historical TIPS returns before TIPS existed using these models, but they failed because this data did such a poor job to correlating to actual TIPS returns. Actual returns indicate that the market doesn't pay that much attention to economists' forecasts of inflation.

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Sat Dec 08, 2018 11:07 pm

stlutz wrote:
Sat Dec 08, 2018 9:22 pm
I don't find the idea that TIPS offer a 1/4 or 1/2% in extra interest from illiquidity to be credible. TIPS trade with quite narrow spreads (< 20 bps). More than nominals, but much less than brokered CDs. Brokered CDs tend to offer about .6% or so of extra interest over Treasuries, but spreads for them are 2-3% (or more).

I would peg the illiquidity premium on TIPS as <10 bps and therefore probably not worth thinking all that much about.

Also, I don't know that you can say that economists' expectations of inflation really relate to market expectations at all. A number of folks on this forum have attempted in the past to construct synthetic historical TIPS returns before TIPS existed using these models, but they failed because this data did such a poor job to correlating to actual TIPS returns. Actual returns indicate that the market doesn't pay that much attention to economists' forecasts of inflation.
The Lehmann Bros experience in 2008 suggests that the liquidity premium on TIPS is much higher. They fared far worse than nominal tbonds. The market is simply not as deep as for regular Treasurys. That was ten years ago, but TIPS issuance is still below that of other treasuries, which becomes especially problematic in a crisis.

fennewaldaj
Posts: 402
Joined: Sun Oct 22, 2017 11:30 pm

Re: Tips breakevens and illiquidity premia

Post by fennewaldaj » Sun Dec 09, 2018 2:10 am

gmaynardkrebs wrote:
Sat Dec 08, 2018 11:07 pm
stlutz wrote:
Sat Dec 08, 2018 9:22 pm
I don't find the idea that TIPS offer a 1/4 or 1/2% in extra interest from illiquidity to be credible. TIPS trade with quite narrow spreads (< 20 bps). More than nominals, but much less than brokered CDs. Brokered CDs tend to offer about .6% or so of extra interest over Treasuries, but spreads for them are 2-3% (or more).

I would peg the illiquidity premium on TIPS as <10 bps and therefore probably not worth thinking all that much about.

Also, I don't know that you can say that economists' expectations of inflation really relate to market expectations at all. A number of folks on this forum have attempted in the past to construct synthetic historical TIPS returns before TIPS existed using these models, but they failed because this data did such a poor job to correlating to actual TIPS returns. Actual returns indicate that the market doesn't pay that much attention to economists' forecasts of inflation.
The Lehmann Bros experience in 2008 suggests that the liquidity premium on TIPS is much higher. They fared far worse than nominal tbonds. The market is simply not as deep as for regular Treasurys. That was ten years ago, but TIPS issuance is still below that of other treasuries, which becomes especially problematic in a crisis.
Wouldn't at least part of Tips performance in that particular crisis be related to fear of deflation. In a bout of serious deflation the real yield of Tips should go up.

stlutz
Posts: 4852
Joined: Fri Jan 02, 2009 1:08 am

Re: Tips breakevens and illiquidity premia

Post by stlutz » Sun Dec 09, 2018 2:34 am

gmaynardkrebs wrote:
Sat Dec 08, 2018 11:07 pm
The Lehmann Bros experience in 2008 suggests that the liquidity premium on TIPS is much higher. They fared far worse than nominal tbonds. The market is simply not as deep as for regular Treasurys. That was ten years ago, but TIPS issuance is still below that of other treasuries, which becomes especially problematic in a crisis.
Were TIPS actually illiquid in 9/2008? By that, I mean if I tried to sell my TIPS at the same time would I have taken a 10% haircut on trading spreads alone?

TIPS are definitely a much smaller market than nominals. That's not the same thing as being hard to trade. The residential real estate market is massive; it's also quite illiquid. If I absolutely had to sell my house tomorrow, I could probably do so. I'd pay a lot to make it happen.

I don't think the same was true of TIPS in Sept. 2008. I could have traded them with larger than normal spreads, but my guess is that such spreads were still lower than corporate bonds at the time. TIPS prices moved significantly because a large holder unloaded their position. Mayhem would happen in the market for nominal bonds if, say, the government of China decided to unload all of their holdings all at once.

Does anyone have any evidence as to what trading spreads for TIPS in 9/2008 actually were? It seems clear to me that we are all guessing here.

Chip
Posts: 2376
Joined: Wed Feb 21, 2007 4:57 am

Re: Tips breakevens and illiquidity premia

Post by Chip » Sun Dec 09, 2018 8:19 am

stlutz wrote:
Sun Dec 09, 2018 2:34 am
Were TIPS actually illiquid in 9/2008? By that, I mean if I tried to sell my TIPS at the same time would I have taken a 10% haircut on trading spreads alone? [...] Does anyone have any evidence as to what trading spreads for TIPS in 9/2008 actually were? It seems clear to me that we are all guessing here.
They weren't illiquid, but spreads did widen significantly. Here's some contemporaneous spread info on one trade I made in 9/08, documented in this thread:
Chip » Mon Sep 22, 2008 1:13 pm (actually 2:13 my time) wrote: Just managed to get an in-the-spread fill at Fido. At least it looks like it. The 2.00 1/26 issue had a bid/ask of about 95.25/95.85. Put an order in at 95.7. It filled in about 3 minutes, at 1:55. Bid/ask right afterwards was about 95.1/95.707 (the ask is exact).
In another post in 11/07 I mentioned that 8-10 year maturity TIPS spreads were .1% and 18-20 year were .2%.

Based on only those two posts, TIPS spreads roughly tripled in those 10 months.

Your point, that downside price volatility in TIPS is often described as/conflated with illiquidity here, is a good one. I've been guilty of it myself.

Before the financial crisis it was a reasonably common theme here that TIPS were a perfectly acceptable choice for 100% of one's bond allocation. That turned out not to be true if one of the expected uses of those TIPS was to hold nominal value to be used for rebalancing "ammunition" during financial crises. I won't forget that stocks were down big, TIPS were down big, and I had precious little available to rebalance into them.

I still have TIPS, but now have a decent allocation to nominal bonds as well.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Sun Dec 09, 2018 8:41 am

Chip wrote:
Sun Dec 09, 2018 8:19 am
Your point, that downside price volatility in TIPS is often described as/conflated with illiquidity here, is a good one. I've been guilty of it myself.

Before the financial crisis it was a reasonably common theme here that TIPS were a perfectly acceptable choice for 100% of one's bond allocation. That turned out not to be true if one of the expected uses of those TIPS was to hold nominal value to be used for rebalancing "ammunition" during financial crises. I won't forget that stocks were down big, TIPS were down big, and I had precious little available to rebalance into them.

I still have TIPS, but now have a decent allocation to nominal bonds as well.
If very few want to buy your asset but you lower your asking price enough someone will eventually buy it. I call that situation illquidity.

In any case I was in a similar situation as you in '08 and sold my entire TIPS ladder to rebalance. I managed to buy some of it back a few months later when they were still on sale. Ten years later I have decided that I don't need that kind of inflation protection any more.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

Chip
Posts: 2376
Joined: Wed Feb 21, 2007 4:57 am

Re: Tips breakevens and illiquidity premia

Post by Chip » Sun Dec 09, 2018 8:47 am

Doc wrote:
Sun Dec 09, 2018 8:41 am
If very few want to buy your asset but you lower your asking price enough someone will eventually buy it. I call that situation illquidity.
I have to disagree, Doc. I believe that if you can easily buy and sell with reasonable spreads that's the very definition of liquidity.

Did you think that stocks were illiquid during the financial crisis?

My poster child for something that is illiquid: timeshares.

User avatar
vineviz
Posts: 2357
Joined: Tue May 15, 2018 1:55 pm

Re: Tips breakevens and illiquidity premia

Post by vineviz » Sun Dec 09, 2018 8:55 am

Chip wrote:
Sun Dec 09, 2018 8:47 am
Doc wrote:
Sun Dec 09, 2018 8:41 am
If very few want to buy your asset but you lower your asking price enough someone will eventually buy it. I call that situation illquidity.
I have to disagree, Doc. I believe that if you can easily buy and sell with reasonable spreads that's the very definition of liquidity.

Did you think that stocks were illiquid during the financial crisis?

My poster child for something that is illiquid: timeshares.
In financial economics, liquidity is defined as the ability to transact without affecting the price. In 2008 the order imbalance of TIPS unquestionably exceeded the liquidity of that asset class at that moment in time.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Sun Dec 09, 2018 9:09 am

I guess there are several definitions of liquidity. Wide spreads is one definition, but it was not the one I was focusing on with regard to the situation with Lehmann Brothers in 2008. The definition of illiquidity in that case was "if you sell, you will move the market." That does not happen in liquid markets. If Lehman Brothers have been trying to dump a similar amount of nominal treasuries, I doubt that the prices would have dropped anything close to what they did during those weeks. In fact, I don't think they would have dropped at all for that reason alone. The reason it happened with Tips, is that the market depth was simply not there, and thus, the market price moved against them when they tried to sell. lacked. I'm not sure if that is still true today, but I suspect it is, because TIPS issuance is still pretty low. That leads me to think there is still a significant liquidity premium earned by tips holders.

Chip
Posts: 2376
Joined: Wed Feb 21, 2007 4:57 am

Re: Tips breakevens and illiquidity premia

Post by Chip » Sun Dec 09, 2018 10:52 am

vineviz wrote:
Sun Dec 09, 2018 8:55 am
In financial economics, liquidity is defined as the ability to transact without affecting the price. In 2008 the order imbalance of TIPS unquestionably exceeded the liquidity of that asset class at that moment in time.
Transact how much? How much price effect? It seems patently obvious to me that there are NO markets that can meet that particular definition of a liquid market given a transaction of sufficient size. Even the nominal Treasury market. If that's the case, how do you differentiate between a liquid and an illiquid market without a transaction that tests that liquidity?

I can move the price of the some of ETFs I buy with a puny 100k order. Would that cause you to consider them illiquid?

User avatar
vineviz
Posts: 2357
Joined: Tue May 15, 2018 1:55 pm

Re: Tips breakevens and illiquidity premia

Post by vineviz » Sun Dec 09, 2018 11:34 am

Chip wrote:
Sun Dec 09, 2018 10:52 am
vineviz wrote:
Sun Dec 09, 2018 8:55 am
In financial economics, liquidity is defined as the ability to transact without affecting the price. In 2008 the order imbalance of TIPS unquestionably exceeded the liquidity of that asset class at that moment in time.
Transact how much? How much price effect? It seems patently obvious to me that there are NO markets that can meet that particular definition of a liquid market given a transaction of sufficient size. Even the nominal Treasury market. If that's the case, how do you differentiate between a liquid and an illiquid market without a transaction that tests that liquidity?
Campbell, Lo, and MacKinlay literally wrote the book on financial market econometrics and they define liquidity this way:
“Ability to buy or sell significant quantities of a security quickly, anonymously, and with minimal or no price impact.”
You can definitely say that liquidity is the measure the price impact from a trade of a given size. For any asset, there is some order size below which the price will not be impacted and above which it will be impacted.

There is a volume of literature on ways to estimate and/or quantify liquidity, all of which is probably beyond the needs of this discussion. If you are interested then I can recommend a 2008 paper by Goyenko et al. that summarizes many of them and compares them.

https://kelley.iu.edu/cholden/Do%20Meas ... -02-22.pdf

Although bid/ask spreads are not a direct measure of liquidity, it turns out that they are a reasonably good proxy (at least for bonds).
Chip wrote:
Sun Dec 09, 2018 10:52 am
I can move the price of the some of ETFs I buy with a puny 100k order. Would that cause you to consider them illiquid?
I don't believe that a single $100k trade CAN move the price of any commonly held index ETF. ETF liquidity is almost entirely a function of the liquidity of the underlying securities, so a tiny trade like that will only have a price impact in the most esoteric of niche asset classes.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

stlutz
Posts: 4852
Joined: Fri Jan 02, 2009 1:08 am

Re: Tips breakevens and illiquidity premia

Post by stlutz » Sun Dec 09, 2018 12:37 pm

Chip wrote:
Sun Dec 09, 2018 8:19 am
Here's some contemporaneous spread info on one trade I made in 9/08, documented in this thread
Good stuff--thanks! :sharebeer

User avatar
Kevin M
Posts: 10306
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Tips breakevens and illiquidity premia

Post by Kevin M » Sun Dec 09, 2018 2:07 pm

stlutz wrote:
Sat Dec 08, 2018 9:22 pm
Brokered CDs tend to offer about .6% or so of extra interest over Treasuries, but spreads for them are 2-3% (or more).
I know this is a tangential topic, but since I happen to have some data on it ...

First, the CD/Treasury yield spread can vary greatly over time (time series variation), and it can vary significantly with maturity (cross-sectional variation). Here is the current picture out to 5-year maturity using yields from Fidelity's yield summary page:

Image
(the "IRA" suffix indicates that the state tax exemption of Treasuries is not considered here).

CD and Treasury yields are pretty much equal out to 1-year maturity, and the CD yield premium increases monotonically at longer maturities.

Second, the average bid/ask spread for CDs where both bid and ask prices are provided has tended to by 0.75%-0.85% range for CDs out to 3-year maturity when I'd downloaded quotes from Fidelity. For my last download from Fidelity, average was 0.75%, max was 2.11%, and minimum was 0.10% (I calculate the spread as ask/bid -1). Out of 1,102 quotes, there were bid quotes for 858 or 78%. However, I wouldn't be surprised if the bid/ask was higher for CDs for which there were no active bid quotes; i.e., when you must solicit bids.
Also, I don't know that you can say that economists' expectations of inflation really relate to market expectations at all. A number of folks on this forum have attempted in the past to construct synthetic historical TIPS returns before TIPS existed using these models, but they failed because this data did such a poor job to correlating to actual TIPS returns. Actual returns indicate that the market doesn't pay that much attention to economists' forecasts of inflation.
The Cleveland Fed model uses inflation swaps, which of course are market based, as well as survey data. It seems to have generated real rates reasonably close to TIPS yields with the notable exceptions being the first few years after TIPS were introduced, and during the financial crisis, when TIPS were "underpriced" relative to the model; i.e., TIPS yields much higher than the model real rate. A TIPS yield illiquidity premium, which is not included in the model, could be one explanation for these periods of deviation.

Just because the TIPS yield illiquidity premium may have been high at times in the past does not mean that it's significant now.

Regarding TIPS being relatively expensive or inexpensive now (they were extremely inexpensive relative to the model in late 2008), the November Cleveland Fed model 10-year real yield was 1.31% with the TIPS 10-year yield at 1.17%. So based on this model, you are paying a bit of a premium for TIPS now (lower yield = higher price). If there were a TIPS yield illiquidity premium, and it was the only factor not captured by the model, one would think that the TIPS yield would be higher than the model yield (as it was in late 2008).

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Sun Dec 09, 2018 5:04 pm

Regarding illiquidity and spreads: It's not only the spread that is the metric but also the size of the bid & offer. A spread of 1 bps on a bid and offer of 100 shares may mean nothing by itself. We need to look at the rest of the book also. And then there has been some mention recently that there are "market makers" whose "bid & ask" don't show in the book at all. If there are no "market maker" buyers we are in a whole different ball game.
Kevin wrote:Just because the TIPS yield illiquidity premium may have been high at times in the past does not mean that it's significant now.
More over if the stock market crashes on Tuesday the illiquidity premium on Monday might be meaningless.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

JackoC
Posts: 360
Joined: Sun Aug 12, 2018 11:14 am

Re: Tips breakevens and illiquidity premia

Post by JackoC » Sun Dec 09, 2018 5:48 pm

Kevin M wrote:
Sat Dec 08, 2018 8:00 pm

1. The unexpected inflation yield premium pushes the nominal Treasury yield up. The TIPS illiquidity yield premium pushes the TIPS yield up. If these two premia were equal, and there were no other factors influencing only one or the other yield, then the breakeven inflation rate would be an accurate estimate of expected inflation.

2. Unfortunately the Cleveland model does not include the TIPS illiquidity premium, so we don't have a presentation of both premia in this one model.

3. To further complicate things, the Cleveland model indicates that the 10-year real term premium is +1.27%, while another model that's been mentioned indicates that the 10-year nominal term premium is negative. I haven't even tried to wrap my head around this yet.
1. I agree. Or in the example previously discussed, if the 10 yr TIPS breakeven is 1.92%, the 10 yr inflation expectation is 2.15% and the inflation risk premium is 0.44%, the illiquidity/other premium would be 0.67%.

The TIPS investor pays (or gives up relative to the nominal treasury yield) 0.44% to get rid of inflation risk, and receives 0.67% to bear the greater illiquidity of TIPS compared to nominal, an expected return 0.23% higher with TIPS than nominal: 1.92% break even with 2.15% inflation expectation.

2. It doesn't really have to because whatever is not explained in the difference between expected inflation and TIPS breakeven, by inflation risk premium, must be liquidity/other. And including 'other', we don't have argue exactly what 'liquidity' means and why some particular definition would be worth 0.67%. It would be all the factors we've seen in TIPS behavior that would make you not want to hold them *if you are going to sell them prior to maturity*. 2008-9 price behavior? That's part of it. It's often given as argument why TIPS suck, right? And it would be if you were going to sell them. But not if you are going to hold them to maturity.

Anyway, there's not even any particular reason to look at model output for inflation risk premium, besides general curiosity. If you believe the inflation expectation output (and that's reinforced at the moment by Philly Fed survey based inflation expectation) and it's higher than the breakeven, and if TIPS are to be fitted into a layer of your fixed income portfolio unlikely to be subject to sale in rebalancing (if you rebalance, nor do you have to rebalance by actually selling physical bonds even if you do rebalance), then TIPS are definitively more attractive than nominals. You're getting inflation protection you know is worth something (whether or not 0.44%), paying for it by bearing illiquidity you've no reason to care about in that part of your portfolio, then getting .23% more after that.

This is not a simple answer to what to buy though for two reasons: you may get paid even more for illiquidity (you've no reason to care about) with best CD's than TIPS, so between those two you might really need to figure what inflation protection is worth *to you*. And, the comparison doesn't tell you what maturity is most favorable.

3. The Cleveland Fed model 1.27% 'real risk premium' for 10 yr and the NY Fed model -0.45% 'term premium' for 10 yr (what it was in November, now it's like -0.60%) are not directly comparable. Though the C terminology is 'risk premium', that number must also include the equilibrium real rate, IOW risk premium relative to getting a return of inflation flat. The NY premium is instead the difference between a rolling investment at the non-risk adjusted expected short rates over 10 yrs v the 10 yr term rate. Those future non-risk adjusted short rates already include a real component on top of expected inflation. The NY model is just trying to tease out what additional premium the investor gives or receives to lock in all those forward short rates now, which would mean a rolling investment at the risk neutral short forward rates simply deduced from today's curve, IOW the same thing as investing in the 10 yr note now.

User avatar
Doc
Posts: 8749
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Tips breakevens and illiquidity premia

Post by Doc » Sun Dec 09, 2018 7:47 pm

I have a vague recollection that the introduction of TIPS was partly based on developing a market prediction of future inflation. If so it was a complete failure as this theard illustrates. :D
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

gmaynardkrebs
Posts: 937
Joined: Sun Feb 10, 2008 11:48 am

Re: Tips breakevens and illiquidity premia

Post by gmaynardkrebs » Sun Dec 09, 2018 7:53 pm

Doc wrote:
Sun Dec 09, 2018 7:47 pm
I have a vague recollection that the introduction of TIPS was partly based on developing a market prediction of future inflation. If so it was a complete failure as this theard illustrates. :D
You’re right. Larry Summers idea. The public doesn’t understand them, or the concept of real returns. Pretty tough sell when there hasn’t been significant inflation in many years.

Angst
Posts: 1874
Joined: Sat Jun 09, 2007 11:31 am

Re: Tips breakevens and illiquidity premia

Post by Angst » Sun Dec 09, 2018 10:06 pm

grok87 wrote:
Thu Dec 06, 2018 9:10 am
The bond market is sending some interesting signals
Right now:

Nominal treasuries;
5 yr: 2.75
10 yr: 2.88
30 year: 3.14

Tips
5 yr: 1.04
10 yr: 0.96
30 yr: 1.15

Breakeven inflation
5 year: 1.71
10 yr: 1.92
30 yr: 1.99
5x5: 2.13
20x10: 2.025

I think long run fair value breakeven inflation would be about 2.25%. Because the fed targets PCE inflation at 2% and CPI usually runs 25 bps higher. Since 20x10 is at 2.02 that implies. 23 bp illiquidity Premium for long dated tips.

Interesting.
grok,
this Campbell, Shiller and Viceira 2009 paper (below) from the Brookings Institution gave what I thought was an interesting, historically contemporaneous perspective to TIPS (and inflation-linked Gilts as well) and the apparent illiquidity premium during the Lehman episode. It rarely allows itself to refer to illiquidity without calling it the "risk and illiquidity premium" and frequently refers to and demonstrates the difficulty in trying to characterize and understand the nature of the premium.

https://www.brookings.edu/wp-content/up ... mpbell.pdf

From the perspective of someone who only cares about building a LMP, I embrace the premium. What's not to like?

hdas
Posts: 413
Joined: Thu Jun 11, 2015 8:24 am

Re: Tips breakevens and illiquidity premia

Post by hdas » Sun Dec 09, 2018 10:36 pm

I believe Ray Dalio and Bridgewater were instrumental in the creation/development of the market:

From 1999:
......But not all professional investors have ignored TIPS. Ray Dalio, CEO of Bridgewater Associates, a private $20 billion-plus fund, is probably the largest holder of inflation-indexed bonds in the world. Out of a total international market of inflation-protected government bonds of $220 billion or so, he holds $2.5 billion......
Stay the course and buy some more.

Chip
Posts: 2376
Joined: Wed Feb 21, 2007 4:57 am

Re: Tips breakevens and illiquidity premia

Post by Chip » Tue Dec 11, 2018 6:07 am

Doc wrote:
Sun Dec 09, 2018 5:04 pm
Regarding illiquidity and spreads: It's not only the spread that is the metric but also the size of the bid & offer. A spread of 1 bps on a bid and offer of 100 shares may mean nothing by itself. We need to look at the rest of the book also. And then there has been some mention recently that there are "market makers" whose "bid & ask" don't show in the book at all. If there are no "market maker" buyers we are in a whole different ball game.
Exactly. Though I'm not talking about TIPS, I've had many experiences where there is little depth to the order book. Perhaps my "moving prices" comment was inaccurate, but it sure feels like prices are moving when my order is filled in pieces at prices that vary a good bit from the midpoint of the NBBO. For example, NBBO is 1x1 at 100.00/100.02. But to get a guaranteed fill of a 1,000 share buy I might have to set my limit order at 100.15. That's the "illiquidity shown in the spread" that I was trying to describe.

Though I will admit that I've occasionally had a market maker swoop in and fill an order at better prices than one would expect from the book. In the example above, I put in the 1,000 share limit buy order at 100.15 and the entire order is filled at 100.01999. In that case I guess there was just the illusion of illiquidity. The real spread isn't on display, as you noted.

Post Reply