Larry & Rick

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
User avatar
Rick Ferri
Posts: 9703
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Post by Rick Ferri »

Cubsfan,

Corporate bonds in general have not delivered the risk premiums that investors expected during the 10 year period that you mentioned. That is true for investment grade corporates as well as high yield. For example, the VG Inter-Term Investment Grade fund only achieve results on par with Treasuries over the past ten years (the VG Inter-Term Investment Grade fund was titled the VG Inter-Term Corporate bond fund until recently).

However, one cannot invest based on near-term past performance. In an efficient market, the expectation of corporate bond investors is that they will achieve a higher risk premium over Treasuries.

Since inception of the Vanguard High-Yield Corporate Fund Investor Shares (VWEHX), the net-of-fee annualized return has been 9.17% through June 2007. In contrast, the return of the Lehman Brothers 1-10 Year Intermediate Credit (Corporate) Index has been 8.90%, the Lehman Brothers Aggregate Bond Index has been 8.74%, and the return of the Lehman Brothers U.S. Government Index has been 8.59% (there are no fees in the indexes). So, there has been a realized efficient market spread.

Do these numbers show stellar results for corporate debt in general? No. However, if the corporate bond market is efficient in the long-run, then the returns for corporate bonds should be higher in the future. If they are not higher, then either 1) I am over-estimating the risk premium paid on all corporate debt regardless of quality (Larry's contention), or 2) spreads on new-issue corporate bonds will permanently widen so that a larger risk premium is paid, or 3) US government securities are actually more risky than people believe.

Perhaps over the last 30 years the global market has increased the premium required on U.S. government debt rather than decrease the premium on U.S. corporate debt. IN other words, in the eyes of global investors, U.S. corporate debt is becoming more attractive than U.S. government debt. If that were true, it would explain the seemingly lower premiums on corporate debt over Treasuries. That scenario is not out of the question!

Rick Ferri
Cubsfan
Posts: 101
Joined: Fri Feb 23, 2007 1:38 pm

Risk vs. reward on high yield

Post by Cubsfan »

larryswedroe wrote:Cubsfan

YOur points are exactly the ones I have been making

And while it might be data mining, high yield in the ten year period with the figures shown underperformed the intermediate treasuries, which are much less risky and have lower correlations to stocks (so better diversifier) and this is period that ends with junk at lowest spreads ever.

One of the reasons for the low performance is the call risk inherent that no one seems to talk about. And it is not just from lower interest rates that calls happen but from improvement in credit ratings.
It isn't a coincidence that we're making the same points. Where do you think I learned that stuff?

The Vanguard high yield fund has been around since 1984, so there is a little more history available. If I go from 1985-2006, the CAGR was 8.76% for high yield vs. 8.11% for intermediate treasuries. Expanding the time period brings in only one other year where Vanguard S&P 500 index fund (VFINX)lost money. That was 1990, and intermediate treasury fund made 9.7% while high yield lost 5.85%. Again, the risk of high yield showed up at the wrong time.

From 1985-2006, the correlation between intermediate treasury and VFINX was .19. Correlation between high yield and VFINX was .62.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Post by larryswedroe »

few quick thoughts before I head off for my trip

Rick keeps saying that I believe the risk premium is overstated, meaning prices are wrong. I have no idea where he gets that from for I have never said that. What I have said is the risk premium is obtained in inefficient ways and that the asset class doesn't mix well with the rest of the portfolio and the goal of fixed income in portfolio.

Second, while indices have no expenses funds do. And with treasuries you don't need a fund as there is no diversification benefit. And same thing true with gov't agencies where you pick up bit more yield. So you could buy them on your own (forgoing bit of convenience) and get the index return but not true for funds. So the gap is actually less than (or can be) then showed. For example, we Rick can buy for his clients individual agency bonds but Rick has to buy a fund to gain exposure to junk.

The spreads you see as yields are not realized because of call risk, which is priced, and the defaults that do occur.

Thus once all this is accounted for there is a very small premium, very small amount of unique risk--and the risk has nasty tendency to show up when it does the most damage--in other words the low correlations rise at the wrong time
User avatar
Rick Ferri
Posts: 9703
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Post by Rick Ferri »

One of the reasons for the low performance is the call risk inherent that no one seems to talk about. And it is not just from lower interest rates that calls happen but from improvement in credit ratings.

That is good point. In a falling interest rate environment, callable corporate bonds will under-perform non-callable government bonds. Corporate interest rates were 18% in 1980. Now they are 6%. Lots of corporates have been called, lowing the returns of that sector. This is also the reason GNMAs have not achieved the returns expected over the period.

Over the next 25 years, with starting rates at 6%, will we see the large number of calls? I do not believe so. In that case, the premium from corporates should be higher.

Thanks for helping make my case, Larry!

Rick Ferri
Valuethinker
Posts: 48944
Joined: Fri May 11, 2007 11:07 am

Re: Risk vs. reward on high yield

Post by Valuethinker »

Cubsfan wrote:
larryswedroe wrote:Cubsfan

YOur points are exactly the ones I have been making

And while it might be data mining, high yield in the ten year period with the figures shown underperformed the intermediate treasuries, which are much less risky and have lower correlations to stocks (so better diversifier) and this is period that ends with junk at lowest spreads ever.

One of the reasons for the low performance is the call risk inherent that no one seems to talk about. And it is not just from lower interest rates that calls happen but from improvement in credit ratings.
It isn't a coincidence that we're making the same points. Where do you think I learned that stuff?

The Vanguard high yield fund has been around since 1984, so there is a little more history available. If I go from 1985-2006, the CAGR was 8.76% for high yield vs. 8.11% for intermediate treasuries. Expanding the time period brings in only one other year where Vanguard S&P 500 index fund (VFINX)lost money. That was 1990, and intermediate treasury fund made 9.7% while high yield lost 5.85%. Again, the risk of high yield showed up at the wrong time.

From 1985-2006, the correlation between intermediate treasury and VFINX was .19. Correlation between high yield and VFINX was .62.
One has to be cognisant of the history of high yield bonds.

Phase 1 - up to the early 1980s - 'Fallen Angels' that were former investment grade bonds that had gone sour. A very bright undergraduate at Wharton in the 1970s (Michael Milliken) picked up on research that showed that such bonds outperformed investment grade bonds, despite the higher risks.

Phase 2 - about 1979-1989. Armed with that information, Michael Milliken at Drexel Burnham Lambert, forged an extraordinary empire raising money for 'risky' companies and risky deals, and moving that debt off to Savings & Loans, insurance companies etc. If you needed to do a HY deal, you had to go to MM and Drexel-- they were the only port of call.

1986-- Milliken the most highly paid man in the world, about $1 billion in salary and bonus.

1989-90. Milliken is arrested for insider trading (see Wall Street). Drexel Burnham Lambert goes bust. Savings & Loan debacle hits. US goes into recession. Junk bonds reach 12% default rate. Market collapses (no market maker).

It turned out that the original case for junk bonds was false. In the sense that the 'fallen angels' were not the same sort of bonds as were issued in the 80s, which were by fundamentally far riskier companies. So recovery rates post default were much lower.

1994 - 'Perfect Storm' in bond markets. Bond prices dropped 20% (hedge funds unwinding positions) and bond funds returned c. -10% (income was higher in those days).

1990s & 00s. High yield becomes an ordinary asset class with lots of players, and much better understood credit risks.

2001 - a lot of credit defaults, but localised into the 'dot com' bubble (Worldcom, Enron), also electric utilities. Warren Buffett takes a position in junk bonds.

So:

- any data pre about 1991 is somewhat suspect. It was a different world in the 80s

- we've only had one really tough, nasty recession in junk bonds (1990).
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Post by larryswedroe »

Rick
So now you are market forecaster? (:-))

Let's address this key point about call risk. First, rates can go a lot lower. Anyone recall just few years ago? How about Japan where rates went close to zero. One mistake even very smart people make is treating even the unlikely (and in this case I would not even argue it is unlikely) as impossible. Note also that the call risk is very large and main reasons over long term corporates have not delivered much if any ex-post risk premiums. Consider the following. IN last ten years we have seen spreads come way down on junk--in fact to lowest historical levels. The best possible outcome for junk. Clearly could have been much worse and spreads could have widened dramatically and losses could have been huge from defaults. Neither happened. The risks did not show up. Yet despite this the fact remains that junk underperformed riskless treasuries for the last ten years. Junk got the environment an investor could have hoped for and did not outperform. OUCH. To me that is almost case closed. But there is much more. Let's consider

Second, the only reasons to buy longer term bonds (that historically have not provided much if any term risk premiums) is to protect (as Swensen makes the case) against deflation risks and to hedge reinvestment risk. Now here we get into how junk mixes with rest of portfolio--the only right way to think about an asset. Now junk is longer term debt. Now if you get deflation, that will kill junk bonds as credit risks show up. The second problem is that if you get deflation you will see all the bonds that can be called because they can refi do so, giving you back the reinvestment risk you were trying to avoid in the first place. Double whammy. OUCH, OUCH.

And of course you have the location problem from tax perspective.
QED
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T »

.
FWIW – here are the rolling 12 month correlation averages between Vanguard’s HY fund and TSM, and LB Intermediate Treasury Index and TSM for periods when TSM’s 12 month rolling returns were positive and for periods when TSM experienced moderate to extreme losses.

January 1979 to Sept 2006 [period for which I have data].

Code: Select all

                               Average 12 month correlations: 
TSM 12 month returns:            HY-TSM       Int. Treasury-TSM
Positive                          0.48             0.24
Between 0% to –10%                0.62             0.18
Between –10% to –20%              0.65            -0.21                  
More than –20% loss               0.63            -0.57 
If the equity allocation is not TSM but is tilted to small cap and value stocks (with a size and value loading of 0.2 and 0.4 respectively) [labeled TSMSV in the table below] then the correlations change to the following.

Code: Select all

                              Average 12 month correlations: 
TSMSV 12 month returns:         HY-TSMSV       Int. Treasury-TSMSV
Positive                          0.53             0.24
Between 0% to –10%                0.65            -0.03
Between –10% to –20%              0.75            -0.34                  
More than –20% loss               0.73            -0.86 
Correlations between the Vanguard HY fund and equities tended to increase during market declines while the correlations between LB intermediate treasuries and equities did the opposite, they shifted to increasingly negative during equity market declines. This pattern was even more pronounced for declines in a small cap and value tilted equity allocation.

If the role of fixed income is to reduce the downside risk of an equity biased portfolio then HY does not seem to fill this role (i.e. correlations with equities increase at the wrong time). In contrast, intermediate treasuries do seem to fill this role (i.e. correlations decrease and turn negative at just the right time), even more so with a small and value tilted portfolio.

Just my take.

Robert
.
User avatar
Rick Ferri
Posts: 9703
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Post by Rick Ferri »

Robert T,

Thank you for that very good analysis. It does a great job explaining the economic risk in HY that is a times similar to equity, and why HY investors should have at least 50% of their fixed income investments in government securities.

Rick Ferri
User avatar
Chas
Posts: 846
Joined: Sat Mar 24, 2007 8:41 pm
Location: America

VG Hi-Yield is BB+ ????

Post by Chas »

Rick Ferri wrote:
Knock off 0.5% for defaults, and 1.5% seems like a adequate risk premium for well managed BB+ portfolio.

Rick Ferri
I have been following this thread waiting for some one else to say this, but no one has, so here it is. The VG Hi-Yield VWEHX fund has about 50% of its holdings in B, below B, and Not Rated bonds according to M* portfolio. What’s going on with all this BB+ stuff?

Chas
Chas | | The course of true love never did run smooth. Shakespeare
User avatar
Rick Ferri
Posts: 9703
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Post by Rick Ferri »

This is straight of the Vanguard website:

Number of Bonds 261
Yield to Maturity 7.1%
Average Coupon 7.6%
Average Maturity 7.4 years
Average Quality* Ba3
Average Duration 4.8 years
Short-Term Reserves 2.4%
Fund Total Net Assets $9.8 Billion

*These are Moody's rating. S&P ratings may differ:

Aaa 4.1%
Aa1 0.0%
Aa2 0.0%
Aa3 0.0%
A1 0.0%
A2 0.0%
A3 0.0%
Baa1 0.0%
Baa2 0.3%
Baa3 1.9%
Ba1 13.3%
Ba2 13.2%
Ba3 15.3%
B1 19.1%
B2 13.7%
B3 9.4%
Caa1 7.4%
Caa2 0.2%
Caa3 0.0%
Ca 0.0%
C 0.0%
DEFAULT 0.0%
NOT RATED 2.1%
Total 100.0%

Caa and non-rated bonds make up less than 10% of the fund.
User avatar
Norbert Schlenker
Posts: 544
Joined: Mon Feb 19, 2007 10:06 pm
Location: The Dry Side of the Wet Coast

Post by Norbert Schlenker »

Rick, Chas' point is correct. Close to half of the portfolio is below Ba3. The median bond is Ba3, but Ba3 (Moody's) = BB- (S&P), not BB+ as you wrote above.

And just to be evenhanded re expert opinions in this thread ...
larryswedroe wrote:First, rates can go a lot lower. Anyone recall just few years ago? How about Japan where rates went close to zero. One mistake even very smart people make is treating even the unlikely (and in this case I would not even argue it is unlikely) as impossible.
Good points from Larry. It's ironic that this argument does not apply when he's discussing TIPS and the real rate risk they have.
User avatar
LH
Posts: 5490
Joined: Wed Mar 14, 2007 2:54 am

Post by LH »

Nice chart Robert T, really adds to the discussion.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Post by larryswedroe »

Norbert
Dont know how you can make that last statement about my view on TIPS and real rate risk. In fact I have written extensively about that, including providing tables to people suggesting investors consider at least shifting allocations and shifting maturity approaches to TIPS.
User avatar
Norbert Schlenker
Posts: 544
Joined: Mon Feb 19, 2007 10:06 pm
Location: The Dry Side of the Wet Coast

Post by Norbert Schlenker »

Larry, I've had discussions/debates/arguments with you in a fair number of threads now about the risk in TIPS. You have repeatedly downplayed the risk, you have denied that there is much risk at all, you have relied on arguments that returns on TIPS should act in a certain way and I have had to demonstrate that, whether you and I like it or not, they have not acted as we think they should.

Now you have a book out - one of the few of yours I don't have - that I am told counsels people to change their allocations of TIPS versus nominal bonds depending on what real rate is being offered. When I described that as market timing, you at first denied it and then grudgingly admitted it.

So let me go back to what you wrote just before you went on vacation (and which you've written many times before about asset classes other than TIPS).
One mistake even very smart people make is treating even the unlikely (and in this case I would not even argue it is unlikely) as impossible.
You're very smart. Consider for just a minute or two whether it is just unlikely or absolutely impossible that TIPS yields go higher, perhaps even back to 4%. Every time you write about TIPS here, the impression you leave is that it's impossible. So I'm teasing you a little with your own words to try to get you to apply your own argument to this one asset class that you appear to believe is exempt from your own rule.

I'm asking for a little consistency from you. That's all.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Post by larryswedroe »

Norbert
Unfortunately nothing you said in your last post was correct.

Let's begin with the first point. The risk for TIPS is basically zero for those that buy and hold. They get the real rate of return they bought. And since they should be held in tax advantaged accounts there is no tax risk as some people try to point out.

Now having said that I have repeatedly suggested to people that there is a reasonable expectation that real rates will stay within some boundary--unlike nominal rates. I have stated that something like 1% on downside and 4% on upside is likely. Note I have also said that there is risk that it could go beyond that range. I have also said that because the market price IMO is likely to be the right price even at 1% the odds of it going up or down is likely to be the same. However, the cost of being wrong (having switched say to nominal bonds at that point) is relatively low compared to the risk on the other side---rates cannot go below zero but they could go to 4% or more. And at 4% while rates could go higher, the worst case is that you have locked in a very high real rate (about rate most think equities will provide or at least close to it), and it is not likely to go much higher because it would kill the economy (though it is possible).

I have been extremely consistent on this issue.

Now to your next point--about me counselling people to shift allocations. I never said that. What I have said is the following

There are two good strategies here IMO. The first is to own a fixed ratio of nominal ST bonds and TIPS. That diversifies real rate risk and forces a buy low and sell high strategy if you rebalance. The other is to adopt a shifting allocation and/or a shifting maturity approach based on the idea that TIPS are likely to trade in a fairly narrow band. Note I have always said LIKELY. IMO this approach, which I personally use, will deliver higher returns. But there is clearly risk it will not.

Having said that personally I dont care about whether it in the end delivers higher returns or not, because I don't have clear crystal ball and the strategy is either right or not ex-ante. IN other words if at 3.5% I go 100% TIPS and all long term TIPS and rates go higher, I don't care because I got what I wanted a high real rate of return with no risk. The only loss is opportunity loss which is that if I had waited and held nominal bonds longer I could have gotten higher return. But I would have met my objective and reduced risk of real rates falling.

There are two risks we have higher--both TIPS and nominal bonds hedge inflation risk, though TIPS do a significantly better job in long term (at the "cost" of interim price risk and higher NOMINAL volatility which IMO is totally irrelevant unless you need the funds and then you should not buy long term TIPS but short term ones to reduce that risk). The difference is that TIPS lock in the real rate and nominal bonds float it. Which is why when real rates are well below historical averages IMO the odds are in your favor to own nominal bonds and vice versa (note I use the word odds).

Hope the above is helpful
User avatar
LH
Posts: 5490
Joined: Wed Mar 14, 2007 2:54 am

Post by LH »

Norbert Schlenker wrote:Larry, I've had discussions/debates/arguments with you in a fair number of threads now about the risk in TIPS. You have repeatedly downplayed the risk, you have denied that there is much risk at all
What is the "risk in TIPS"?


Great thread,

LH
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Post by larryswedroe »

LH

There is INTERIM price risk as real rates can rise causing loss in market value. If you hold to maturity there is no risk.

One could also say there is opportunity cost risk, if real rates rise you might have been better off being in short term treasuries that are nominal bonds.

Now if course you can basically eliminate or minimize this risk by buying SHORT TERM TIPS instead of longer term ones.
Post Reply