Bogleheads

[larryswedroe]

This one covers a longer period, so we now have data all the way back to 1960.
Bradford Cornell and Kevin Green, “The Investment Performance of Low-grade Bond Funds,” Journal of Finance, March 1991.

The study covered the thirty-year period 1960–89. The authors concluded: “On a risk-adjusted basis, the returns on low-grade bond funds are not statistically different from the return on high-grade bond funds.”

[larryswedroe]

the combined studies now give us 48 years of data on junk bonds.
all with the same basic results--they don't call it junk for nothing

[drjdpowell]

[larryswedroe]

Sorry Dr Powell I forgot to include the reason why it supports the other studies. This study includes the period when almost all of it was original issues (which Fridson showed did deliver but the FAs did not). We only got original issues in late 80s basically.
And of course for individuals you have the location problem for those with choices.
BTW-Milken made his name on the original issues-that was where his research was. Before him there was basically little/no original issue junk market.
Larry

[larryswedroe]

Just to confirm the facts I went and checked and the first year of Original Issue high yield debt exceeded $1b was 1977.

[mephistophles]

Aside from the studies I can personally attest to the risk of high yield bonds. We had some in my father's portfolio about ten years ago. They were bought for yield to help pay for in home 24/7 health aides instead of putting parents into a nursing home. The plan worked for awhile and I believe the yields were above 10%. Then all hell broke loose when the tech bubble burst and the value of these high yield bonds declined between 30 and 40%. This money was never recaptured when the markets improved. And now for the double whammy, these bonds received a stepped-down basis at death so a tax loss could not be taken.

Their is no free lunch. The high yields of junk bonds come at the price of huge risk and potential loss. I cannot see any situation that justifies ownership of junk.

Regards,
ole meph

[HerbertSitz]

[kenschmidt]

[alec]

[HerbertSitz]

[alec]

Herb,

Sorry for any confusion. On any diversifiation benefits to adding HY bonds to a portfolio, I was referring to this post by Larry.

[larryswedroe]

Ken
Yes that is just what it said
And prior to late 70s there was almost no original issue market for junk. Milken changed that. It wasn't until 1977 when it finally crossed the $1b in a year. So most of the data for the period of that study was the returns of the fallen angels. But now OIs account for the vast majority of the market (about 70% if my memory serves)

[larryswedroe]

Herbert
I did not say that one should disregard if risk adjusted returns are the same.
The returns of the FAs seem to be the same on risk adjusted basis but not the OIs which dominate the market now. They have much more risk of calls and clawbacks. That seems to lead to reduction in returns.

Now having said that even if they had the same risk adjusted returns you have the location problem for individuals which institutions do not have.

And of course you can buy Treasuries direct and you have to use a fund to access junk (to get diversification you need) and funds not only have operating expenses but trading costs as well--both of which can be avoided.

And then you have the issue of location problem institutions don't have--

And then you have the problem of the correlations rising at the wrong time.

So IMO much better to simply avoid owning junk bonds. You can build superior portfolios without them, and you avoid all the negatives.

Now perhaps if you could find a fund that was passive and well diversified and low cost and only bought FAs and you had to hold all the assets in tax advantaged accounts anyway, then you might consider adding some hi-yield. But even then you are adding a very tiny amount of unique risk, risk not explained by the tax differential at the state level, expected losses and the three factor model. Why bother?

[larryswedroe]

Herbert
The mistake people make when looking at the correlation of junk to stocks and bonds is that you should see low correlation because it is hybrid.
Split the two parts correctly and you will get high correlation. There is almost no unique risk here you cannot accomplish by using stocks and Treasuries or investment grade bonds, without the negatives of junk

[HerbertSitz]

[larryswedroe]

Herbert
The fact is that junk has had very little unique risk.That is the conclusion reached by the academic papers on the subject.

The paper I cited did not look at that issue.

But that paper IMO confirms the findings of the Fridson study as it covered period where most of the outstandings were fallen angels--before the 80s almost all of the market was fallen angels.
The studies beginning in the late 80s and 90s show poor risk reward, and even negative in one study, because they contain mostly original issues (which now make up about 70% of the market)