Seeking Alpha
Profile| Send Message|
( followers)  

There are several issues with corporate junk bonds that many investors are unaware of or do not sufficiently understand. This is the first of two articles describing these issues, providing a mathematical projection for how junk bonds are likely to perform, and showing how junk bonds match up against other fixed-income investments with similar years-to-maturity. Corporate junk bond funds and individual corporate junk bonds are covered.

The Higher Interest Rates on Junk Bonds Are Very Deceiving

Some junk bond investors think the higher interest rate they receive on junk bonds is well reflective of the actual return they can expect to receive. If you are unknowledgeable enough to think this is the case, you should not be investing in junk bonds. Please allow me to explain how junk bonds actually work.

You receive a much higher interest rate on junk (non-investment-grade) bonds than you do on investment-grade bonds. This is because junk bonds default at a much higher rate than investment grade bonds. In theory, junk bonds default enough more that it largely negates the additional interest you are paid for holding junk, rather than investment-grade, bonds. In theory, you end up with a small amount more in total return than you would have with investment-grade bonds because you were willing to take on additional risk. This is the theory. Practice does not always follow theory.

Recently, a lot of people are chasing higher yields because interest rates, in general, are very low. This can lead to an environment where junk bonds, in relation to investment-grade bonds, are overvalued. In an environment like the current one, junk bonds may be a less profitable investment, not a more profitable one.

Based on data from page 183 of this Credit Suisse report, if you begin from the middle of 1985, average annual default losses on corporate junk bonds through the end of 2011 were 2.67%. The best year was 2007 at 0.22%. The worst year was 2002 at 11.51%. This is default losses, so it takes into account that some money is usually recovered (about 42% on average for corporate junk bonds) when bonds default. Beginning from 1985 is a good thing to do because this is, about at least, the beginning of the modern corporate junk bond era. Prior to 1985, there were no, or very few, new-issue corporate junk bonds. Prior to 1985, the corporate junk bond market was much smaller and consisted of investment-grade bonds that had fallen from grace (i.e., "fallen angels"). Data from prior to 1985 is not as relevant to the modern junk bond market.

The 2.67% figure is consistent with the healthy amount of other corporate-bond-default-losses data I have reviewed. This is far from an end-all-discussion figure though. For instance, as you can see in looking at the bottom of page 12 in this (pdf) Moody's report, the credit qualities (i.e., Ba, B, Caa-C) associated with corporate junk bonds were far stronger in 1985 than they are today. From this perspective, we can see that 2.67% is too low, since Caa-C bonds default much more, and with a slightly lower recovery rate, than Ba bonds. On the other hand, iShares iBoxx $ High-Yield Corporate Bond ETF (NYSEARCA:HYG) and SPDR Barclays Capital High-Yield Bond ETF (NYSEARCA:JNK), which I am about to evaluate, hold only about half of the percentage of Caa-C (Moody's) or CCC-C (S&P) bonds as the corporate junk bond market in general; so 2.67% seems about fair when applied to HYG and JNK.

There are some other important things to note about the 2.67% figure. For one, there is no guarantee defaults were not undercounted somehow. For another, we are assuming there is no ratings scam, as there was with MBSs a little while back. If there is such a scam, the impact will probably be much greater in the junk world versus the investment-grade world. For yet another, the near future is projected to be better than 2.67%; but you are extremely unlikely to get advance notice prior to the near future not looking better. In all likelihood, junk bond prices will have already reacted prior to your being notified. The worse year for corporate junk bond default losses from 1985 until now was 11.51%, but it can be far worse than this. Finally, past performance does not necessarily reflect future performance; so 2.67% may be significantly high or low as an average going forward. The risk is, of course, that it is low.

Below is a spreadsheet I used to calculate the Adjusted True Yields of six different investments. Adjusted True Yield is True (actual) Yield adjusted for personal expenses, fund expenses, default/downgrade losses, and taxes. For an explanation of True Yield, please see my article entitled 'The True Yield of Your Bond Investments". For a better understanding of Adjusted True Yield, please see my article entitled "Determining the Best Bond Funds: True Future Total Return". (Adjusted True Yield is the same as True Future Total Return minus the capital-gain-or-loss-due-to-interest-rate-changes aspect.) The six investments are:

  • HYG
  • JNK
  • Vanguard Intermediate-Term Corporate Bond ETF (NASDAQ:VCIT)
  • SPDR Barclays Capital Intermediate-Term Corporate Bond ETF (NYSEARCA:ITR)
  • A combination of VCIT and ITR that provides a good comparison to HYG because it has the same average years-to-maturity.
  • A combination of VCIT and ITR that provides a good comparison to JNK because it has the same average years-to-maturity.
ETFYield Type (YTM, YTW, or OAY)YieldSharesPriceAmountPersonal ExpensesPersonal Expenses %Holding Period (Years)Annualized Personal Expenses %Fund Expense RatioAnnual Default / Downgrade LossesLast DistributionDistributions Per YearInterest RateInterest Tax RateAnnual Interest Tax LossCurrent Price vs. Par Value Differences Gain or LossCapital Gain or Loss Tax RateAnnual Capital Gain or Loss Tax EffectAnnual Adjusted True YieldAs of Date
HYGYTW6.27%1081$92.49$100,000$16.73-0.02%6.650.00%-0.50%-2.67%$0.49126.29%25.00%-1.57%-0.02%15.00%0.48%2.01%10/26/12
JNKYTW6.28%2,482$40.29$100,000$16.73-0.02%6.780.00%-0.40%-2.67%$0.22126.61%25.00%-1.65%-0.33%15.00%0.51%2.07%10/26/12
VCITYTM2.80%1,129$88.56$100,000$16.73-0.02%7.400.00%-0.14%-0.16%$0.24123.21%25.00%-0.80%-0.41%15.00%0.11%1.80%10/26/12
ITRYTM2.01%2,857$35.00$100,000$16.73-0.02%5.170.00%-0.15%-0.13%$0.09123.03%25.00%-0.76%-1.02%15.00%0.20%1.17%10/26/12
VCIT/ITR HYG ComparisonYTM2.54%1,416$70.60$100,000$16.73-0.02%6.650.00%-0.14%-0.15%$0.19123.18%25.00%-0.80%-0.65%15.00%0.14%1.59%10/26/12
VCIT/ITR JNK ComparisonYTM2.58%1,357$73.67$100,000$16.73-0.02%6.780.00%-0.14%-0.15%$0.20123.19%25.00%-0.80%-0.61%15.00%0.14%1.62%10/26/12

Notes:

(1) Assumes there are no applicable state or local taxes.

(2) Assumes all capital gains or losses are long-term.

(3) HYG's yield to worst (YTW) was calculated using data available on the BlackRock iShares Holdings webpage for HYG. This YTW was adjusted slightly to account for the minor price difference between HYG on 10/26/12 versus 10/24/12, the date for the Holdings webpage data.

(4) HYG's Holding Period is its weighted average maturity in years, as calculated using data available on the BlackRock iShares Holdings webpage. Its years-to-worst should be a little less.

(5) JNK's YTW is based on the underlying index's 6.30% YTW as of 9/30/12. The 6.30% figure was adjusted slightly to account for the minor price difference between JNK on 10/26/12 versus 9/30/12 and the minor difference in yield to maturity (YTM) for JNK and its underlying index as of 10/24/12.

(6) JNK's Holding Period is its (weighted) Average Maturity in Years. Its years-to-worst should be a little less.

(7) JNK and ITR have variable fees for cash creation/redemption that I did not see explained in the JNK fact sheet, prospectus, or annual report. The potential impact of these fees is not reflected in this spreadsheet.

(8) I used YTM versus YTW for VCIT and ITR because of its availability and the fact that VCIT and ITR seem to hold no or almost no bonds that can be prepaid (e.g., called).

(9) VCIT's Holding Period is its average maturity in years.

(10) ITR's Holding Period is its (weighted) Average Maturity in Years.

(11) I calculated the Annual Default/Downgrade Losses for VCIT and ITR based on data from page 38 of a Moody's report entitled "Annual Default Study: Corporate Default and Recovery Rates, 1920-2011". To get this report, you need to register at www.moodys.com and, then, search for the report on the Moody's website. These Annual Default/Downgrade Losses are based on investment-grade versus junk default rates after 7.4 years-to-maturity for VCIT and 5.17 years-to-maturity for ITR. The figures calculated are probably a little high because the data used is from 1983-2011 versus midway through 1985 through 2011 and there are no 7.4 years figures from 2004 forward or 5.17 years figures from 2006 forward. Annual default losses should be far less than the figures calculated; but the figures also accommodate the fact that some investment-grade bonds will lose their investment grade and need to be sold at, what will probably be, a lower price.

(12) The formula for the Annual Capital Gain or Loss Tax Effect field is an improvement upon the formula for this field as described in my article entitled "Determining the Best Bond Funds: True Future Total Return". This formula also treats Annualized Personal Expenses %, Fund Expense Ratio, and Annual Default/Downgrade Losses as capital losses. (It was important to make this improvement now because the Annual Default/Downgrade Losses and Fund Expenses Ratios for HYG and JNK are relatively rather high.)

The above spreadsheet is set for the 25% U.S. income tax bracket, but I also ran the results for the other tax brackets. The results were as follows.

ETF

True Yield

Holding Period (Years)

Adjusted True Yields

U.S. Income Tax Bracket

0%

10%

15%

25%

28%

33%

35%

HYG

6.27%

6.65

3.10%

2.47%

2.16%

2.01%

1.82%

1.50%

1.38%

JNK

6.28%

6.78

3.21%

2.55%

2.22%

2.07%

1.87%

1.54%

1.40%

VCIT/ITR HYG Comparison

2.54%

6.65

2.24%

1.92%

1.76%

1.59%

1.49%

1.33%

1.27%

VCIT/ITR JNK Comparison

2.58%

6.78

2.29%

1.97%

1.81%

1.62%

1.53%

1.37%

1.30%

Simplified, the results look like this.

ETF

Adjusted True Yield Differences

U.S. Income Tax Bracket

0%

10%

15%

25%

28%

33%

35%

HYG vs. Invest.-Grade Comparison

0.86%

0.55%

0.39%

0.42%

0.33%

0.17%

0.11%

JNK vs. Invest.-Grade Comparison

0.92%

0.58%

0.41%

0.44%

0.34%

0.17%

0.10%

There are a couple of additional factors that should be considered though. For one, I only had years-to-maturity figures for HYG and JNK. I did not have years-to-worst figures. Some of the HYG and JNK bonds can/will be called or the like (i.e., paid off early) by the borrower, so the lending periods for HYG and JNK are not quite as long as is reflected in the data I used. To make the comparison fairer, I conducted the same exercise using YTM, versus YTW, for HYG and JNK. The other additional factor that should be considered is HYG's portfolio of bonds has better credit quality than JNK's portfolio of bonds. This being the case, I equalized the YTMs for HYG and JNK so the only difference is that which is caused by the difference in years-to-maturity. The final results are as follows.

ETF

Adjusted True Yield Differences

U.S. Income Tax Bracket

0%

10%

15%

25%

28%

33%

35%

HYG vs. Invest.-Grade Comparison

1.04%

0.73%

0.57%

0.57%

0.48%

0.32%

0.26%

JNK vs. Invest.-Grade Comparison

1.21%

0.87%

0.70%

0.69%

0.58%

0.41%

0.34%

So, in the end, my statistical analysis indicates there is, currently, a small reward for taking on the additional risk associated with owning HYG or JNK versus an investment-grade junk bond ETF with a like years-to-maturity. This reward is greater in the lower tax brackets because the relatively high interest rates HYG and JNK pay have higher tax consequences.

If I had done an evaluation of individual corporate junk bonds to be held to maturity versus individual corporate investment-grade bonds to be held to maturity, the results would have come out a little more in favor of the junk bonds. The higher trading costs associated with the junk bonds would have hurt them some, but the absence of the higher junk bond ETF fund expense ratios would have more than offset this. (I say this assuming the bonds are held for the holding periods listed previously.)

Do not jump for joy because the junk bonds won or vice versa yet. Please wait for the next article. There is more to this story than just this one issue.

Source: The Issues With Corporate Junk Bonds