I want to discuss the Junk Bond idea
again, as I am trying hard to kill this investment idea. I will start first with some commentary by Bill Miller from his recent communication with his investors regarding credit spreads:
One of the most important bullish signs has been little remarked upon. The monetary base, which consists of cash in circulation or in banks, had been decelerating during the entire [crises] time the Fed had supposedly been injecting liquidity into the system since last August. Thus, the amount of what economist Milton Friedman called high-powered money to stimulate the economy was decelerating, and so was the economy as the crisis continued. Now, though, the base is exploding as the Fed has finally turned up the liquidity pump. Since just after the GSE [Fannie Mae and Freddie Mac] seizure, the Fed began expanding the monetary base, so far by over $300 billion, an unprecedented increase. It takes a while for all that liquidity to find its way into the system, but find it, it should, and the transmission mechanism is typically through capital markets first. As it does so, the odds are very good credit spreads will begin to decline sharply and equity prices rise.
Right or wrong in his view, I always view macro economic forecasting as a roll of dice at best. I try hard to never merit any investment idea based on economic forecasts; the idea has to stand on its own merit. But I though it was a good mention.
Unlike equities, fixed income investing holds finite number of risks. These risks predominately are:
- Liquidity Risk (ability to find bids on your bonds): If I hold individual names this would be a big one.
- Interest Rate Risk: With fixed income, higher rates usually kills returns for bonds. However, junk bonds typically behave like stocks coming out of recessions.
- Refinancing Risk: Ability for businesses to refinance their borrowings. This one will be acute since there is almost no credit circulating for good companies let alone junk status ones. And in an era of vanishing leverage junk companies will get no financing to roll their debt so defaults will be higher than previous crises.
- Credit Risk: Generally this is a big one. And it is more paramount in this period than historically. We are coming out one of the largest credit bubbles in history, so undoubtedly writing standards for junk bonds in recent years have deteriorated from past years. That means that any historical averages for default and recovery rates will be almost irrelevant. Historical recovery rates averaged around 40%, I think that rate should not be remotely expected in this crises. Furthermore, the highest default rates on record were 15% recorded during the great depression era. Another point to consider is that the default rate is a blended rate across all issues from BBB rated to CCC and non rated issues. So with Junk bonds you have to assume higher rates. I will assume it will be 17.5-20%. The most recent years high default rates were recorded in 1990-91 as they topped the 10.14% and 10.27% respectively.
However the Case for Junk Bonds is as follows:
History tells us that U.S. Treasury securities generally outperformed spread sectors when financial markets were descending into crisis and investors were seeking a flight to safety. But history also tells us that these periods of out-performance were generally short-lived and difficult to time. And as markets emerge from these periods, and valuations begin to normalize, the yield differential of spread product over U.S. Treasuries could offer compelling value.
A look at the graph, (
Source), we can see a clear relationship between Junk-Treasury spread and future incremental returns over treasuries. In all years where spread moved over the average spread of junk-treasuries, investors were rewarded with higher returns in junk compared to treasuries. For example, when the spread reached 10.5% in 1990, the forward year junk return was 26% over treasuries.
As the crisis atmosphere recedes and the economy improves, logic says that it will happen sooner or later, default risks are likely to recede. In this environment, spreads tend to normalize, which can lead to better relative performance from high yield bonds.
The question is, are the wide spreads pricing enough compensation in the form of higher yields to assume this risk of high defaults and low recoveries?
Breakeven Yield AnalysisI use simple breakeven analysis, demonstrated by Altman and J. Bencivenga (1995), to show the breakeven yield (BEY) that must be promised in order to compensate for expected default rates and recovery rates.
The end result is a comparison between actual yields at a point in time and the breakeven yield. This shows the yield premium (if any), at any point in time (i.e., the amount to compensate investors for risks, other than expected default risk, involved, e.g., liquidity, unexpected losses, flights to quality, etc).
The results are as follows:
Recovery rates
Default 20.00% 30.00% 40.00%10.00% 12.42% 11.31% 10.20%
15.00% 18.11% 16.35% 14.58%
17.50% 21.22% 19.10% 16.97%
20.00% 24.52% 22.02% 19.52%
If we assume Armageddon with default rates of 17.5% and 20% recovery rates, I need to be promised a yield to maturity of 21% to breakeven with treasuries. Current yields to maturity of Junk bonds (JNK) is 21%, which is the breakeven point for my base case scenario. Anything better, junk bonds will deliver handsomely, and if these assumptions are too optimistic then it is not worthwhile investment. It seems that the promised yield for junk is at the edge, leaving no margin of safety for errors in my assumption and calculations.
IRR Cash Flow AnalysisThe problem with BEY analysis is that it does not take account for holding periods and cash flows. So I will conduct an IRR cash flow analysis, it is the same that I made in my earlier post. Again the assumption, and they are unrealistically pessimistic:
- 5 year holding period
- 17% default rate throughout the holding period of 5 years, never will happen. You might get two years of high defaults.
- 20% recovery rate throughout the holding period.
- I assume I am buying Junk at 60 cents to the dollar.
- Average coupon rate of either JNK or HYG is 8.7%
- defaulted bonds are assumed non workable and will be removed from books, another unrealistic scenario, some of the defaulted bonds will be put to special service firms that will rework payments.
The results are as follows (all figures are with $1):
Year: 1 2 3 4 5 Par value $ 1.00 $ 0.83 $ 0.69 $ 0.57 $ 0.47
Coupon pmts $ 0.07 $ 0.06 $ 0.05 $ 0.04 $ 0.03
Defaults $ 0.17 $ 0.14 $ 0.12 $ 0.10 $ 0.08
Recovery $ 0.03 $ 0.03 $ 0.02 $ 0.02 $ 0.02
Net cash Flow $ 0.11 $ 0.09 $ 0.07 $ 0.06 $ 0.52 (return of principal:$.47)
My IRR is 9.57%, that return can be achieved by investing in more secured debt instruments.
Conclusion:
I have backed away from my earlier conclusion in my previous post and I am less enthusiastic about junk bonds. The investment appeal rests with the assumption of defaults and recoveries. If you think that defaults and recoveries are going to be better than priced by the market then you have a good investment. If not then High Yield debt is appropriately priced by the market.
My problem with the junk bonds is that I am not getting a bargain. There is no margin of safety for potential errors in my analysis. To do this deal I need better promised yield from junk or better debt seniority, which will entail better recoveries and margin of safety.
