In an informative post, #33695, Howard said that, these days, he’s not looking at anything priced under 50 and that the rule is keeping him out of trouble. I’d suggest that the cutoff-price could be bumped up to 60 as an easy way to avoid bonds that probably offer more grief than reward. It would be nice if there were also an upper price above which it isn’t prudent to buy. But there doesn’t seem to be one, as you can verify for yourself (just as you should also verify the utility of the lower cutoff price). But let’s focus for a bit on the upper-end of the price-range. If you grab the 500 highest-priced bonds, what you will notice is that a mere 11.8% fail to offer a real-rate of return even when discounted for a 5% inflation-rate. So, clearly, the fact of selling at a premium to par shouldn’t necessarily, in and of itself, be a buying deterrent. Do all of us hate paying a premium to par? Of course. But what we should really be hating to pay is a premium to value. That’s what is unwise. How is ‘value’ determined? Ah, grasshopper, now you’re beginning to ask a question worth answering and one that your investment plan (which you do have, right?) will tell you.
I’d suggest that the cutoff-price could be bumped up to 60 as an easy way to avoid bonds that probably offer more grief than reward. Charlie,That’s probably true, but from my experience, I’ve found that bonds in the 50’s have offered me some of my best risk/reward results. I’ve bought many bonds (with attractive CY’s) in the 50’s that have struggled, but so far managed to keep paying their coupons. I especially like high-coupon bonds selling in this price range. An example of what I’m talking about is Realogy 12-3/8’s of ’15. I bought this issue for 56.5 and my interest-adjusted cost is now down to 27.6. The bond is currently marked at 90 and still paying the high coupon. Overall, I’ve had very good results buying bonds in the 50’s and I’ve had very poor results from my purchases below 50.I’ve come to the conclusion that market-price is a very good indicator of a bonds chances of default...Better than agency ratings or YTM. In fact, in my junk world, I interpret “Price-Ratings” something like this :)< 50 likely to default50-59 hanging on by their finger nails60-69 high risk of default70-79 probably will survive80-89 conservative junk90-99 blue-chip junkAs far as a high-price limit is concerned, I do NOT buy above par. I stick with this rule only because it makes my buying decisions easier. It means I don’t have to check and worry about call-dates and whether YTM numbers are accurate. I'm sure to miss some good buys, but I’m always trying to simplify my scanning procedure with as few parameters as possible.The rules above were derived from my personal experiences and results in order to refine my scanning and buying process. They do not apply to everyone. I probably should consider them more guidelines than rules, but it’s easier to scan using strict rules.Howard
< 50 likely to default50-59 hanging on by their finger nails60-69 high risk of default70-79 probably will survive80-89 conservative junk90-99 blue-chip junk
I’ve found that bonds in the 50’s have offered me some of my best risk/reward results.Howard,Thanks for your detailed reply. If I had to name a price-range that offered my best rewards, that would be the sub-35's, like Internet's Cap's long-gone, somethings of sometime (I don't remember), that gave me 100% per year, or Xerox's 8's of '27 (recently called) that returned about the same, or even Hanson's 6.125 of '16 (which Scott and I got into mid-40). But that stuff's not available these days. So I was adjusting my prices to today's market. Last week, I would have said 50 was a good, lower cutoff. Below that, there's one or two issuers that a 'Speculative' buyer (in Graham's sense) could consider. But even an 'Enterprising' one (in Graham's sense) shouldn't be looking there and, for sure, not a 'Defensive' one, either. This week, I'd argue 60 is the lower cutoff, now that prices for Sears' bond have rebounded. (Even the bids this week are higher than last week's offers). But I'd argue that bond-buyers who self-identify themselves as 'Defensive' (those whose chief goal is "safety and freedom from worry") do themselves a disservice by excluding from their scan bonds trading at a premium to par. Yes, by definition, that group isn't low-priced. But does the group lack 'value'? That's where things get tricky. When do bonds (issued at par) come to trade at a premium to par? When interest-rates drop. That's most of what the premium can be attributed to. Interest-rates were higher at issue than presently. For sure, adverse calls have to be avoided. But all brokers warn customers of that by providing the YTW. So that's not a number that even needs to be calculated (though it should always be checked). Give me a couple of minutes (since this doesn't look to be a 'buying day'), and I'll run a frequency-distribution on the entry prices of my holdings. Again, thanks for your detailed reply. Charlie
Charlie,Thanks for taking the time to tabulate and present your data. I see this thread has now morphed into 3 threads that talk mostly about buying bonds above par. I will stick to our original discussion about default rates and investment returns in the various junk price brackets. Below I list my data which shows why I’ve come to my conclusion. I’m not trying to convince anyone, I’m just describing what works for me. As you can see my <50 bracket shows a 37% default rate (on a small sample). The 50-59 bracket shows the highest total return and the >90 bracket shows the lowest return. I’m a numbers guy, so this is why I said what I said, and do what I do. Price Trans Issues Defaults TotRet% Defaulted Issue< 50 15 7 2 8.7% Newpage, Norske (large loss)50-59 11 9 1 74.8% US Concrete(small loss)60-69 21 19 1 25.1% Dynegy (small profit)70-79 30 25 1 28.8 Dynegy (small loss)80-89 18 16 0 20.7 90-100 8 7 0 1.0 Totals 103* 76** 4*** 25.8***** includes all transactions (open, closed and defaulted) except bonds bought AFTER default (Tribune, Abitibi)** 7 issues were bought in split price brackets, so totals don’t balance with sum *** Dynegy was bought in 2 price brackets**** total return over 3 years of gradual accumulationI think the main reason we tend to differ in our buying approach is that bonds are your primary investment class, so you need to cover as many areas as possible in order to reinvest your free cash flow (in Lynch analogy, you need more rocks to turn over).For me, bonds are a sideline to my equity investing, so I focus on what works best. I’m just trying to add a bit of diversification and increase my income by leveraging my equity knowledge. It also gives me a market to shop before and after the stock market is open. On earnings announcement days, this can be very beneficial. So I’m quite happy to limit my focus to a part of the bond market that performs best and compliments my other investments.I agree with you that “everyone needs to trade their own game” (or something like that)Howard
Price Trans Issues Defaults TotRet% Defaulted Issue< 50 15 7 2 8.7% Newpage, Norske (large loss)50-59 11 9 1 74.8% US Concrete(small loss)60-69 21 19 1 25.1% Dynegy (small profit)70-79 30 25 1 28.8 Dynegy (small loss)80-89 18 16 0 20.7 90-100 8 7 0 1.0 Totals 103* 76** 4*** 25.8***** includes all transactions (open, closed and defaulted) except bonds bought AFTER default (Tribune, Abitibi)** 7 issues were bought in split price brackets, so totals don’t balance with sum *** Dynegy was bought in 2 price brackets**** total return over 3 years of gradual accumulation
Howard, Your reported 37% default-rate for issues bought below 50 is consistent with Moody's work on historical default-rates. So, no surprises there. Part of the reason I was exploring the lower limits (and the upper limits) of prices is that I'm trying to put together a set of guidelines (not rules) for my daughter that aren't based merely on what seems to be working now, but point toward risk-management principles that apply to any asset-class, anytime. (Risk is risk, no matter where it is found). But, yes, I don't do anything else but bonds, because they provide enough interesting puzzles and decent returns. US Concrete was a bond I dithered over and then backed away from, as was Dynegy. Like you, I did get slammed by Norske. But the loss (when the court settles with creditors) will be tolerable, because the position is a single, as is any of my speculative positions, and I try to keep exposures per position for such issuers under one-eighth of one percent of AUM. In other words, if I lose one or two, as has to be expected, the impact amounts to a rounding error. OTOH, the contribution they make toward total gains, when such positions do work, isn't insignificant, as you well know. Again, thanks for your detailed follow-up.Charlie
Charlie, I'm trying to put together a set of guidelines (not rules) for my daughter that aren't based merely on what seems to be working now, but point toward risk-management principles that apply to any asset-class, anytime. (Risk is risk, no matter where it is found). I would say that in most times (recession troughs excluded), bonds under 50 offer similar risks as stocks under $1.00. I don’t touch penny stocks (anymore). I try to keep exposures per position for such issuers under one-eighth of one percent of AUM You have a great advantage there (especially for junk). I’m now trying to get my exposure down below 3% of my bond holdings. 1-2% would be even better, but that’s still a ways off.Thanks for bringing up this subject. It motivated me to do a thorough analysis of my buy-prices vs default/return rates. I should’ve done this earlier. Howard
Howard, Instead of charging your bond-risks against just your bond portfolio, charge them against your whole portfolio. Risk is risk, no matter where it comes from, and if a position is too big, then it’s too big. Likewise, if it’s too small, then it’s too small, which is where most investor fail themselves by not taking on sufficient risk (for all of the well-studied reasons). But the net-result in either case is the same, a loss of nominal-principal due to price reversals, or a loss of purchasing-power due to inflation. Either one will kill a portfolio as surely as the other. One is just less obvious and, thus, less feared. Suggestion: find Moody’s white paper on defaults and construct a default table for yourself. What you’ll find is that the rates approximate a Fibonacci series. Normalize it and then down-grade each rating one notch (i.e., increase its default-rate by the next number in the Fib series). That becomes the rule of thumb benchmark against which you do your buying. In other words, in the same way a Black Jack player can count cards and adjust his bet size, a junk bond buyer can attempt to do the same. I say “attempt”, because there is never a statistically valid sample of any of the lower ratings for sale, nor will any but the largest of portfolios ever be able to hold a statistically-valid sample (whose size can be argued about, but, trust me, the responsible work in the field, not the MPT idiots, suggests the numbers are huge). The workaround, obviously, is to attempt to achieve statistical validity diachronically. In other words, retrospectively, the buying will appear to have been less than optimal. OTOH, when the black swans fly, no more than a tolerable number will land in your pond. Taleb hammers on this stuff constantly, and that's how risk has to be managed and portfolios built, to survive the damage you know will happen, but about which you cannot know "How soon?" or "How much?" Therefore, the possibility of total failure has to be accepted and effectively hedged. But then, you know that. You just don't think you do. Charlie
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |
BATS data provided in real-time. NYSE, NASDAQ and NYSEMKT data delayed 15 minutes.
Real-Time prices provided by BATS. M