Remember the old days, when retractible issues yielded less than perpetuals? That inspired one of my first articles, in which I examined the question of just how bad things had to get before the tortoise outpaced the hare.
And, Assiduous Readers will recall, BNA.PR.C often exhibits puzzling behavior.
These two concepts have now met, with (at the closing bid) BNA.PR.C priced below BAM.PR.N.
BNA.PR.C yields 9.37%, based on a bid of 16.83 and a hardMaturity 2019-1-10 at 25.00, while BAM.PR.N yields 7.13% based on a bid of 17.00 and a limitMaturity. The former issue is a split share based solely on BAM.A, with asset coverage of 3.3+:1 as of July 31, according to the company. As of March 31, 2008, there were 19,032,000 Units outstanding, each unit comprised of one capital share and one preferred share – each series of preferreds has a $25 liquidation value. The company owns 46,161,000 shares of BAM.A, so the BAM.A:Unit ratio is 2.4:1, so at today’s closing price of $31.52, asset coverage is a hair over 3.0:1. Give or take.
Now, one thing that makes the BNA issues intrinsically fascinating is the fact that they are so well covered by a relatively poor credit. I wish I could be as poor a credit as Brookfield, at Pfd-2(low) according to DBRS, but it’s still worse than the banks! This means that the credit rating of BNA is constrained by the rating of the BAM prefs – which makes all kinds of sense. As a rough approximation – for conceptual purposes only – we can say that BNA prefholders get hurt when the common is below $10, while the BAM prefholders get hurt when the common is below $0.
Anyway, the upshot is:
- the BNA prefs may be thought of as being junior to the BAM prefs, but
- the BNA prefs have a fixed maturity date, while the BAM perps are … perps
I invite criticism on this point, but I suggest that the two influences cancel out, leaving credit quality of the two issues approximately equal for investment purposes.
But the spread between these issues has varied all over the place:
The wideness of the current spread really is most peculiar. The fund has recently swapped its BAM.PR.N holding for BNA.PR.C … we shall see how well it works out!
Update, 2008-09-06: BAM.PR.N was recently affirmed at Pfd-2(low) by DBRS.
Since you invite criticism I shall make two points:
1. BNA and BAM prefs are quite different beasts, so there is no reason for Sr/Jr (which I don’t buy — see below) to “cancel out” perps vs an 11 year fixed maturity.
BNA prefs are short an 11-year put option on BAM common at $10. BAM prefs do not have this liability (the most directly comparable assumption would be a put option at $1 forever, subject to not being called by BAM). I consider this much worse than a Sr/Jr relationship to the same assets/business.
We can argue about how much the BNA put option is worth, but the market gives it a value (resulting in higher pref yield more than 80% of the time — looking at your YTW curves). BNA.PR.C and BNA.PR.B are trading at reasonably consistent levels recently (comparable yields within the huge b/a spreads).
In times of financial market stress and increased volatility, option time premiums go up — even if stock prices remain the same (BAM has actually lost some ground this year).
Although your YTW curve goes back only 10 months, it suggests to the amateur arbitrageur lurking here that there may be 6-10% relative pricing difference that can be recovered IF values go back to “normal”. Of course, the maximum arb profit could be as high as 29% if YTW’s converge to the same value (which has happened before, but is not the average and may not reflect b/a spread).
Another thing I see in the YTW data is that we can expect at least a couple of months for relative pricing to change to make an arb trade profitable.
This would not be a pure arb trade because long BNA.PR.C / short BAM.PR.N has residual credit risk from the short put. It may make a decent relative value trade (your objective), but the spread could be adversely affected by worsening yields and credit conditions — which a pure arb trade wouldn’t be. You’ll do fine if the ongoing pref rally continues.
Point 2. Aside from the short put explanation for the existence of a spread, I believe spreads should be wider when YTW are higher. This is because I believe spreads are better expressed as ratios of yields (not differences). Today’s YTW ratio (9.37/7.13 = 1.29 is not much out of line with earlier data of 1-1.25 in your figure. I suspect this is a minor factor compared with the put option value, but both go in the same direction at the same time.
Sure, I recognize that the parallels between BAM and BNA prefs are not exactly straight lines. But there is more analytical overlap between them than with most perpetual/retractible issues (excepting those pairs that are from precisely the same issuer, of course) and I consider them to be exposure to the same credit from a risk-control perspective (which I assert is a qualitatively true statement, regardless of the precise quantitative relationship) … so I can’t resist trying to figure out, even if only in conception, just how similar they are.
In times of financial market stress and increased volatility, option time premiums go up
Sure, but I assert that this should affect the perps more – at least normally. One thing that this crisis is introducing us to, among many other newfangled notions, is the concept of an inverted CDS curve (where the rate for a 1-year contract is higher than the rate for a 2-year contract). So “normally” doesn’t necessarily mean a lot in terms of the annual rate, but – if the strikes on the options were the same, which would be the case if the issuer was identical and the prefs parri passu – the total present value of the option should increase monotonically with time.
This would not be a pure arb trade…
Yeah. There’s a lot of former Wall Street Wonderboys who are currently trying to figure out just what makes a perfect hedge different from an imperfect one.
I believe spreads should be wider when YTW are higher
Maybe. It’s hard to justify in purely mathematical terms though. I think. Have you got a model that spits out this result?
1. Option time premiums go up in times of stress. You seem to think the perps are more affected, but the split share pref should be more affected as there is more time premium (in absolute dollars) in a $10 put than a $1 put. The expected life of a perp is not infinite because the put is conditional on the pref not being called beforehand. Even if the percent increase in a $1 perp put is higher than the % increase in a $10 split share pref in times of stress, the dollar value and dollar value increase in times of stress will be greater for the $10 put.
2. The spread I have studied most is the real interest rate — difference between Fed Funds, say, and inflation. When interest rates were high in the 1980s, the real interest rate was also high (e.g. 16% nominal – 8% inflation = 8% real). Later in the 1990s, inflation fell and we had 6% nominal – 3% inflation = 3% real. The real rate was proportional to the inflation rate. Although real rates fluctuate over the short and medium term as Fed policy is adjusted, they also experienced a cyclical rise in the 1970s and extended fall from 1980 – 2005. A better fit with higher R2 and lower residual error is obtained by fits with the real interest rate spread = a multiple of inflation rather than an additive amount over inflation. If you view inflation as a risk to nominal returns for which investors demand compensation, then it is not surprising that risk is proportional to inflation rate and therefore that real rates should be proportional to inflation — a spread that is multiplicative.
I haven’t performed fits of corporate bond spreads with inflation, but if we view a corporate bond yield as inflation + real Treasury yield + corporate spread, it is not inconceivable that the corporate spread could also be proportional to the real treasury yield, which is, in turn, generally proportional to inflation (over a business cycle; though not necessarily true in recessions when real yeilds are negative).
Certainly in the case of prefs, we have the pref bond equivalent yield = inflation + real gov’t yield + corporate bond spread + pref spread (for further subordination). Recent behaviour in this market makes it look like the pref spread is proportional to the corporate bond spread (so we get a double widening). We don’t see wide corporated bond spreads and narrow pref spreads over corporates, so the additive model seems to break down.
Finally, many things in life and investing are described by a log normal distribution, because negative values are physically impossible. A log normal distribution is a ratio scale, while a normal distribution describes additive components. Intuitively it makes sense that spreads could be log normally distributed, have variable volatility, etc. just like stock prices.
Most of the time you never see this effect, but when comparing periods of substantially different yields, the possibility that spreads are multiplicative rather than additive should be considered. I believe they are multiplicative, and that standard textbooks and practitioners don’t want to deal with the added complexity so stick to the simpler, traditional additive model.
[…] Asset coverage of 3.3+:1 as of July 31 according to the company. Now with a pre-tax bid-YTW of 9.34% based on a bid of 16.88 and a hardMaturity 2019-1-10 at 25.00. Compare with BNA.PR.A (6.17% to call 2009-10-31) and BNA.PR.B (8.88% to 2016-3-25). See also a comparison with BAM perps. […]
[…] is due to purchases of BNA.PR.C net of a small sale of WFS.PR.A. Assiduous readers will recall that I consider exposure to BNA to be equivalent, for credit risk control purposes, to exposure to BAM. These readers will not be surprised, therefore, to learn that the purchase of BNA.PR.C was not […]
[…] Compare with BNA.PR.A (7.02% to 2010-9-30) and BNA.PR.B (8.92% to 2016-3-25). Note that, given 2.4 shares of BAM.A per BNA preferred and a price of 28.49 on BAM.A (see above), asset coverage is now […]
[…] BAM! Scare stories about Brookfield crushed their prefs – and their dependent structures: Commercial real estate could be next on the economic hit list because of Wall Street’s […]
[…] Compare with BNA.PR.A (7.26% to 2010-9-30) and BNA.PR.B (9.14% to 2016-3-25). Note that, given 2.4 shares of BAM.A per BNA preferred and a price of 27.51 on BAM.A (down 4.35% from yesterday), asset coverage is now 2.6+:1. […]
[…] 3.2+:1 as of August 31 according to the company. Coverage now of 2.6+:1 based on BAM.A at 27.40 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 8.10% based on a bid of 24.26 and a hardMaturity 2010-9-30 at 25.00. […]
[…] 3.2+:1 as of August 31 according to the company. Coverage now of 2.7+:1 based on BAM.A at 28.38 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 10.25% based on a bid of 15.90 and a hardMaturity 2019-1-10 at […]
[…] of August 31 according to the company. Coverage now of just under 2.6:1 based on BAM.A at 26.98 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 12.36% (!) based on a bid of 22.50 and a hardMaturity 2010-9-30 at […]
[…] of August 31 according to the company. Coverage now of just under 2.8:1 based on BAM.A at 28.97 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 11.68% (!) based on a bid of 14.36 and a hardMaturity 2019-1-10 at […]
[…] 3.2+:1 as of August 31 according to the company. Coverage now of 2.4+:1 based on BAM.A at 25.27 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 12.13% (!) based on a bid of 16.53 and a hardMaturity 2016-3-25 at […]
[…] 3.2+:1 as of August 31 according to the company. Coverage now of 2.2+:1 based on BAM.A at 23.37 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 13.02% based on a bid of 13.11 and a hardMaturity 2019-1-10 at […]
[…] to the company. Coverage now of just under 2.1:1 based on BAM.A at 21.75 (also up on the day) and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 9.68% based on a bid of 19.01 and a hardMaturity 2016-3-25 at 25.00. […]
[…] 3.2+:1 as of August 31 according to the company. Coverage now of 2.4+:1 based on BAM.A at 25.16 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 11.66% based on a bid of 17.02 and a hardMaturity 2016-3-25 at […]
[…] as of September 30 according to the company. Coverage now of 2.4+:1 based on BAM.A at 25.44 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 17.55% based on a bid of 20.75 and a hardMaturity 2010-9-30 at […]
[…] as of September 30 according to the company. Coverage now of 2.1-:1 based on BAM.A at 21.72 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 12.79% based on a bid of 13.39 and a hardMaturity 2019-1-10 at […]
[…] as of September 30 according to the company. Coverage now of 2.O-:1 based on BAM.A at 20.50 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 11.18% based on a bid of 17.55 and a hardMaturity 2016-3-25 at […]
[…] as of September 30 according to the company. Coverage now of 2.0+:1 based on BAM.A at 21.10 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 15.42% based on a bid of 21.60 and a hardMaturity 2010-9-30 at […]
[…] as of September 30 according to the company. Coverage now of 2.2-:1 based on BAM.A at 22.51 and 2.4 BAM.A held per preferred. Now with a pre-tax bid-YTW of 17.24% based on a bid of 21.00 and a hardMaturity 2010-9-30 at […]
[…] Compare with BNA.PR.A (7.02% to 2010-9-30) and BNA.PR.B (8.92% to 2016-3-25). Note that, given 2.4 shares of BAM.A per BNA preferred and a price of 28.49 on BAM.A (see above), asset coverage is now […]