GDP, M&A, EBITDA, P/E, NASDAQ, Econo-thread Part 13

[DanS]

Reading the just-released Economic Report of the President, there are some interesting arguments. For instance...

1) Stronger correllation than normal '95-'00 between equity prices and corporate investment;

2) Eliminating capital income taxes should address a distortion on the propensity to invest v. consume;

3) There is not a housing bubble, due in part to increased demand from immigrants, a reduction in the default risk premium. See pg. 43;

4) The accumulated current account deficit is far from a point where we need to be concerned about it. See pg. 59;

http://w3.access.gpo.gov/usbudget/fy2004/pdf/2003_erp.pdf

[Dinner]

There probably isn't a nation wide housing bubble but home prices here in SanDiego have been appriciating at around 15% per year for the last three years and 10%-12% for the 10 year period before that. Population went up but so did the amount of housing stock so it seems there is a slight bubble forming at least in southern California.

[HershOstropoler]

This could be interesting, just don't have the time to read it at the moment:

http://www.ofheo.gov/docs/reports/sysrisk.pdf

"SYSTEMIC RISK: FANNIE MAE, FREDDIE MAC
AND THE ROLE OF OFHEO"

And it seems the boss of ofheo was fired over this ?

[Sten Sture]

On corp bonds...

The return numbers are not directly comparable because of the differences in maturity structure that we in place during those years; those same differences are in place today. Differences in bond duration on the "long" end of the curve can be substantial, and especially at low interest rates, a 10yr bond can have radially different return characteristics than a 30yr bond. Also differences in prepayment penalties can impact returns. For a substantial portion of those years a large number of long corporate bonds had call provisions which were different from the government provisions. A bond that is 30yrs to maturity but callable in full after ten years will perform much different than either a 10yr bond or a 30yr bond.

The way we deal with these return differences now is to look at each individual corporate bond and compare its return to a similarly structured treasury bond to derive "Excess Return". The math for evaluating call options in bonds wasn't really in play until Black/Scholes and Merton developed their model in the late 70s, and Excess Returns have only been actively used in the past 5 years, so to my knowledge we haven't had anyone go back and run historical return information using modern methodology.

The credit composition of corporate bond indecies has changed dramatically over time as well. For instance just 10 years ago the percentage of traded corporate bonds outstanding relative to Treasury bonds was about 35%, now it is about 125%. And the BBB component has grown the fastest of the corporate credit buckets.

So now we look at corporate bonds using their (option adjusted) spread to similar treasuries, and that weighted spread is about 175 basis points, or a 3.10% on Treasuries versus a 4.85% for all corporate bonds.

So the spread needs to compensate investors for assuming default risk, and the relative level of nominal rates is important as well. With Treasuries at a 3.10, +175 is a big percentage improvement, however if Tsy was 15.00% then 16.75% might not seem like such a great deal.

Historically, even with a disasterous 2002, the default rate on investment grade bonds (BBB or higher) has been almost negligible. In fact for BB rated junk bonds it has barely been material. So returns that are very close to Treasury returns are not that difficult to accept.

[DrSpike]

Originally posted by Sten Sture
The math for evaluating call options in bonds wasn't really in play until Black/Scholes and Merton developed their model in the late 70s, and Excess Returns have only been actively used in the past 5 years, so to my knowledge we haven't had anyone go back and run historical return information using modern methodology.

Yeah....there is bagloads......I don't have my stuff here with me now but I know Lo and Mackinlay (huge empirical finance geezers) have quite a few articles applying modern math to timeless problems.

[DanS]

Thanks for that nice explanation, Sten. I'll have some follow up questions. But for now, here's a post for Banned GP:

Questions:

1. Does long term mean the same thing for the different groups?

2. What grade of corp bonds (AAA or junk?)

3. What grade of government bonds (US treasuries or local governments or foreign debt).

Discussion:

It would be interesting to look at actual default rates and see if that difference is reasonable or not. I haven't seen this done for the exact comparison you have here. But I have seen it done for different ratings of corp bonds. I.e. BBB vers BB, etc. Broadly speaking the results showed that the different ratings for the bonds and different yeilds correlated reasonably well with default rates. You can take the numbers and do risk-adjusted rates of return. And the answers seemed to make sense. Not completely clean.* But made sense.

*I think there were more AAA bankruptcies than AA. But otherwise, the rankings seemed to match well with the incidence of bankruptcy. and the different yeilds matched as well.

---------------------------

Didn't we already cover all of this?

[Sten Sture]

does this work...


Nope.

How about I just type it in.

Since 1983, annual default rates by rating subcatagory

Ba1 0.70% (highest junk rating)
Ba2 0.60%
Ba3 2.70%
B1 3.80%
B2 6.70%
B3 13.20%
Cs 19.40% (Caa - C)


GP - long term generically means maturing in 10yrs or more, so it includes about 15% of the broad domestic bond market ex-money market stuff.

Generally Corporate bond indecies just include investment grade stuff Baa/BBB and higher. Though recently there has been a lot of all-in type of numbers generated as the high yield (junk) market has matured.

Government bonds refers to both direct US Treasury and indirect US sponsored Agency note and bonds, but not state issued, or agency mortgage backed issued stuff.

[Sten Sture]

Originally posted by DrSpike

Yeah....there is bagloads

Thanks, I should have assumed as much. I've been brainwashed into thinking that no one cares about bonds.

[DrSpike]

Hehe right about now I wouldn't have thought that would be a problem.

[Adam Smith]

DanS:

Do you have a link for the Financial Times article on valuing Iraqi oil reserves? I did not find it on a search here or elsewhere, and I think their analysis may have missed something.

[DanS]

Here's the link, but apparently after a certain amount of time, only subscribers are allowed to access it.

http://news.ft.com/servlet/ContentServer?pagename=FT.com/StoryFT/FullStory&c=StoryFT&cid=1042491848625&p=1012571727092

[Adam Smith]

When he calculated the effects of lower prices, do you recall whether that applied to Iraqi oil, imported oil, or all oil? Should have been applied to all oil.

[DanS]

I don't recall.

Edit: I now have the article in front of me. I'll try to PDF it later on today. In short, the author makes a narrower argument. He assumes that Iraq's production will not increase or decrease substantially from current levels.

10% discount rate on 2.5 million barrels/day production. $25/barrel.

[DanS]

Let's see if this works.

[Saras]

Bump!

And a link about the Roland Stenish crowd

http://www.foulds2000.freeserve.co.uk/economists.htm

How do you guys like the EUR/USD rate?