> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
> Behalf Of John Williams
> Sent: Friday, September 26, 2008 10:29 AM
> To: Killer Bs (David Brin et al) Discussion
> Subject: Re: Do low-probability catastrophic risks justify the bailout?
> 
> Dan M <[EMAIL PROTECTED]>
> 
> >> But, you talk as though A2 represents risky companies.
> 
> They are riskier than AA. That is the point.
> 
> > This is for a market that fully expected a bail-out.
> 
> No, it was not full, only about 75%.
> 
> > For example, do you really think the odds of GE defaulting on 1 day
> paper is
> > significantly higher than it was 2 weeks ago?
> 
> I have no idea. The market evidently decided that, distributing the
> available funds among available comapnies, that GE was less desirable 
> than some others.

OK, I can understand if that happened; but that's not what happened.  The AA
rated paper didn't have their interest rates drop (which is what happens
when more people want to buy those bonds); they held steady.  
The A paper price rose precipitously.  The one bond that went up (lower
yield) was the Fed note.)  This indicates that people are not switching
their desired targets for purchase from one company to another, it's an
indication that money is being pulled out of the commercial paper market.

> Do you really think you know what the interest rate should be for lending
> to companies?

No.  I'm not arguing that the markets are inefficient.  What I am saying is
that in a system where everyone acts in a purely rational market, changes to
Honda's, GE's, Ford's, Catipiller's 1-30 day interest rates should be
related to either a general rise in interest rates or to things that
directly involve these companies.

Now, it is very rational for all banks to be somewhat suspect, since in the
present atmosphere, we don't know (perhaps no-one knows) their exposure.
But, companies that are not involved with things like credit default swaps,
rationally, should be minimally effected by problems with credit default
swaps.

We know historical cases were irrational bubbles and panics existed.  For
example, all the folks who told me that investing in companies because
either they were making a profit or had a reasonable plan to make a profit
in the future was the old paradigm: companies that would never make a profit
were good investments because the new paradigm said so.

That's irrational, but lotsa folks bought into that before the bubble burst.

Now, since the market is efficient, I cannot predict when the bubble would
burst, but I know investing in a company with no business plan to make money
ever ever ever is not a good idea.  To do so, I'd have to predict the
behavior of others.

And, with all of its problems, I agree with you that a market based economy
is a lot better than centrally planned economy....a whole lot better.

I think we agree that panics exist. Where I think we differ is whether
markets can stop panics on their own, and whether we can see signs of a
panic...and whether interventions by folks with big pockets can stop such a
panic.

When GE pays a 1.5% premium on a 1 day note, investors are saying they are
seeing a much higher risk for GE not being able to pay the note the next
day.  I would argue that it's irrational to even give a 1% probability to
the proposition that GE would be unable to pay that note the next day.  I
think the jump in the spread to 400 basis points from about 40 basis points
in a weak between AA and A short term paper is a measure of panic.

But, let's assume that it is fully rational.  That means that about 2% to 3%
of the companies that have A 30 day paper should default on that paper in
the next month (it's less than 3.5% due to the premium people have always
paid for low risk).  That's a catastrophe on the order of which we haven't
seen in 80 years.

Is this what you are arguing, that there is a very good chance that a
significant fraction of non-financial companies will be in default in a
month?

Dan M. 

_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l

Reply via email to