----- Original Message ----
From: John Williams <[EMAIL PROTECTED]>
To: Killer Bs (David Brin et al) Discussion <[email protected]>
Sent: Tuesday, September 23, 2008 8:48:12 PM
Subject: Re: Meltdown
Gautam Mukunda <[EMAIL PROTECTED]>
> From: John Williams
> Gee, really? It couldn't possibly be just a little more complicated than that?
>
> Me:
> There are more factors,
Yes, that is what I was referring to.
> but in the short-term money market, not that many, really.
Surely the laws of supply and demand still hold? In other words, if
the supply of money to be lent (in aggregate) goes down, then the
price (interest rate) will have upward pressure.
Me:
Yes, but again, it's a circular problem. Why did the supply go down? Because
people were afraid of losing money in the money markets (even though _no one
had_). So there was a stampede for the exits. Fine, but when there's a
stampede lots of people get trampled. Was the risk of defaults across the
market so large that _GE_ was at a higher risk of default? Obviously not. But
such a stampede could cause lots of otherwise fine companies to go under, which
would cause still more companies to go under, and so on, until finally the only
things left would be companies like, well, GE and Microsoft and Pfizer. We
know what financial collapses look like. They look like 25% unemployment rates
and a decade of disaster. It's hard to imagine _anything_ worth the risk of
going through that when interventions can prevent it.
> We had enormous market events followed
> immediately by a pretty-much unprecedented increase in money market interest
> rates, paired (presumably not coincidentally) by a massive flight of
> investors
> from the money markets
Several large firms were in imminent danger of failing. I would pull my money
out,
too, if I thought I might lose it or it might get tied up in a bankruptcy.
Me:
Sure, that's fine. The problem is that when everyone does it these actions can
_cause_ a bankruptcy that will not otherwise occurred. This is a standard
collective action problem. You have described exactly the mechanics behind a
back run. If there's a run on the bank, you want to be first in line. But
since everyone wants to be first in line, you can get runs on banks for no
reason at all. That's not a market functioning perfectly, and if you can
prevent it, you should. That's what the FDIC is for, and here we had a similar
problem.
> This flight was particularly odd given that _no person_ lost money in such
> investments.
Do you wait until you have been in an accident to put on your seat belt?
Me:
No, but _putting on my seat belt cannot cause an accident_. This is the
fundamental problem with your analogy.
> JPMorganChase (run by Jamie Dimon) and Goldman Sachs (run by Lloyd Blankfein)
> didn't. If the government were responsible, you'd have to explain why they
> were
> immune to pressure. Instead they - brilliantly - handled this potential
> crisis
> exactly right.
Which would be great if the government didn't interfere. The strong companies
that
made good decisions would survive, and the others would fail. But instead we
have
the government bailing out the bad companies.
Me:
Well, we don't yet know. _But_. If the run on the money markets had
continued, we would have seen strong companies go under. Like, for example,
Goldman, which probably was in some danger for a while there. That's what
contagion means. You have people who didn't do anything wrong going under.
The only financial institution which I have really high confidence in right now
is JPMorganChase, and even they're not invulnerable.
> For example, we had a series of events occurring. We saw the
> mark-to-market value of financial instruments constructed based on subprime
> mortgages drop to near zero. Functionally that explains the collapse of
> Lehman
> and AIG. Although the value of these instruments is presumably substantially
> lower than their purchasers thought they were, a true value of zero is
> implausible at best (absent strange
> leverage constructions _unviersal across the instruments_ this would imply a
> default rate on subprime mortgages of nearly 100%, which is clearly not going
> to
> happen). This collapse forced Lehman to declare bankruptcy while it was
> technically still solvent - an unprecedented event, so far as I know.
As I mentioned in another post, there is a practical definition of solvency and
a
technical definition. The market seems to follow the practical one. Extreme
leverage necessitates a probabilistic definition of solvency.
Me:
That's fine, but I don't think you've thought through the implications of such
a definition in a period of extreme ambiguity. You can have situations where
_no one knows_ if you're solvent or not. If that's true, and people have
suddenly ramped up their risk aversion, then companies that are, in fact,
solvent can be rendered insolvent simply by the existence of these concerns.
That's what we're trying to avoid.
> But, because of how
> investment banks are structured, if
> people demand higher interest rates for short term loans from them, they
> would
> be forced to go under immediately. Note the problem here: No increase in
> interest rates, no bankruptcy. Increase in interest rates caused by fear of
> a
> bankruptcy, however, can _cause_ the bankruptcy. Any individual investor,
> knowing the bankruptcy would not occur, would happily lend the money. But
> because that certainty does not exist, the money might not be lent.
The 5 big investment banks entered uncharted territory: leverage of 20x or 30x
with large amounts of money-market funding. Evidently, this is not a reliable
structure. It is also a little scary that Paulson (ex Goldman), who certainly
had
involvement with that risky structuring, is now asking for a $700B or more
blank check to buy risky assets.
Me:
Well, it's not that uncharted. Barclay's and Deutsche Bank are both over 50x,
I believe. Now I absolutely think - and have thought for a very long time -
that we need regulatory limits on leverage, probably around the 12x range. But
I doubt that's what you think we should have - I hope I'm not misinterpreting
you. Because you can't get clearer-cut government intervention in markets than
that. But if one institution that has such leverage makes a mistake it can
take down other institutions that make no mistakes. That's a problem, and the
only institution that can do something about it is the government.
> In the face of such events, government intervention is an
> entirely reasonable - indeed, necessary, thing to do.
You sound like a politician. We can save you! We have the technology!
Too bad the track record is emphatically otherwise.
Me:
Well, I think I sound like someone with some experience with markets and some
knowledge of finance. I'm certainly not a politician. But on what do you base
your belief that the track record is otherwise? You haven't cited a single
case. Not one. I would say that the track record of Sweden in the 1990s, or
the Fed with LTCM, or Hong Kong in 1998, to pick three of many examples,
suggests precisely the opposite. Well-structured, skillfully-handled bailouts
can prevent economic collapse at minimal cost to the taxpayer (Sweden and Hong
Kong turned a profit, and the Fed actually didn't lose a penny since it just
played a coordinating role), and even very little if any additional risk of
moral hazard. So I'm standing on an empirical record of success on this one.
Are you familiar with the Swedish case? If you are, why do you think the
lessons there are not applicable to this one?
Here's another way to think about it. You're the manager of a money market
fund. Yesterday you were churning along, doing a fairly boring job, making a
little bit of money on each transaction. No big deal. All of a sudden you see
your friend who runs Fidelity's fund - a guy who, honestly, you've always been
a little jealous of, because that's a more prestigious job than most money
market managing ones - has just had his fund evaporate out from under him.
He's lost his job. His company had to pony up to make his investors whole.
But you know that, in his shoes, you would have made the same decisions,
because money market funds _never lose money_. Now, today, you've got to make
investment decisions. Intellectually you know that things are pretty safe.
Goldman's not going to go under. But, that's what he thought about Lehman.
Now, if you do invest, you won't make much money - because money market funds
have really thin margins. And there's a
chance that you too will go under. A small chance. But pretty clearly it's
bigger one than you thought it was yesterday. On the other hand, if you
_don't_ invest today, you won't make the money you would have made by lending
to them. But you won't lose your job either if you're wrong. What would you
do? Now, all by yourself, that decision doesn't matter too much. After all,
if you don't invest in Goldman, someone else will. But today - every other
money manager is thinking the same things. And if Goldman doesn't get that
money, it's going under. If Goldman goes under, it will take a lot of other
people along with it. When that happens, _everyone_ might go under. But it's
not your job to prevent that. And the worst case outcome is if you invest, but
no one else does. Then you're out of a job. So what do you do? Most managers
won't invest. Given the scale of the downside (you get fired) versus the
upside (you and your investors make a
small amount of money) that's an individually rational decision. But it's not
a collectively rational decision. Most of the time we want people to pursue
their individual best interests. That's how markets work. But sometimes when
everyone does that we can get into a collective action problem like this one,
and here the stakes are as high as it is possible to get. And so now we're
looking into the abyss. It is almost impossible to overstate the stakes of the
potential damage. I guess one way to put it is to say that, so far as I can
tell, even the most aggressive estimates of the damage global warming might do
are almost insignificant compared to the damage from the collapse of the US
financial system. Now you can fairly answer this might not happen. It's true.
Maybe the damage would be contained. But you can't possibly know that. I'm
not saying that about you in particular. _No one_ can possibly know that. We
don't understand the economy
well enough to answer that question. We just know that the last time
something like this happened - 1929 - it didn't end well. My argument is that
almost nothing is worth the risk of the Great Depression all over again, that
economic and finance theory show clear mechanisms through which a panic can
occur and be mitigated through government interventions, and that the
historical record suggests that such interventions can be, and have been,
conducted successfully. I'm afraid I'm not clear on what basis - either
theoretical or empirical - you disagree.
Best,
Gautam
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l
