The subprime meltdown and what's really wrong with the banking system

44 views
Skip to first unread message

Ryan Fugger

unread,
Aug 11, 2007, 12:19:11 PM8/11/07
to rippl...@googlegroups.com
I just finished reading this article I found on reddit.com,

http://www.lewrockwell.com/orig8/wallach1.html

and something tweaked in my head. The lenders -- banks and other
investors in loan-backed securities -- allowed real estate to be
massively overvalued, and now, instead of admitting their mistake,
letting the currency remain stable, and taking their lumps in the form
of defaults, they are injecting money into the markets, creating
inflation so the current overly-high price of housing becomes correct,
and the price of everything else rises to come into line relative to
it. In doing so, they maintain the viability of their loans and
reduce their obligations to their depositors, maintaining their
profitability.

But that is not the justification for their actions. Their
justification is that the system must be stabilized to avoid a crisis
of confidence. The banks must appear to have been right and in
control all along, otherwise the monetary system could collapse, and
the economy along with it, as cascading defaults lead to bank runs
like at the start of the Depression -- when the central bank actually
tried to get banks to take their lumps. The fact that we depend on
continued mass confidence in the banks' system means that when the
banks screw up, we must all change *our* ways to remain in step, lest
it become known that the Emperor has no clothes.

This is how the banks control us. They said, "build houses," and
those that built houses got rich on the backs of the rest of us,
regardless of whether those houses had value or not. (With peak oil
looming, it increasingly seems that suburban houses may not have much
value in the future.) Our labour is about to be taxed through
inflation without our consent to retroactively pay those who did what
the banks wanted. We must collectively obey or the system that feeds
us will grind to a halt. That system is a confidence game run by
banks in which we maintain our quality of life only by collectively
doing as the banks say. It is a system driven by threats and fear.

When maintaining the appearance of competent leadership becomes more
important to the operation of the system than actually heading in the
right direction, we have a recipe for a system that cannot
self-correct and will sooner or later head right off a cliff. The
central banks are in a tight spot, because they are in a position
where to save the system in the short term they must allow it to
continue on the unsustainable path they have set it on: continually
increasing production of consumer goods of dubious value in the face
of dwindling resources. It seems that unless something changes in the
way we pay each other, we will be forced to destroy our resources in
order to support the system that we use to distribute them.

Ryan

David Watson

unread,
Aug 15, 2007, 6:01:08 PM8/15/07
to Ripple users
Ryan,

I've believed for a while now that we are on a path of managed
collapse of the dollar. Meaning, the banking elites essentially know
the dollar will eventually collapse and are trying to manage it's fall
without causing catastrophic instability and by acquiring as much real-
world wealth as they can in the process. Many, high level investors
have already begun diversifying. I speculate that this is towards the
end of further erosion of national sovereignty and putting in place
the NAU, or some other EU style arrangement here in the U.S.

Mike Hearn

unread,
Aug 19, 2007, 5:43:58 PM8/19/07
to rippl...@googlegroups.com
> I just finished reading this article I found on reddit.com,
>
> http://www.lewrockwell.com/orig8/wallach1.html

It will be interesting to see if the Feds own report on the money supply:

http://www.federalreserve.gov/releases/h6/current/

concurs with the authors estimate of money supply increase when they
update it to cover the period in question.

Daniel Reeves

unread,
Aug 19, 2007, 6:32:25 PM8/19/07
to rippl...@googlegroups.com
I'm skeptical of Ryan's theory (cautiously skeptical as I believe Ryan
is smarter than me!).
But I'd be willing to bet $1000 that the dollar won't collapse.
Oh, wait, it's impossible for me to lose that bet. :)

How about if we denominate the bet in, say, gold? Any takers?
I'm serious. (Payout can still be in dollars, just using whatever the
current dollar-value of gold is. And of course we'd have to pin down
exactly how and when the bet would be resolved which I won't bother
with unless anyone is interested.)

I'm actually doing research on new kinds of betting mechanisms now
which is why I'm excited about the bet idea. And of course it makes
disagreements much more interesting when the participants can quantify
their confidence in their own positions.

Daniel


--
http://ai.eecs.umich.edu/people/dreeves - - search://"Daniel Reeves"

Ryan Fugger

unread,
Aug 23, 2007, 12:13:44 PM8/23/07
to rippl...@googlegroups.com
I'd probably take that bet, depending on the terms. How would we
decide who won?

Ryan

Daniel Reeves

unread,
Aug 23, 2007, 2:21:05 PM8/23/07
to rippl...@googlegroups.com
Could we predict the future official price of gold as measured in US$
(equivalently, the price of dollars as measured in gold)?
We could take today's price as the benchmark and predict the percent
increase for whatever time horizon you think is appropriate (but has
to be committed to ahead of time, of course).

I have a new betting mechanism where the interface is just "how much
money (in our case, gold) do you want to risk and what is your
expectation of this random variable (price of gold in our case)?"

Would anyone else want in on this?

(For those just tuning in, Ryan predicts the collapse of the dollar,
which means he predicts that the price of things that don't care about
the value of the dollar, gold say, will surge (as measured in
dollars). I'm offering to bet that he's wrong.)

Ryan Fugger

unread,
Aug 24, 2007, 1:32:21 PM8/24/07
to rippl...@googlegroups.com
OK, that's not a bad idea. Gold is at $668 right now. I guess for
there to truly be a dollar collapse, gold would be steadily pushing
towards $1000 a year or two (or maybe three) from now. Although,
looking at historical data
(http://www.kitco.com/charts/historicalgold.html), if it breaks $750
that would really be something. I guess I'd need to know the betting
mechanism before I place my wager :)

For the record, I certainly would be happy to see the banks
successfully defuse this crisis by walking the line between
deflationary depression and continued money-printing. I'm just not
sure they'll be able to pull it off this time, especially if oil
supplies show stronger and stronger signs of peaking, and I think
they'll lean more towards money-printing rather than lethally tough
medicine.

As a thought to counter the conspiratorial bent of my original post,
the banks' power to control the economy should ultimately be
constrained by our own willingness to go along. If the current system
fails, there are alternatives that can grow to fill its shoes :)

Ryan

Daniel Reeves

unread,
Aug 27, 2007, 6:39:51 AM8/27/07
to rippl...@googlegroups.com
Ok, I've got a prediction market / betting pool set up. The interface
is really clunky but all you need to do is specify an amount of money
(in today's dollars, ie, when gold is $668/ounce) that you're willing
to put at risk, and your prediction of the future price of gold. You
can then change your prediction at any time. Or if you don't want to
have to keep thinking about it, we can agree to just put in one
prediction now and not touch it (I guess I like that better). I'm
also fine with moving the payout date as far out as you want. Right
now it's set to one year from now.

The mechanism is incentive compatible in the sense that you maximize
your expected payout by predicting your true expected price of gold,
bounded between $668 and $750 (so predicting outside those bounds is
the same as predicting at them).

I put in $10,000 on my prediction of $668 (I believe what the markets tell me :)

http://yootles.com/wager/gold/

Details about the mechanism: (But before you try to grok the
explanation it may be simpler to just try some hypothetical bets and
hypothetical outcomes with the -- did I mention painfully clunky? --
interface above. Anything in the big text area you can edit,
including the "hyp" variable which lets you give a list of
hypothetical outcomes to compare the resulting payouts. Once we're
satisfied we'll set all the parameters and past bets in stone.)

----- gory details below; feel free to skip this part ----
(and I'll add that I promise that what it shows for the hypothetical
outcomes is what will really happen -- there's nothing particularly
tricky, or stochastic, or even discontinuous)

The mechanism matches our bets like in poker when you go "all in".
For example, if just us two are betting and you put in less than me
then any excess amount I put in is just returned to me. Once we're
matched with the same amount of money, a *proper scoring rule* is
applied to our predictions -- specifically, your "score" is the
negative squared difference between your prediction and the outcome,
rescaled to be between 0 and 1. It then adjusts our scores by
subtracting the other person's score. Multiply that adjusted score by
the common amount of money put in and that's your payout. It gets
more complicated to explain (but not more complicated to use) with
more than 2 people or with people updating their predictions or bets.

For full details, see sections 2.1 and 4 of this paper:
http://yootles.com/outbox/betbud.pdf

Daniel

Reply all
Reply to author
Forward
0 new messages