[Vision2020] [spam] Debate about the bailout

Andreas Schou ophite at gmail.com
Fri Sep 26 17:18:09 PDT 2008


Ken --

Mark-to-model of any sort redefines 'value' and creates illusions of
liquidity. While the underlying accuracy of marking a financial
product to a model may be greater, the company's ability to reduce
that asset to currency is far, far more important to the company's
creditors and shareholders. Once you allow a company to value their
holdings using any reasonable model, you've escaped the bounds of a
market economy: things are worth what the owner says they're worth,
rather than the purchase price of the item. I much prefer the
invisible hand of the market to determine these things, rather than
the incompetent hand of the CFO.

With financial instruments like futures and options, I can understand
the impulse to mark them to model: their value is in limbo until some
specified time in the future. Unfortunately, economics, like physics,
has a problem of observation. A model may accurately predict the the
market's current behavior, but once a significant number of market
actors begin to follow the model, the model itself becomes an economic
actor.

A perfectly accurate, entirely universal economic model would
instantly make itself worse than useless. Let me give you an example:
If, for instance, a model used by every bank accurately predicts that
housing prices will peak on a certain date and then slowly fall, every
rational economic actor will want to dispose of those assets on or
before that date. The banks begin to dispose of those assets prior to
that date. The date of the peak changes. If the model changes to
follow the date, the peak likewise follows it.

Only a secret or retrospective model -- a useless model -- has no
effect on the market.

Frankly, I think the best way to make the market match the model would
be for more people to take advantage of arbitrage opportunities
between complex financial instruments and their  constituent parts*.
That solves the problem of market/model mismatch without allowing
unscrupulous or optimistic corporations to overestimate their assets.

-- ACS

(1) I am not a hedge fund manager, and I am told that this is already
actually done, but that

On Fri, Sep 26, 2008 at 4:27 PM, Kenneth Marcy <kmmos1 at verizon.net> wrote:
> On Friday 26 September 2008 11:15:39 Andreas Schou wrote:
>> > By the way, efforts to throw the blame on the accounting rules are bunk -
>> > Mark to market was the means by which this whole house of cards was
>> > revealed.  Had we not had mark to market rules in place, this whole mess
>> > might have been undetected for some time and the consequences even more
>> > severe.
>>
>> Oh, hell. It pains me to say it again, but -- again -- I agree with Jeff.
>>
>> If banks had been allowed to use a 'mark to model' valuation of their
>> CDO assets, which essentially means to put them on their books as
>> being worth what the bank predicts them to be worth, the crazy
>> debt:asset ratios that banks are carrying would never have come to
>> light. The fact that they're forced to put them on their books at the
>> actual cost the market will pay is  shining sunlight on the problem,
>> not creating it.
>
> Mark-to-market is merely an inventory valuation methodology used specifically
> for financial assets; thus, it is analogous to first-in-first-out (FIFO), or
> last-in-first-out (LIFO), or various weighted average methods used for
> physical inventory, and can hardly be blamed for an industry-wide confidence
> shortage. Other factors need to be sought and examined.
>
> If one asks the question "Why are some not confident in their investments?"
> part of the answer seems to be that the investors don't really know of what
> their investments consist. Really? Why not? Don't we have accounting rules
> that specify the necessary parameters to identify and to value an asset?
>
> Well, as it turns out, for some of these newly-invented financial instruments,
> the answer may be no. Sad, but true. If that is the case, we have been here
> before (in the early 1930s, before securities regulation) and, after better
> accounting rules were adopted for the prevalent assets of that time, those
> who wanted to value firms for investment purposes had better tools with which
> to do so. Here we are again, and what do we need? Better rules to identify
> the nature of these new types of assets, so they can be valued appropriately.
>
> In the 1930s, the word computer applied to a person whose job it was to do
> arithmetic work, regardless whether it was in a large engineering office or
> some other computationally intensive employment. A mechanical, or an
> electromechanical calculator may have been available, but an electronic
> computer was not. In the twenty-first century we assume that computer refers
> to a digital electronic device, and many people who regularly work with
> computers assume that they are connected to one or more networks comprised of
> many other computers. 1930s computers worked at the speed of thought; 21st
> century computers work at speeds many orders of magnitude faster, with more
> accuracy, precision, reliability and predictability, given that their
> software is thoroughly debugged and well-maintained.
>
> We should be using these 21st century computers to better identify and value
> financial instruments. How so? Modern financial instruments have many
> characteristics because often they are composed of many other financial
> instruments sliced, diced and divided into different portions for the purpose
> of repackaging various levels of financial characteristics of the underlying
> levels of financial instruments. If one goes to the grocery store and picks
> from a shelf a box of some foodstuff, one can read a label on the box that
> indicates of what the foodstuff is composed. An analogous action prior to
> purchasing a complicated financial product is often not possible to do. I
> suggest that with the power available in 21st century computers, we can
> determine the ingredients of a financial product as it is designed and
> packaged, and we can keep track of those ingredients' characteristics as the
> product is offered for sale, and as it progresses through its financial life.
>
> In my view there is no reason why, save for the not-yet-developed software,
> financial instruments cannot be tracked from cradle to grave, so to speak,
> and information about those instruments can be made available either
> individually, or in any collection, or part of a collection, that an analyst
> or investor may wish to consider. The necessary computational hardware is
> available, if not necessarily yet installed where it would be useful, and the
> necessary algorithmic theory is available, if not yet packaged in appropriate
> application software. What is needed is the design vision, the project
> management, and the research and development funding to create, test, and
> implement intra- and inter-firm, and intra- and inter-governmental systems to
> better accomplish the various financial transactions of the 21st century.
>
>
> Ken
>



More information about the Vision2020 mailing list