« I Can Hardly Wait, She Lied | Main | Economists On The Candidates »

October 04, 2008

Comments

Why would the government encourage revaluation of mortgage assets (by scrapping m2m) just before the Treasury gets ready to purchase billions of dollars of them?

Don't a lot of people think mark-to-model was responsible for the decade-long recession in Japan in the 90s?

I also don't see the value of changing the asset/debt ratio required. I mean, a lower ratio only means that the firms will get even more leveraged before insolvency, right? That seems like doubling down on a bad hand.

Both mark-to-model and changing the ratio are ad hoc solutions that might have worked in this particular situation, but I can't see how either of them are sustainable policies for the long-term.

And I wonder if we really need anything structural of this sort. If all these banks got caught up in these faulty assets, doesn't that just mean their portfolios weren't diverse enough w/r/t holdings of the entire market? That they simply weren't accurately assessing the price risk?

On the other hand, I can see how saying "Look all we need is a better assessment of risk" sort of begs the question here. The whole problem with asset bubbles is that people don't accurately assess risk beyond a certain point, and bank's are going to get caught up in this psychology just like other investors.

"the second is to change the requirement that they sell those assets whenever their asset/debt ratio gets too low. The second seems obviously preferable: it solves the problem directly, while allowing us to have as much information about the companies we invest in as possible."

I agree with you:

http://www.marginalrevolution.com/marginalrevolution/2008/10/prophets-of-acc.html

"We show that a shock in the insurance sector can cause the current value of banks’ assets to be less than the current value of their liabilities so the banks are insolvent. In contrast, if historic cost accounting is used, banks are allowed to continue and can meet all their future liabilities. Mark-to-market accounting can thus lead to contagion where none would occur with historic cost accounting."

Or, we could have both, and allow exemptions in a crisis on a case by case basis.

Or try McTeer:

http://www.bob-mcteer-blog.com/why-mark-to-market/

"Critics of some mild regulatory forbearance during this most serious of crises since the Great Depression will no doubt cite transparency as essential, but the two can go together. I'm not advocating hiding temporarily impaired assets. They can remain on the balance sheet with a footnote explaining the intent to hold to maturity."

In any case, this problem seems solvable, even if I don't know how to do it.

Ara: I don't think changing the debt to asset ratio is a good idea either, really. But I do think that if people want to argue that we should do something about the spiral described in the post, which turns on a reaction they don't like to information, the response should be not: don't provide the information, but: change the supposedly bad reaction.

The cases in which this is not true are, I think, cases in which the reaction is both effectively impossible to prevent and really, really awful. The fact that people might respond to the availability of plans for hydrogen bombs by building one would be a reason to keep them secret: in that case, you wouldn't want to rely entirely on preventing people from using the information in the wrong way. In this case, however, I don't see anything that really defeats the presumption against trying to solve the problem by making information unavailable.


I also don't see the value of changing the asset/debt ratio required. I mean, a lower ratio only means that the firms will get even more leveraged before insolvency, right? That seems like doubling down on a bad hand.

I think the idea is to temporarily suspend the requirement that firms must sell a portion of their less liquid assets to raise cash, in order to get back into compliance with an asset/debt ratio which they are currently out of compliance with due to a sudden devaluation of their assets. But they shouldn't be able to issue new loans until they are back in compliance with capital ratio requirements. So it is a one way suspension.

I think hilzoy is correct.

If marking to current market price those assets which were brought onto their books during the bubble will trigger an out of control deleveraging of these firms, better to suspend the requirement that they sell off assets than to give up transparency in accounting for the value of these assets.

Ideally we can then have a slower and more controlled deleveraging in stages which is less likely to result in panic selling and irrationally low prices for the assets being sold, i.e. it is better to fall down a flight of stairs than to fall off a cliff.

I'm okay with suspending mark-to-market, with two large caveats. The first is that, any asset that isn't marked-to-market must be disclosed in all of its details. Suspension should not be a way to avoid transparency.

The second is that in the future, we must be as suspicious of asset bubbles as we are the bursting of asset bubbles. I'm tilting at windmills here, and it's never going to happen, but this can't be unidirectional. The real crisis isn't that these assets are deflating in value; it's that they inflated like they did in the first place. This is one place that I disagree with Brad DeLong. Years ago, he recognized that there was a housing bubble, but decided that it was necessary in order to undo the consequences of the 2001 recession. He was wrong. In the long run, the bubble did too much damage.

Can anyone explain, in terms appropriate for an intelligent layman, the difference between mark-to-market and fair-value accounting? I'd suspect the former is an aspect of the latter, but I don't know for certain.

Can anyone explain, in terms appropriate for an intelligent layman, the difference between mark-to-market and fair-value accounting? I'd suspect the former is an aspect of the latter, but I don't know for certain.

Mark-to-market is one of the ways of implementing fair-value accounting. Financial Accounting Statement 157 requires fair-value accounting for all assets that are revalued. (Some are carried at purchase price; that's not the controversial part.) It mandates that mark-to-market accounting should be used wherever possible. If mark-to-market is not possible, it mandates that the asset holder use estimates of future cash flows to find the most likely value of the assets.

As a statistics major who is now almost finished with an accounting degree, I find FAS 157 a bit amusing. It uses statistical methods, and then ignores the main point of the whole discipline: exact answers aren't possible. I think that one of the biggest weaknesses of accounting that's been exposed by this crisis is that its demand for precise answers easily ends up misleading.

I'm spending some of my free time thinking about how one would move to risk-based accounting, in which the entire distribution of possible values of an asset is considered. I don't have anything close enough to an idea to make it worth sharing at this point.

"I can't see why anyone would think that what we really need right now is less transparency." Unless, of course, you're Bear Sterns, or Lehman or AIG. Remember they asserted they were solvent until some one, namely potential lenders, peeked into their books and then finito la musica. If you love zombie companies and Charles Ponzi, then you're a fan of "mark to model."

I'm far from an accountant or economist, but I figured I'd re-post this comment I made on another thread and see what someone knowledgeable in this area thought:

Regarding whether to use mark-to-market or mark-to-model, why not just let the firm choose either one? As long as you have rules for sticking to one or the other for a certain period of time (similar to subchapter-S designation) and it's public information, wouldn't that force a credibility test onto their officers?

I mean, having seen what's gone on lately, any company that in the future CHOOSES TO use mark-to-model is going to get a thorough body cavity search by any potential investor, whereas the *safer* method of mark-to-market would give potential investors more assurance that bonus-driven executives aren't cooking the books, making it easier for that firm to attract capital.

Thus, the market could be in actuality the discipliner that it's only been in theory up until now.

Hi hilzoy. Those who voted for the bailout deserve to get reminded of it frequently for the next 31 days. I've posted their contact information and links to possible challengers over here and my reasons for opposing it over here. In the aftermath of getting steamrolled I think a lot of people may not be inclined to talk about it, and those who made it happen are going to very discreetly celebrate their wonderful fortune (notice how it's gotten back to all horse race all the time in the big outlets?) But those of us who are outraged about it need to keep making noise.

" If mark-to-market is not possible, it mandates that the asset holder use estimates of future cash flows to find the most likely value of the assets."

Forgive me if I'm wrong, but isn't this precisely the point? I have yet to see anyone definitely show that the market values are "wrong" based on what the future cash flows will be if we revert back to historical means in housing/debt prices.

Mark to market may be bad if there is a distressed asset sale (except there is a caveat for this) and it may even accelerate the decline to what "should" happen based on historical values, but what I am picking up is that the market is simply predicting that the bubble is going to burst back to historically sustainable levels and the market prices reflect this fact.

The reason why people are complaining is that is for levels that are worse than the present and people are hoping we don't go back to those levels.....and the accounting rules are pulling us towards them.

Forgive me if I'm wrong, but isn't this precisely the point? I have yet to see anyone definitely show that the market values are "wrong" based on what the future cash flows will be if we revert back to historical means in housing/debt prices.

Note that I said that this applies if mark-to-market is impossible. Not undesirable; impossible. It happens sometimes.


Mark to market may be bad if there is a distressed asset sale (except there is a caveat for this) and it may even accelerate the decline to what "should" happen based on historical values, but what I am picking up is that the market is simply predicting that the bubble is going to burst back to historically sustainable levels and the market prices reflect this fact.

I think one of the problems we are grappling with now is due to the tranching of MBS rather than just the accounting method(s) used to value them.

The junior tranches have risk profiles that are much more volatile than the housing market as a whole, because the risk of defaults and the losses which may ensue when they occur are preferentially concentrated in the junior tranched MBS. This has the effect of amplifying small swings in the default rate of the underlying pool of mortgages so that many of the junior tranched MBS may in the end prove to be either close to par or completely worthless or somewhere in between, and the outcome is very hard to predict. A change in default rates of just a few percentage points can make or break a given set of MBS depending on the details of their tranching.

From an engineering standpoint, you really could not possibly do any worse than to design a system which is composed of tightly coupled subsystems*, which contain positive feedback loops (e.g., pro-cyclical leverage ratios), and which also contain elements that are highly sensitive to small variations in input and amplify them before passing them on to other subsystems (e.g., tranched risk profiles per above).

If you were deliberately trying to design a machine which was capable of blowing itself up due to systemic instability flaws, this would be a nearly perfect design. And that is more or less how our financial system was put together. In retrospect I'm amazed at how long it has lasted.

* Regarding tight coupling, there has been a long running sidebar discussion in comments and sometimes in top level posts at nakedcapitalism (one of which was in part triggered by a comment by yours truly) over the issue of whether this is a good description of the global financial system or not, and to what degree it may also be described as a complex/chaotic system, and what that might mean.

Much of the subsequent discussion has been well over my head, but to the best of my ability to follow along it seems to me that our current system is tightly coupled, in the sense that maximal exploitation of almost every imaginable form of arbitrage and the use of credit default swaps to spread counterparty risk as widely as possible has ensured that it is impossible to perturb any part of the global financial system without sending ripple effects throughout the entire system, and in at least some cases without noticeable attenuation of the amplitude of the signals which are propagated.

You see this playing out in the immediate crisis right now, in the way that every time the fiscal and financial authorities intervene to try to stabilize one part of the system they end up causing risk to migrate somewhere else and destabilize another part of the system.

For example the new discount window and auction facilities the Fed opened in the early stages of the crisis helped those firms with access to them but at the cost of leaving out in cold some firms which did not and accelerating the liquidity crisis suffered by the latter. The crisis in the market for commercial paper appears to be unfolding along similar lines. Attempts to stabilize money market funds are having the unintended consequence of casusing capital flight from riskier investments. The Irish guarantee of bank deposits is creating problems for banks in England. The ban against short selling equities creates a risk of blowing up hedge funds which were relying on that technique to balance their risks, etc., etc.

It seems to me as if every time we try to stabilize one part of this system we end up making another mess somewhere else. That to me is a sign of a system that is suffering from dangerously tight coupling between its parts, and ultimately the central banks and governments trying to stave off a collapse may find themselves with too many leaks in the dike and not enough fingers with which to plug them precisely because of this broad scale architectural flaw.

I'm wondering if a "smoothed" version of mark-to-market might be useful, say an average (possibly weighted) of the closing prices for the last 90 days. When an asset's price is dropping precipitously, this gives its holder a chance to stabilize it before the full effect is felt on the balance sheet.

Assets like mortgages are different from many other kinds of assets. For instance, your share of stock, assuming it doesn't pay a dividend, only has value when you sell it, so mark to market makes complete sense. But a mortgage has value in two ways - either by selling it, or by holding it and receiving the payments. Normally, we'd expect the market value to reflect, in some fashion, the value of holding the asset.

And that is a problem right now - those valuations don't match - mainly because the value of holding the mortgage depends on the creditworthiness of the homeowner, and that information isn't trustworthy, in part because there's been so much fraud by lenders (to inflate the value of these mortgages) and in part because the information isn't clearly retained when you bundle up mortgages.

But the market value *can* be wrong for those assets that aren't intended for sale but for which there is no buyer. How can a mortgage be considered worthless at the moment the bank receives a payment? Yet that's completely reasonable in mark-to-market, when it shouldn't have a value less than the payment amount.

Now, mortgages aren't exotic things. I don't see why Fannie/Freddie can't develop a model that institutions can use as an alternative to market price. It can even be a very conservative model that tends to underprice, which institutions are likely only to use when the market gets really out of whack, but at least it would be stable and we wouldn't be relying on institutions to game their own books.

ThatLeftTurnInABQ VERY good points about how the tranching/coupling makes a "chaotic" system. My day job involves using systems theory to try to better understand biological rhythms so I fully understand your points and have seen some of it in action.

I had read some of the basics on how our system is the worst you could possibly design, but hadn't thought about the impact of having the tranching.

From my perspective this crisis has the potential to completely revolutionize economic theory since it is showing that increased connectivity and liquidity are inherently destationary rather than what you would expect from the current models.

Like you, I don't necessarily understand a lot of the details of the current situation, but on the other hand the problem is so blindingly obvious when you look at it from a systems perspective it is flabbergasting that no one in charge foresaw this.


the problem is so blindingly obvious when you look at it from a systems perspective it is flabbergasting that no one in charge foresaw this.

Very well put.

It makes me wonder if the financial whiz kids who have been putting together this system for a couple of decades suffer from a rather large set of intellectual blindspots, particularly with regard to low frequency / high amplitude events (this blindness is what Taleb coined the phrase "black swans" to describe), and we would be better off if they had possessed a broader set of interests ecompassing other disciplines which are more concerned with systemic stability issues.

Here's a fascinating interview with a hedge fund manager where he expresses the same idea - intellectual diversity matters and its absence can be very costly:

Pt. 2: Black Boxes Enslave Brains


n+1: How do you know all this stuff?


HFM: How do I know all what stuff?


n+1: All the stuff that you know. Did you go to—


HFM: I didn’t go to business school. I did not major in economics. I learned the old-fashioned way by apprenticing to a very talented investor, so I wound up getting into the hedge fund business before I think many people knew what a hedge fund was. I’ve been doing it for over ten years. I didn’t even know what a hedge fund was when I first had this opportunity. I’m sure today I would never get hired.


n+1: Really?


HFM: Yeah, it would be impossible because I had no background, or I had a very exiguous background in finance. The guy who hired me always talked about hiring good intellectual athletes, people who were sort of mentally agile in an all-around way, and that the specifics of finance you could learn, which I think is true. But at the time, I mean, no hedge fund was really flooded with applicants, and that allowed him to let his mind range a little bit and consider different kinds of candidates. Today we have a recruiting group, and what do they do?—they throw resumes at you, and it’s, like, one business school guy, one finance major after another, kids who, from the time they were twelve years old, were watching Jim Cramer and dreaming of working in a hedge fund. And I think in reality that, probably, if anything, they’re less likely to make good investors than people with sort of more interesting backgrounds.


n+1: Why?


HFM: Because I think that in the end the way that you make a ton of money is calling paradigm shifts, and people who are real finance types, maybe they can work really well within the paradigm of a particular kind of market or a particular set of rules of the game—and you can make money doing that—but the people who make huge money, the George Soroses and Julian Robertsons of the world, they’re the people who can step back and see when the paradigm is going to shift, and I think that comes from having a broader experience, a little bit of a different approach to how you think about things.


n+1: What’s a paradigm shift in finance?


HFM: Well, a paradigm shift in finance is maybe what we’ve gone through in the sub-prime market and the spillover that’s had in a lot of other markets where there were really basic assumptions that people made that, you know what?, they were wrong.


The thing is that nobody has enough brain power to question every assumption, to think about every single facet of an investment. There are certain things you need to take for granted. And people would take for granted the idea that, “OK, something that Moody’s rates triple-A must be money-good, so I’m going to worry about the other things I’m investing in, but when it comes time to say, ‘Where am I going to put my cash?,’ I’ll just leave it in triple-A commercial paper, I don’t have time tothink about everything.” It could be the case that, yeah, the power’s going to fail in my office, and maybe the water supply is going to fail, and I should plan for that, but you only have so much brain power, so you think about what you think are the relevant factors, the factors that are likely to change. But often some of those assumptions that you make are wrong.


n+1: So the Moody’s ratings were like the water running…

HFM: Exactly.

"I have yet to see anyone definitely show that the market values are "wrong" based on what the future cash flows will be if we revert back to historical means in housing/debt prices."

I can't definitively show anything in this area, it being beyond my expertise. But a couple of points are relevant:

First, some of the complicated derivatives don't actually have histories of the sort that would allow us to say what their historical norms are. They're just too new and too complicated.

Second, as Martin says:

"But a mortgage has value in two ways - either by selling it, or by holding it and receiving the payments. Normally, we'd expect the market value to reflect, in some fashion, the value of holding the asset."

My sense is that some people who are in trouble because of this are people who do, in fact, want to hold mortgage-backed securities to maturity, who are getting income from them, who have no interest in selling them, but who nonetheless have to mark them down to near zero. They are in something like the position of the hypothetical homeowner I wrote about, who has to track her assets and maintain a given asset/debt ratio: even if she has precisely no interest in selling her home, is getting exactly what she wants to out of it, etc., she can find that a serious housing crash forces her to sell it.

My general point was: when you have a problem that takes the form: people have to disclose a certain type of information, and as a result bad things happen, you can take two routes: (a) prevent the disclosure of the information, and (b) prevent its bad consequences. And (b) should be presumed to be better.

In these cases, that might mean something like this: firms should disclose the market value of their assets, but regulators/creditors/whoever should be open to the idea that if a firm has a (documented) history of holding a certain type of security to maturity, and it can produce a reasonable basis for assuming a given level of foreclosures or other events that will cause some or all of that security to default, then it should be able to use a valuation based on the income stream, discounted by the likelihood of default/whatever.

As I am sort of making this up, I'm completely open to the possibility that this would not work, for some reason. But it is an example of my general assumption that one should target the bad response to information, not the information itself. Especially at a time when lack of transparency is one of our problems.

I can't remember where I saw it now, but one of the financial bloggers had a link to a video in which an economist applied Churchill's quip about democracy being the "worst system except for all the others" to mark-to-market accounting.

Yeah, flabbergasted. I'm flabbergasted that people are so nonchalant about a policy explicitly intended to increase the "demand" money supply by about a third just to create a "command" economy for derivatives. That strikes me as crazy. Soviet Union in the 70s crazy. It might work, but it's a little too close to spending the milk money on powerball tickets for my taste.

Like mikkel, I view economics from a theoretical biology background and as far as I can tell, this bailout business is exactly wrong in practically every way. In effect, we're betting about a third of our "real" (on-demand) money supply that a particular small group of conflicted-interest individuals will be able to construct a predictive model of a system that's very obviously on the verge of a phase change. It is, quite literally, like betting that people who have never seen H2O below 0°C will be able to predict that it not only solidifies but expands. Or else that they'll be able to prevent 60cc of water from freezing by adding 1cc to it, without knowing what the temperature of the 60cc is or what else is affecting it.

The reason we have statistics in the first place is to deal with systems like this, that are too easily perturbed, and resist modeling. Systems where the only way to predict their future behavior is by extrapolating from their past behavior and constructing expectations post facto (though that's pretty perilous too, as Nick Taleb, or any legitimate prospectus, reminds us ;-) The other alternative (to statistics) is natural selection, sometimes aka "market forces."

* mikkel, would you be at all interested in discussing what your day job is? I wrangle bits for a living, but my passion is living systems, and Steve Strogatz is right up there with Varela and Axelrod in my view.

The interview with the hedge fund guy is interesting, and mostly right on. I would add something else about most of the people that go into finance these days. Most of them go into it because they want to be rich, not because they have any particular fondness for the subject. The people I sit in business classes with don't really have any intellectual curiosity. Not like the people I sat in history or math classes with.

I think that this blinds them to a lot of the pitfalls. I tend to think that I have an advantage on them by entering the field without the motivation that I thought it would make me rich. I was just trying to figure out what I wanted to do with a statistics degree, and took a class on the math of options, mostly because I could take it during the evening, unlike almost all the other math classes. It turned out that I loved the math. I still don't understand a lot of it, but I love it. So, I went into options trading.

The intellectual flexibility helps, but the motivation leads to fewer blind spots, I think.

My sense is that some people who are in trouble because of this are people who do, in fact, want to hold mortgage-backed securities to maturity, who are getting income from them, who have no interest in selling them, but who nonetheless have to mark them down to near zero.

This is absolutely correct (IMO). Interestingly, this relates to something that I didn't realize until yesterday afternoon when I was trying to explain to my boss why I thought the bailout was a bad idea (Disclaimer: I work at a brokerage). If the m2m requirement goes away, there should (theoretically) be no need for the government to spend any money at all. If there's no m2m requirement then there's no need for anybody to buy or sell these derivatives in order to balance their books. And therefore no need for anybody to buy them.

First, some of the complicated derivatives don't actually have histories of the sort that would allow us to say what their historical norms are. They're just too new and too complicated.

Again, this is absolutely (IMO) correct. Only that's a problem, not a solution. The only value these derivatives ever had is based on actuarial predictions. If those actuarial predictions are not (reasonably) accurate then nobody knows what the mark-to-maturity value of the derivatives is. If they are (reasonably) accurate (as far as the banks which hold them are concerned) then why would the banks holding them want to sell them, once the m2m requirement was no longer an issue?

The potential for the downward spiral that Hilzoy described was recognized when the mark to market rules were put into place. The solution was to say that distressed sales don't count. If a bank is forced to sell assets immediately in order to meet capital requirements, the prices of those sales should not be used as the market price when valuing similar assets. If the only sales that are taking place are distressed sales, then there is no market price and the accountants stop using mark to market accounting. Once the markets return to normal operation, mark to market accouting resumes.

One can argue that this is not sufficient to prevent the downward spiral, but a lot of proponents of abandoning mark to market accounting simply ignore the existing solution, rather than arguing that it doesn't work.

Like mikkel, I view economics from a theoretical biology background and as far as I can tell, this bailout business is exactly wrong in practically every way. In effect, we're betting about a third of our "real" (on-demand) money supply that a particular small group of conflicted-interest individuals will be able to construct a predictive model of a system that's very obviously on the verge of a phase change.

I disagree with this. This is not a bet that people are going to be able to build such a model. They already tried, and failed. This is about cleaning up after that failure. There is a huge amount of deleveraging that is going to take place. This is a bet that throwing a huge pile of money into the system can cushion the blow of that.

The reason we have statistics in the first place is to deal with systems like this, that are too easily perturbed, and resist modeling. Systems where the only way to predict their future behavior is by extrapolating from their past behavior and constructing expectations post facto (though that's pretty perilous too, as Nick Taleb, or any legitimate prospectus, reminds us ;-)

My argument is that statistics are inherently flawed for looking at this. The mathematical models are doomed to work right up to the point that they don't, and if the financial system is built on them, it will always fail. There are two major problems that are going to defy modeling.

The first is that statistical distributions depend upon the assumption that the events that make them up are independent. The problem in finance is that the constituent events stop being independent at the worst possible times. Take mortgages. All of the models depended upon the idea that each default was independent of all of the other defaults. This was true as long as house prices kept going up. The instant they started to decline, the defaults were suddenly highly correlated.

The other problem is that the use of the models directly affects the universe they are trying to model. It isn't just that the future isn't like the past. It's that the methodology guarantees that the future won't be like the past.

The whole business model of the finance industry is flawed. Once upon a time, banking was a boring, low margin business. Among other things, that's because those of us in the finance industry don't really create any wealth. Done right, we create an environment that makes it easier for other people to create wealth. In the end, we won't fix the problem until banking goes back to being a low margin, boring business.

This is a bet that throwing a huge pile of money into the system can cushion the blow of that.

That's a fine way to describe it. Question: how do you cushion a blow without knowing what the trajectory of the falling object is? Answer: by putting cushions everywhere. Question: what if you run out of cushions?

That "huge pile of money" is only a small fraction of the size of the derivatives "market" but it's a nontrivial fraction of the "real wealth" of these United States. Mortgages + mortgage derivatives ~= $60t if I remember correctly. Please feel free to correct me if I'm wrong. The main point being that nobody, and I do mean nobody, actually knows exactly what book value is currently represented by assets which would eventually be "non-performing" if held to maturity. It could be $1t, in which case all we have to do is hope that Hank Paulson and the bankers can figure out which $1t, or it could be $30t, in which case we might as well switch to a different currency right now.

In the end, we won't fix the problem until banking goes back to being a low margin, boring business.

Yep. I've read your comment carefully several times now and I still have to agree with practically all of it. :)

Mortgages + mortgage derivatives ~= $60t if I remember correctly.

That's the notional value. Notional value has nothing to do with the amount of money actually at risk. Honestly, I wish everyone would stop citing it, because it's meaningless.

Just looking at this from a control theory standpoint, replacing M2M with a rolling average does look like a good idea, it would produce an effect closer to PID control than what the markets have now.

BTW, if I may ask, what's with all the media outlets continuing to describe this as a "$700 billion bailout"; Did the extra $150 billion the Senate added somehow escape their notice?

Hilzoy said:
"First, some of the complicated derivatives don't actually have histories of the sort that would allow us to say what their historical norms are. They're just too new and too complicated."

This is true, but derivatives "derive" their value from underlying assets. I was talking about the underlying "assets" i.e. housing resorting back to norm.

J. Michael Neal mentioned that when a bid is impossible that's what they are supposed to go by and LeftTurn pointed out that the derivatives fluctuate in value WILDLY because they are too sensitive to changes in the underlying assets and so they are intrinsically hard to price.

So you could argue that the derivatives muck everything up because a 28% peak to trough decline may be wildly different than a 30% peak to trough decline and that is just too small of a change for us to predict correctly. Leverage strikes again...

I agree with J. Michael Neal completely and it's something I've been hammering on for a while.

Using statistics that assume normally distributed independent processes just doesn't make sense for nearly any complex real life system and yet it is the cornerstone of science and economic understanding.

Also economic modeling even from a systems standpoint is unlikely to work for the point he mentioned. It can guide us to what *not* to do, and give qualitative information, but I don't think there will ever be a definitive quantitative relationship between variables like traditional economists try to derive. It's hard enough coming up with a nonlinear system model when you can do tons of experiments and the underlying parts don't change...

(radish, briefly my job is to help design a standardized system to investigate physiologic data using traditional techniques and also look for places when they fail, then use that information to help doctors, physiologists and neuroscientists develop diagnostics and basic models. If you want more of an explanation my email address is igalo and the domain is yahoo.com : I'm actually always on the lookout for new fields to apply it.)

Everybody yet read up what IRS authority was granted in the bailout bill? Permanent undercover authority, sharing of info with other governmental agencies, all sorts of goodies.

Notional value has nothing to do with the amount of money actually at risk.

Sure it does. Notional value gives you a sense of how much real money a market might be able to "absorb" at any given time. How many cushions would be needed in order to guarantee a safe landing (roughly speaking) in the event of a collapse. Knowing the notional size of a market is handy if you're trying to figure out whether you'll be able to influence it.

The whole point of the bailout exercise is to reduce the risk of participating in that marketplace, without the usual tedious and embarrassing need for bankruptcies to remove that risk in the usual way. We cannot simultaneously claim that it's a plausible exercise, and also that the money being used to implement it is not itself at risk.

TLTinABQ, thanks, wow, those nakedcapitalism threads are fascinating...

mikkel, thanks, email to follow...

Brett: what's with all the media outlets continuing to describe this as a "$700 billion bailout"; Did the extra $150 billion the Senate added somehow escape their notice?

The extra $150 billion is not bailout money, and might have passed on its own. Henry Paulson has access to $700 billion, not $850 billion.

The other $150 billion is a combination of things, none of it bailout-related: hurricane relief, AMT reform, renewable energy tax credits, mental health parity, the notorious wooden-arrows industry favor, etc. etc.

Not sure how the big increase in account limits insured by FDIC is figured in; could be almost no real cost if there are not several more big bank failures, or could be pretty huge if events cause anything like a run on the banks.

@Gary: I don't like those IRS authorities one bit, either, but they seem to be something that's actually been in force for most of the time since 1998. Which doesn't make them any better, but does change their aspect as a terrifying new development.

The Obama administration has to make its priority auditing the taxes of those with money. It would be surprising if the Bush-Cheney regime hadn't politicized the IRS along with the Justice Dept, EPA, etc.

radish: one of the financial bloggers had a link to a video in which an economist applied Churchill's quip about democracy being the "worst system except for all the others" to mark-to-market accounting.

It was Alan Blinder or Peter Orszag, in a panel from this past April. Video at the link here. A worthwhile panel, in which much of what's now happening is foretold, as the worst-case scenario...

The comments to this entry are closed.

Blog powered by Typepad