.. continuing a ramble prompted by the Normal / Levy thread :
This deceptively simple question is at the heart of many modelling issues.
Interpreted at a shallow level, this could be a question about whether the values of some variables are within the range they have moved in historically.
At a deeper level, the question is whether models built on history will continue to be fit for purpose in future application.
Models cover a wide range from the deterministic ones of physics, that embody precise understanding of the mechanisms, through to empirical models that merely claim to capture useful patterns. It could be an interesting thread to fill in this continuum with examples; financial models, and especially econometric models, would be up the “empirical” end. There may be various nomenclatures for this kind of discussion – noting for example the von Mises (/Austrian) link provided earlier where “time invariant” is used as a descriptor of certain models.
So, sure, there is wide confidence that the models of physics will still work the same way in the future, but only qualified confidence in (especially) those models that have anything to do with human behaviour, or other complex systems.
Empirical models tend to depend heavily on the choice of the data that “calibrates” them (enough for another thread on this topic). Also, to the extent that they rely on patterns without understanding the drivers of those patterns, there may come a time when they unexpectedly perform less well – perhaps even catastrophically so – than before.
Footnote: I like Andrew’s comment “models are good tools but bad masters” and another well known one “all models are wrong, but some are useful”. Perhaps this latter one should have an addendum “…some of the time”.
17 comments
29 July, 2009 at 08:16
ABOM
Is this piece useful here?
http://mises.org/story/3582
29 July, 2009 at 11:52
ABOM
I also note that the historical data doesn’t even fit within the existing (redundant?) models:
Mandelbrot, in his book “The (Mis)Behaviour of Markets” (2005), lays out the data for the empirical distribution of stock price changes and compares it against the predictions of the bell curve based EMH (any version). Clearly the empirical distribution has fatter tails than the normal distribution implied by EMH. To quote from page 13:
“In fact, the bell curve fits reality really poorly. From 1916 to 2003, the daily index movements of the Dow Jones Industrial Average do not spread out on graph paper like a simple bell curve. The far edges flare too high: too many big changes. Theory suggests that over that time, there should be fifty-eight days when the Dow moved more than 3.4 percent; in fact, there were 1,001. Theory predicts six days of index swings beyond 4.5 percent; in fact, there were 366. And index swings of more than 7 percent should come once every 300,000 years; in fact, the twentieth century saw forty-eight such days. Truly, a calamitous era that insists on flaunting all predictions. Or, perhaps, our assumptions are wrong.”
There is also now clear evidence of a correlation between high volatility and leverage (debt):
http://forum.globalhousepricecrash.com/index.php?act=attach&type=post&id=4127
I merely repeat my plea to consider a new paradigm, where the long term bell curve is replaced with “bifurcating bell curves” as debt levels increase.
This, I believe, would “close the circle” on the models and more accurately predict price volatilities (although not movements) in a whole range of markets.
It would be impossible, a priori, to predict whether a particular market will be captured by a bubble with overleverage – the bubble moves. Dot com. Then housing. Then oil. Then gold. The some other essential commodity.
But accepting that the financial system has Minsky-like Ponzi dynamics would be much more fruitful than assuming it has stabilising bell curve foundations. It has been conclusively proven that it does not.
Get over it.
29 July, 2009 at 14:32
ABOM
Why not throw the modelling geeks another bone?
http://www.nakedcapitalism.com/2009/01/woefully-misleading-piece-on-value-at.html
29 July, 2009 at 16:57
Andrew
ABOM,
Mandlebrot’s piece does not say that you cannot model the markets (in fact it says precisely the reverse, but it does say that the normal curve is not the right one to use – which is exactly the point I have been making.
Can you please decide whether you believe that models are useless (which is the point I think you have been trying to make) or whether they are useful, the only problem being that the normal curve is not the right one to use – as you seem to be arguing here by quoting Mandelbrot approvingly?
29 July, 2009 at 17:41
ABOM
Andrew,
My rule is: If you can’t replace a model with something that “works” (i.e. at least fits with the historical data), don’t use any model at all. Just look at the markets and see what “imbalances” exist and bet against the “imbalances” because they don’t last.
A simple application of the rule would be to short whatever market went up the most last year and go long whatever market went down the most last year (if you do the analysis this actually works out to be a surprisingly successful investment strategy!).
It is the very definition of madness to continue using a model that doesn’t fit with the historical data (at minimum).
Therefore, using this test, anyone using the bell curve in financial markets is literally insane. I don’t mean that metaphorically. I can’t recall who said it (Einstein? Ben Franklin? Chinese proverb?), but someone said the definition of insanity is doing the same thing over and over and expecting a different result. Using the bell curve when the historical data doesn’t fit the distribution fits within this definition (excuse the statistical pun).
I’ve suggested another “model” (bifurcating bell curves) but lots of work needs to be done here. What is the rate of bifurcation? How does this correlate with levels of increasing leverage/debt/gearing? What markets are vulnerable to bubbles? (I suggest markets closest to essential human needs – housing, oil, food…however tulips and dot com I don’t understand – it’s more a social/fashion issue).
Really all this model would do is correlate volatility with leverage. Big deal. That’s not particularly useful from a modelling perspective but it is very powerful from a policy viewpoint. It suggests debt is dangerous and socially destabilising. Policy-makers should be much more concerned about absolute and relative levels of debt, as well as the rate of growth and the marginal productivity of debt (as Fekete brilliantly points out).
Finally, this perspective signals a more fundamental message. Years ago, after 1987, as a student, I searched for an alternate “model” to picture the probability distribution because I was acutely aware we were entering a new world. This came from my interest in the Austrian School understanding of “money” and Mises’ (and Hayek’s) numerous writings on the history and importance of “free market money” (generally gold and silver) and the evils of embezzling FRB.
The lesson is this: The future is MOST LIKELY to be RADICALLY DIFFERENT from the past when the institutions around the creation of distribution of “money” change. If you don’t pick up the importance of these changes you’ll miss the seismic shifts these institutional changes bring about.
August 15, 1971 was a revolution only the Austrians seemed to understand. Repeal of Glass-Steagall – again seismic. These two changes alone should have sent alarm bells screaming “Something DRAMATICALLY different from the past (good? or bad?) is going come from these changes”. Yet no one from the mainstream seemed to understand the importance of what was occurring because none of them understood the importance of “money” and how far from the historical free market in money we had moved.
Again, a tiny tip from a suicidal madman (genius?):
Be prepared for maximum change when the institutional framework around money production (banking and exchange rates) changes. These changes generally manifest their results with a lag of 10 years.
However, in the case of Roosevelt’s criminal confiscation of America’s gold, the lag in its devastating effects has been 80 years:
http://www.fff.org/freedom/fd0610a.asp
http://www.professorfekete.com/articles/AEFForgottenAnniversaryHauntsTheNation.pdf
If I die soon from alcohol poisoning, someone should continue this research. Seriously, my opportunity to continue this research has long past and I throw it out there for others to pick up and run with. Good luck.
30 July, 2009 at 13:26
Andrew
ABOM,
No model every anywhere (at least in finance or economics) will ever completely fit any set of circumstances. Under those circumstances I am not prepared to advocate running a bank based on a really simplistic trading strategy – like the one you advance above.
You cannot run a bank in the same way you seem to think you can run a personal trading strategy. For example, how do you decide on appropriate liquidity levels based on your ideas above? How many fixed loans should we write compared to variable rates? How do we price option contracts?
All of these questions (and many, many more) need answers that simply cannot be done on the seat of the pants. It would be a very quick road to ruin.
Sorry, but all I see when I look at what you are saying is that banking should be a lot less internally controlled – a position I cannot agree with.
30 July, 2009 at 13:50
ABOM
Andrew,
Please don’t put words in my mouth. The strawman has been beaten up but I’m left standing to one side wondering what the Hell you’re doing with the stawman. It’s looks a little perverted from where I’m standing, I must say.
I never advocated running a bank on the investment strategy I suggested above (although a little more counter-cyclical lending by the banks may lessen the stupidity of future bubbles forming, but that’s an issue for Ponzi professionals that will probably elude them).
Nor would I advocate running a bank based on ridiculous, unrealistic “scientific” risk models based on the normal distribution.
Q4, 2008 (and the trillions of dollars of taxpayer support for the dumb banker stuff-ups) is testament to the absolute stupidity of that idea.
Can you accept that the road to Q4, 2008 was paved with faulty models or not?
It’s like you’re attacking me for my lack of rigorous modelling, when you’ve just burnt down the entire banking warehouse. Can you see the destruction behind you? Do you take ANY responsibility for the mess? Do you see the billowing smoke, the fire engines, the pools of murky water, the taxpayer dollars, the economic dislocation, the bankruptcies, the poverty, GM, AIG, Lehman, Bear Sterns, Iceland, Ireland, Eastern Europe, California…. do you see ANY of that? Are you suggesting that the bell curve-based financial models used around the world and the associated securitisation of debt have NOTHING to do with this disaster?
Are. You. Blind?
Isn’t it better to run a bank not via backroom geeks with models and the paper swapping of CDSs and SIVs and securitised toxic trash, but with the traditional 5 “c”s?
http://blogs.sas.com/fairlending/index.php?/archives/56-5-Cs-of-Credit-in-the-Industry-Spotlight-Top-of-Mind-for-Customers.html
http://www.amazon.com/Credit-Risk-Assessment-Borrowers-Investors/dp/B0028T8T36
Impersonal mathematical modelling (which you appear to espouse) has proven to be the most ridiculous, unscientific, useless, dangerous option of all!!!
I’d prefer my way to your “scientific” methodology.
Ha. Ha. Ha.
30 July, 2009 at 21:03
Andrew
ABOM,
So far you seem content to appear to take the role of a critic – criticise the current models but content yourself with that. Any attempt I have made (and there have been many) to get you to say what you actually want to do (other than your seeming half-hearted support of the dirigiste full reserve banking) merely elicits a stream of invective (as above) against the current models. Other than indicating that what may be your preferred option (I think “bifurcating bell curves” – and then saying even that needs a lot of work) that has been it.
Please make a positive statement of what you actually want – this constant wriggling is getting boring, at best.
31 July, 2009 at 09:30
ABOM
1. Absolute banning of naked CDS.
2. Regulation of OTC derivatives, especially in the commodity space. Having a derivative exchange where trades are not OTC but are mediated by a transparent market with adequate capitalisation by the market participants.
3. Audit of all major central banks.
4. Return to the 5 Cs as the major methodology for lending money from the banks (the slow death of securitisation and off balance sheet lending – which is happening anyway, so doesn’t need me or anyone else to push it).
5. Forcing all off-balance sheet sh*t back on balance sheet – increasing financial transparency. Hedge funds and private equity are just bank agents so they should be regulated and brought back on balance sheet.
6. Forced liquidation (in future) of all busted, illiquid or insolvent financial institutions, no matter how big, and immediate forced break up of TBTF institutions.
That’s the minimum. It’s too late for a lot of this stuff (OTC is out of control so how do you kill that beast now?). We’re freaking doomed already, so I’m not really “pushing” for any changes.
It’s like asking me what I would do with a patient that has terminal lung cancer because he smoked for 30 years. There is actually no “problem” to solve. A problem suggests a range of outcomes that will result in dramatically different outcomes. He’s going to die no matter what you do, so he may as well do whatever he likes – smoke, have a party, whatever. I worry only when a decision has to be made. Fortunately (unfortunately?) such a dilemma doesn’t exist in this case.
Your blinkered mindset is a perfect example of why another crisis is inevitable. I raised the 5 Cs and gave clear references for further contemplation by you and I guarantee you haven’t touched the references. You don’t really have any interest in changing your mind, but you keep asking me for solutions when the ones I’ve already given you don’t even consider seriously. What’s the point of talking past each other all the time?
With a terminally ill patient, all you can do is laugh.
31 July, 2009 at 09:54
Andrew
Wow – you are in favour of massive re-regulation. So much for being a libertarian.
As for the 5 Cs, they (or a similar structure) are a common part of just about every single credit scoring consideration methodology I have ever seen. You may think that these are wonderful and novel but to me that just means you have never actually worked in a bank. If you had you would know better.
In this context, though, I was looking more at your position on modelling – on one hand you quote Mandlebrot approvingly, the next you are again excoriating modelling and arguing they should not be used until they can fit historical data. (The fact that backtesting, the method used to confirm this, is a part of every model test I have ever seen is neither here nor there, obviously)
Then you say that the bifurcating bell curves is the way to go, but they need some more work (i.e. according to your theory they cannot be used).
Is it too much to ask for some clarity here, or is there none to be had?
31 July, 2009 at 10:01
ABOM
I’m so lonely. You’ve confirmed your soul has been taken by banking zombies. The 5 Cs are only employed at the coalface – securitisation has destroyed the link so the APPLICATION of the 5 Cs is a joke now. Remember “no doc”? Remember liar’s loans in the US? Remember Godzilla Mozilla?
I’ve said at least a thousand times:
Free banking and abolish the central banks OR heavy re-regulation of banking (even nationalisation).
Both are ridiculous/unacceptable/insane/unreasonable to your tiny banker-brain. So you scream “inconsistency!”, or blubber some other rubbish.
I have provided solutions. Now you don’t like the message. So you kill the messenger.
You’re the one who’s inconsistent, my friend. You are the free marketeer who cannot even imagine a free market in money production.
You are the hypocrite.
31 July, 2009 at 10:20
ABOM
And just on the topic of hypocrisy and blind banker “morality”…
http://www.ft.com/cms/s/0/ff921710-7d2a-11de-b8ee-00144feabdc0.html
Lord, have mercy on the souls of embezzlers….
31 July, 2009 at 13:59
Andrew
You do appear to be lonely, ABOM – up there on the pedestal where you have put yourself.
I do remember all of those, ABOM. Are you talking about the USA or Australia?
31 July, 2009 at 15:31
ABOM
Aust is the US, with a 10 year time lag.
31 July, 2009 at 18:56
Andrew
So – we will have a few hundred disparate and conflicting bank regulators in 10 years? If we follow your advice, perhaps.
31 July, 2009 at 20:51
ABOM
Keeping scoring cheap points on the edge of the abyss, blind man.
I’m the one that’s going to end up laughing the loudest, my friend:
http://www.goldensextant.com/SavingtheSystem.html
31 July, 2009 at 22:20
ABOM
Ha Ha Ha! Even the dumb, arrogant “internationalists” are worried!
http://www.cnn.com/video/#/video/bestoftv/2009/07/26/gps.recession.near.end.cnn