A friend of mine runs a small bank. He said the regulators came in, and said that they had too much money invested in brokered deposits. Like any businessman he wanted a spotless review, because any negative marks could imply he could not do certain things, such as expand a new office, acquire or be acquired. Bad reviews give the government an undefined option to meddle, veto, who knows? So he asked, what level would be alright with you Fed guys? They said, 'that's a business decision". My business friend noted they were very clear that they do not give advice or anything that could be construed as advice.
Translation. We are suspicious of your exposure, but do not want to defend our suspicions.
This is government in action, afraid to make any hard decisions. They just want to s be invited to the boardroom during large corporate events, like some clueless executive eager to seem relevant by being at all the key meetings. They do have a veto power, enough so that people won't laugh at them, but they so fear screwing up, they won't apply it (except with hindsight, where they will be laser-focused on investments that went down).
How could 'more' of this kind of oversight be helpful?
Tuesday, November 10, 2009
Sunday, November 08, 2009
PC Bias on Fort Hood Shooter Extends to Mortgages
The Fort Hood shootings are an example of how an event can be considered a anomaly, or a broad pattern, depending on the preconception. The major media and the White House have a strong belief that Muslim extremism is idiosyncratic and mainly irrelevant. As Chris Matthews stated, 'we may never know if religion was a factor at Fort Hood' (who's this 'we' kemosabe?). Obama cautioned against 'jumping to conclusions. The New York Times has the headline articles "Army Chief Concerned for Muslim Troops, and also Little Evidence of Terror Plot in Base Killings, and finally, Painful Stories Take a Toll on Military Therapists. The narrative they want to tell is that a poor psychologist was overwhelmed by the stress of listening to troops discuss their stress by going over to Iraq, and that the biggest problem created by this event is anti-Muslim bigotry.
Consider that when 4 college kids were killed at Kent State in 1970 it was quickly decided this was the signature event of how the government war machine was killing unarmed American kids, as opposed to an unintended accident caused by students bent on increasing anarchy until something happened. James Byrd was a black man murdered by some white supremacists in 1998. There are prime time documentaries, foundations, and major references by politicians and pundits on this event, as if it signified a broad issue. Actually it was highly unusual, most interracial violence involves black perpetrators and white victims, the disparity in crime propensity is on the order of the male/female difference.
Events are either anomalies or examples of a pattern based on a simple politically correct view of how the world works, still based on Marx's class lens: the dominant class is responsible for everything bad done by everyone, either directly or indirectly. The Statistical Abstract of the United States has lots of tables on crime victimization by race, but not the perpetrator by race, because we don't want to blame the victim.
How does this relate to finance? The Fed keeps easy to read data on mortgage rejection rates by race, but hides default rates by race, which has led to innumerable simplistic newspaper stories that have 'proved' rampant discrimination by banks. As banks and regulators addressed the mortgage disparity, it was seen as simple justice. Fed Governor Edward Gramlich noted in 2004:
The same principle is involved in education, crime, and borrowing, that of seeing any behavior by socially disadvantaged groups as more evidence of their victimization by the dominant majority. Policies predicated on mistaken assumptions make things worse. By promoting the belief that bigotry accounts for most of every disadvantaged group disparity, the PC elites are doing more harm than good to everyone. Their bad solutions are then applied to everyone, creating a race to the bottom based on great intentions.
Consider that when 4 college kids were killed at Kent State in 1970 it was quickly decided this was the signature event of how the government war machine was killing unarmed American kids, as opposed to an unintended accident caused by students bent on increasing anarchy until something happened. James Byrd was a black man murdered by some white supremacists in 1998. There are prime time documentaries, foundations, and major references by politicians and pundits on this event, as if it signified a broad issue. Actually it was highly unusual, most interracial violence involves black perpetrators and white victims, the disparity in crime propensity is on the order of the male/female difference.
Events are either anomalies or examples of a pattern based on a simple politically correct view of how the world works, still based on Marx's class lens: the dominant class is responsible for everything bad done by everyone, either directly or indirectly. The Statistical Abstract of the United States has lots of tables on crime victimization by race, but not the perpetrator by race, because we don't want to blame the victim.
How does this relate to finance? The Fed keeps easy to read data on mortgage rejection rates by race, but hides default rates by race, which has led to innumerable simplistic newspaper stories that have 'proved' rampant discrimination by banks. As banks and regulators addressed the mortgage disparity, it was seen as simple justice. Fed Governor Edward Gramlich noted in 2004:
Given the generally low level of serious delinquencies, a purely numerical analysis seems to suggest that significant net social benefits have resulted from the rise in credit extensions and homeownershipAt the end of the article he notes
Rising to these challenges will ensure that continued subprime mortgage lending growth will generate even more social benefits than it seems to have already generated.The endgame to this was inevitable because no one was going to stop the trend towards easier lending criteria, let alone reverse it. If the statistical disparity was simply due to bigoted discrimination, closing the gap would be costless. As they say, things always end badly, otherwise, they wouldn't end. Once subprime blew up, sticking with the Marxist narrative right-thinking people were quick to blame banks for forcing ill-advised mortgages on minorities.
The same principle is involved in education, crime, and borrowing, that of seeing any behavior by socially disadvantaged groups as more evidence of their victimization by the dominant majority. Policies predicated on mistaken assumptions make things worse. By promoting the belief that bigotry accounts for most of every disadvantaged group disparity, the PC elites are doing more harm than good to everyone. Their bad solutions are then applied to everyone, creating a race to the bottom based on great intentions.
Thursday, November 05, 2009
Innovation is Not Rewarded
That is, for the innovator. We all benefit from the Pasteurs, the Fords, even the Bill Gates. They create things with spillovers. Yet, you average innovator is simply wrong, hearing dog whistles others don't, and like Van Gogh getting his appreciation after dying, most innovators do their bidding for those they do not know. Deviating from the consensus produces all the things that has elevated us from hunter-gatherers, yet, the record for the individual innovators themselves, is decidedly negative.
Look at any nonfiction bookshelf, and there you will find lots of new ideas, most irrelevant, the rest just wrong. Important new ideas are extremely rare, which makes sense because their marginal cost, after being discovered, is trivial, so they spread. We don't have to independently discover the wheel, crop rotation, or why a Republic dominates a monarchy. If good ideas were discovered all the time, our heads would explode; alas, most weeks, even months, we learn nothing new, and not for lack of trying.
There's an interesting article on whether Liberals or conservatives are smarter in The American. I liked this snippet related to how IQ relates to religiousity:
So, the fact that most atheists in America, who are the minority, are smarter, does not imply that Atheism is necessarily a smarter choice, but rather, that in general elites tend to deviate from the default choice. Eventually, the masses emulate in the elites in subsequent generations, as Christianity or civil rights starts out as the belief of a small vanguard, and then becomes the modal choice of the great unwashed.
In my book Finding Alpha, I define risk as a deviation from the consensus. This is what is implied by a status oriented utility function, because you can always remain in the same spot by doing what everyone else is doing. Taking risk can mean either having zero exposure to the stock market, or double the average exposure; standard exposure is 'no risk'. This is why risk, and return, are not correlated empirically (see videos here, read book here).
I personally have known a lot of really smart people and have to say they are more unconventional in their ideas, yet most of their ideas are crazy. If you have ever been to a Mensa meeting (IQ but little formal education), you realize how things like homeopathy, or truthers, get their bearings. If you have ever hung out with PhDs, you know how limited their competence scope is (at research universities they have the same IQ as Mensans, but are more disciplined and less creative). It's no wonder guys like stereotypical MBAs, who are not so analytical but rather personable and articulate, tend to dominate society. I suspect MBA rule is less catastrophic than PhD or Mensa rule, if only because they aren't as certain of themselves. This all gets back to the idea there is an optimal IQ, and it's not 180, but rather, say, 125 (probably the modal IQ for any large group leader, such as Presidents and CEOs).
Being smart is a good thing, and I'm happy when my kids do well on cognitive tests because of what this portends for their life (as Charles Murray noted, most people would prefer their kid had 15 more IQ points than get $1 million on their 21st birthday). Yet highly intelligent people tend to innovate more, and such innovation tends to be counterproductive for the innovator. So, the fact really smart people can answer a question faster or more accurately than others, is at some point offset by the fact that when they have to supply the question--as is the case once they leave formal schooling--they will be attracted towards less conventional, usually irrelevant or wrong, paths. For every Steve Jobs or Albert Einstein there were many who lived and died in obscurity; for every Black-Derman-Toy there are hundreds of insanely convoluted, in-house models, of no value.
So, if you attempt to innovate, take pride that you are doing it with blatant disregard for your narrow self interest. Objectively, you will fail with a high probability. Like thinking about the lottery, there are happy delusions of grandeur considering your improbable future success. You can counter those who say, you should be feeding orphans in an African village, in that your efforts have large positive externalities. Most importantly, it's fun.
Look at any nonfiction bookshelf, and there you will find lots of new ideas, most irrelevant, the rest just wrong. Important new ideas are extremely rare, which makes sense because their marginal cost, after being discovered, is trivial, so they spread. We don't have to independently discover the wheel, crop rotation, or why a Republic dominates a monarchy. If good ideas were discovered all the time, our heads would explode; alas, most weeks, even months, we learn nothing new, and not for lack of trying.
There's an interesting article on whether Liberals or conservatives are smarter in The American. I liked this snippet related to how IQ relates to religiousity:
there would still be a positive correlation between IQ and atheism. The correlation exists not because smart people have necessarily rejected religion, but because religion is the "default" position for most of our society.
This same principle works in places where the default and iconoclastic beliefs are reversed. Japan, for example, has no tradition of monotheistic religion, but the few Japanese Christians tend to be much more educated than non-Christians in Japan. By the logic of someone who wants to read a lot into the Stankov study, Christianity must be the wave of the future, perhaps even the one true faith! But, of course, the vast majority of educated Japanese are not Christians. Just as with atheism in the West, the correctness of Christianity cannot be inferred from the traits of the minority who subscribe to it in Japan.
So, the fact that most atheists in America, who are the minority, are smarter, does not imply that Atheism is necessarily a smarter choice, but rather, that in general elites tend to deviate from the default choice. Eventually, the masses emulate in the elites in subsequent generations, as Christianity or civil rights starts out as the belief of a small vanguard, and then becomes the modal choice of the great unwashed.
In my book Finding Alpha, I define risk as a deviation from the consensus. This is what is implied by a status oriented utility function, because you can always remain in the same spot by doing what everyone else is doing. Taking risk can mean either having zero exposure to the stock market, or double the average exposure; standard exposure is 'no risk'. This is why risk, and return, are not correlated empirically (see videos here, read book here).
I personally have known a lot of really smart people and have to say they are more unconventional in their ideas, yet most of their ideas are crazy. If you have ever been to a Mensa meeting (IQ but little formal education), you realize how things like homeopathy, or truthers, get their bearings. If you have ever hung out with PhDs, you know how limited their competence scope is (at research universities they have the same IQ as Mensans, but are more disciplined and less creative). It's no wonder guys like stereotypical MBAs, who are not so analytical but rather personable and articulate, tend to dominate society. I suspect MBA rule is less catastrophic than PhD or Mensa rule, if only because they aren't as certain of themselves. This all gets back to the idea there is an optimal IQ, and it's not 180, but rather, say, 125 (probably the modal IQ for any large group leader, such as Presidents and CEOs).
Being smart is a good thing, and I'm happy when my kids do well on cognitive tests because of what this portends for their life (as Charles Murray noted, most people would prefer their kid had 15 more IQ points than get $1 million on their 21st birthday). Yet highly intelligent people tend to innovate more, and such innovation tends to be counterproductive for the innovator. So, the fact really smart people can answer a question faster or more accurately than others, is at some point offset by the fact that when they have to supply the question--as is the case once they leave formal schooling--they will be attracted towards less conventional, usually irrelevant or wrong, paths. For every Steve Jobs or Albert Einstein there were many who lived and died in obscurity; for every Black-Derman-Toy there are hundreds of insanely convoluted, in-house models, of no value.
So, if you attempt to innovate, take pride that you are doing it with blatant disregard for your narrow self interest. Objectively, you will fail with a high probability. Like thinking about the lottery, there are happy delusions of grandeur considering your improbable future success. You can counter those who say, you should be feeding orphans in an African village, in that your efforts have large positive externalities. Most importantly, it's fun.
Tuesday, November 03, 2009
RiskMetrics Goes Corporate
I always have admired RiskMetrics, because they offer straightforward advice about how to measure risk. My career as a risk manager started as an attempt to implement the then-new (1994) RiskMetrics Value-at-Risk methodology to the KeyCorp trading operations. Sure, for any nuanced strategy, they are 'academic', in that they do not understand a lot of parochial issues related to, say, risk arb, but for setting a benchmark, understanding certain tools are useful, and being clear, I give them an A+.
Thus, I was saddened to see their acquisition of the KLD Research & Analytics, "a leader in environmental, social and governance (ESG) research and indexes for institutional investors", because it highlights the essence of Risk Management as a corporate enterprise. KLD's politically correct indices are lame, allowing companies to justify holdings to investors via moral preening. Having the imprimatur of some socially conscious busybody just says you are trying to get by on good intentions as defined by conventional wisdom.
Who's to say if investing in, say, Al Gore's latest initiative is morally superior, better for the world's people, than giving more money to Exxon? After all, Gore's senate tie breaking vote back in 1994 started this corn-based ethanol boondoggle, which is now an entitlement that will take decades to undo. Meanwhile, farming for ethanol takes more energy than it produces when you add up all the indirect cost (as any good economist should). Further, it raises the price of corn, and hurts natural aquifers. In contrast, Exxon produces oil that can then be used by centralized power plants or machines, instead of burning biomass, which is much less efficient.
That RiskMetrics thinks moral posturing complements their activities highlights how senior risk management as a profession is mainly about appearances and going through the motions. To be specific, official risk management, which is what RiskMetrics now advises upon, is mainly staffed by prigs and props. A prig is someone wedded to a theory that explains everything, and can be highly mathematical and conventional, like Phillipe Jorion, or highly obtuse, like Nassim Taleb's insight that 'risk is what you don't expect' [which I find not profound, but irrelevant]. The key is they are true believers, have credibility as experts, and are true idiot-savants: they have a lot of knowledge, and a lot of ignorance.
Then there are the props, the suits who actually run senior risk management posts, and they basically have a full time PR job, catching flak from outsiders about decisions not made in their presence. Official Risk Management futility comes from the fact they are graded on their ability to generate reports read by people three levels removed from a business, so that why the props need the prigs, for references. The prigs need the props to add credibility to what they do, though I think it would be most accurate to say there's some truth in what prigs do, but only if you know where to find it. Nuance, the things that really matter for a businesses survival, don't travel up the chain well, so everything is fit into cookie cutters based on the most recent conspicuous financial failures.
I wonder if the RiskMetrics management has convinced themselves this really addresses risk, or they realize it's simply the way the game is played, and they have to make a living. I'm too cynical to pull it off convincingly. I'm saddened by the whole thing because I understand a lot about actual risk management, but realize how irrelevant that knowledge is to risk management as practiced by large, regulated firms.
Macro, Risk Management. These are two fields I know enough about to avoid.
Thus, I was saddened to see their acquisition of the KLD Research & Analytics, "a leader in environmental, social and governance (ESG) research and indexes for institutional investors", because it highlights the essence of Risk Management as a corporate enterprise. KLD's politically correct indices are lame, allowing companies to justify holdings to investors via moral preening. Having the imprimatur of some socially conscious busybody just says you are trying to get by on good intentions as defined by conventional wisdom.
Who's to say if investing in, say, Al Gore's latest initiative is morally superior, better for the world's people, than giving more money to Exxon? After all, Gore's senate tie breaking vote back in 1994 started this corn-based ethanol boondoggle, which is now an entitlement that will take decades to undo. Meanwhile, farming for ethanol takes more energy than it produces when you add up all the indirect cost (as any good economist should). Further, it raises the price of corn, and hurts natural aquifers. In contrast, Exxon produces oil that can then be used by centralized power plants or machines, instead of burning biomass, which is much less efficient.
That RiskMetrics thinks moral posturing complements their activities highlights how senior risk management as a profession is mainly about appearances and going through the motions. To be specific, official risk management, which is what RiskMetrics now advises upon, is mainly staffed by prigs and props. A prig is someone wedded to a theory that explains everything, and can be highly mathematical and conventional, like Phillipe Jorion, or highly obtuse, like Nassim Taleb's insight that 'risk is what you don't expect' [which I find not profound, but irrelevant]. The key is they are true believers, have credibility as experts, and are true idiot-savants: they have a lot of knowledge, and a lot of ignorance.
Then there are the props, the suits who actually run senior risk management posts, and they basically have a full time PR job, catching flak from outsiders about decisions not made in their presence. Official Risk Management futility comes from the fact they are graded on their ability to generate reports read by people three levels removed from a business, so that why the props need the prigs, for references. The prigs need the props to add credibility to what they do, though I think it would be most accurate to say there's some truth in what prigs do, but only if you know where to find it. Nuance, the things that really matter for a businesses survival, don't travel up the chain well, so everything is fit into cookie cutters based on the most recent conspicuous financial failures.
I wonder if the RiskMetrics management has convinced themselves this really addresses risk, or they realize it's simply the way the game is played, and they have to make a living. I'm too cynical to pull it off convincingly. I'm saddened by the whole thing because I understand a lot about actual risk management, but realize how irrelevant that knowledge is to risk management as practiced by large, regulated firms.
Macro, Risk Management. These are two fields I know enough about to avoid.
Geanakoplos' Theory is No Paradigm Shift
The WSJ has a Malcom Gladwellesque article on John Geanakoplos as an iconoclastic genius who topples staid conventional wisdom. These are popular because we like the idea, but alas, what makes for a good read is often not true, not that most readers really care (who goes back to check which alarmist Time Magazine cover stories turn out accurate?).
A good way to assess someone's philosophy is to see what it implies. If you read John Geanakoplos's "Solving the Present Crisis" paper, we see the following insights and proposals:
First, he notes financial crises are typified by periods of easy lending, followed by periods of hard lending. The boom is manifested by easier terms (less collateral required for the lender as a percentage of the loan amount) and lower rates. This is hardly revolutionary. There would have to be some greater specificity on that mechanism to be useful.
The fact that banks offer greater leniency via two factors--the loan rate, and the amount of collateral required--is not really that radical. Many goods and services are multidimensional, such as when coffee is served in a nice cup or with free half&half. If you then say, if you give more leverage to 'natural buyers', they will push equilibrium prices above their 'true value', you need some metric of identifying this deviation ex ante. Everyone knows all things housing related pre 2007 were overpriced, and I don't remember Geanakoplos ringing the bell that some metric of overpricing was happening in real time. I don't see any metric here one can use to identify these situations, which makes sense because it implies an absence of arbitrage (there are no $20 bills on the ground).
But consider his solution to our crisis last year, his paper had three prongs.
1) Put a floor under housing prices.
His mechanism is unimportant, but highlights he is a true macroeconomist: he looks at aggregate prices, not relative prices. The problem, he notes, is that too much lending for housing occurred, so I don't see how keeping prices out of line fixes anything but the symptom, and delays the inevitable adjustment.
2) Provide easier lending for mortgage backed securities.
I don't have a problem with this. As I argued in the April about the PPIP Treasury plan that Stiglitz said was a massive give-away, but was eventually shelved for a total lack of interest: if you allow modest financing, the 'put option' you write is not so valuable, and so it is not a give-away. But, given all the extra obligations that come with having Uncle Sam as your investing partner, it probably is not worth it. The plan won't help, but it won't hurt.
3) Put equity into financial organizations, and have the government get involved in management decisions, and getting banks to lend more.
This is a disaster. Look at which financial institutions are pegged to have the largest losses: Fannie and Freddie. They have the dual objective of doing good, as defined by politicians, and making money. Those objectives don't work well together. Note how once they took over IndyMac they immediately stopped foreclosure, a classic populist approach to banking that is not viable long term.
For what is presented as a new paradigm, it sounds a lot like a standard Democratic talking points on the crisis. Not that this is necessarily bad, just that there's nothing really interesting there that's new or clever.
No Money Down Mortgages Continue
It seems many people are taking advantage of two major programs at the low end of the housing bubble: the FHA is aggressively promoting lending with only 3.5% down, and the $8k tax credit for buying a house less than $200k. A good realtor can apply the tax credit to last years taxes, making sure that the buyer actually gets the money right away, and the HUD is actually OK with using the $8k to cover the down payment. Remember horror stories of sellers who would pay the slim down payment or closing costs, and leave the stupid investors with losses? Well, today that game is over, except for the government, which proves that stupidity in the private sector actually loses people money, causing them to change their ways. For the government, it's just more incentive to double down.
A person who can't afford a down payment should not be in a home. One needs capital to pay for routine maintenance, and most importantly, if something major happens, like if a heater breaks. A renter is someone who does not have the wherewithall to handle these large, unanticipated expenses. It is better for everyone if these people are renters, because otherwise a bad break leaves the property in poor shape, leading to a 'broken windows' problem.
No private bank would lend in such a manner, but FHA wants to take the risk, because unlike the greedy bank, it sees the 'bigger picture', presumably.
The government's program reminds me of the technique children independently discover to make it look like they've eaten up hated peas or carrots: spread them around the plate. Thus, the recent economic debacle is primarily centered on housing, especially lower-end housing financed by overeager lenders. The unavoidable endgame to this problem is fewer houses, and lower prices for those houses; demand was artificially high. But, we wouldn't want people to adjust, because every good Keynesian knows that all misallocations of resources in a complex economy can be solved via top down injections o fiat money, or "G"--they have the multipliers to prove it!
To the government every hangover needs a little more hair of the dog. A friend of mine says his rental real estate business is slow at the top end, because those renters are attracted to the FHA loan/$8k tax credit. Those kind of effects are basically unmeasurable, and what can't be measured is not counted. What about the complex dynamics implicated by the finite nature of the $8k tax credit? Preventing the pain with a temporary subsidy merely prolongs the amount of time spent in the doldrums (eg, 1933-39 was a period of very high unemployment).
Bad governments prioritize the seen over the unseen, the direct over the indirect. The sad fact is popular policies usually have highly concentrated benefits and highly distributed costs, and the politicians act like little children, hoping those watching do not notice stuff that's spread around.
Sunday, November 01, 2009
The Theory of Relativity, Updated
Watch this video on YouTube...the woman has this funny theory. She starts out asking, you do know what H2O is? You have heard of Einstein. Basically she notes
1) E=mc^2
2) As the diameter of protons, neutrons is about 1E-15 m, while the diameter of the atom is 1E-11. Thus, all the mass in the galaxy would fit into a small ball.
3) Thus, the mass being small, we can ignore it, and so E=c^2.
This plays into her theory that we are all mere energy, which somehow means homeopathy works.
Only using mere words can one can make this mistake.
1) E=mc^2
2) As the diameter of protons, neutrons is about 1E-15 m, while the diameter of the atom is 1E-11. Thus, all the mass in the galaxy would fit into a small ball.
3) Thus, the mass being small, we can ignore it, and so E=c^2.
This plays into her theory that we are all mere energy, which somehow means homeopathy works.
Only using mere words can one can make this mistake.
Adult Kill Joys in Action
From the New York Times:
And in keeping with the theme of healthy eating inside the White House that Mrs. Obama has promoted, the children were also given a dried fruit mix of cherries, apricots, pears, apples and papayas.
I bet the kids were thrilled to get dried papayas.
Friday, October 30, 2009
Happy Halloween
Halloween is an awesomely fun holiday. I like it. My kids like it. And young women use it as a temporary insanity defense for acting slutty, much to young men's benefit.
Too bad my kid's elementary school canceled Halloween costumes because the new Somali immigrants find it offensive to their sensibilities.
Thursday, October 29, 2009
The Scientific Conceit
This interview with David Berlinski contains this precious observation:
The idea that science is a uniquely self critical institution is of course preposterous..scientists are no more self critical than anyone else, they hate to be criticized and never criticize themselves...There are local mechanisms of criticisms in science, within established theories if somebody publishes data that don't work out in a certain way, if there are mathematical flaws in a certain theory, these tend to get know, but large global criticisms of the scientific enterprise are very difficult to find, and certainly not being promulgated by the scientists with any ebullience or enthusiasm ... these people are only human, they hate criticism--me too! The idea that scientists are absolutely eager to get beaten up that's one of the myths, put out by the scientists, and it works out splendidly so that they can avoid criticism.
You know a naive or tendentious science writer when they start talking about how science is so different than other professions in how they objectively present their work for criticism. The journal publication process does filter out a lot of errors and unsubstantiated assertions that a journalist might get away with, but that's really a very narrow domain, it's like noting a New Yorker piece is meticulously checks for grammar (unlike my blog!). It's ruthless criticism in a very specific domain.
Look at Freakonomic's author Steve Levitt's work on the abortion-crime link. No referee thought, gee, how does this relate to the different black-white abortion rate relate to the difference in Black-White crime rate over the next 20 years? How did that relate the crime rate differential between 17 year olds and 35 year olds, 17 years after Roe? How does the small difference in fertility post Roe relate to an selectivity effect (clearly, as abortions went up, so did conceptions)? What if you adjusted for population growth? All of these are large, glaring points against Levitt, really common sense type rebuttals, yet he never had to address them because he presented a panel regression with interaction terms.
I have refereed papers with interaction terms, and they are almost always garbage, because they generate a lot of collinearity, and the nonlinear correlations manifest themselves in a bunch of significant coefficients in linear regressions, because if the 'true' relation is y=x^2, and x=A+2B+e, then a regression on y with A and B will have a positive coefficient on A, and a negative on B. Levitt's study has state*age, year*age, and state*year as explanatory variables. Why stop there, why not state*year*age? With tens of regressors, many product terms, you get garbage. Yet, I've refereed reports from professors at Harvard with this stuff, so it's not something that's necessarily wrong, just practically stupid. There has been no important result that shows up only via interaction terms in regressions, just as there has never been an important relationship evinced solely via 3-stage least squares, or the Generalize Method of Moments (GMM). Ever. Rather than correcting error terms for heteroskedasticity or something inside-the-box, there should be more skepticism applied to these kitchen-sink approaches.
But, Levitt is still considered a top-level researcher, and he has a very thin skin. For example, reading his coauthor's response to criticism of his Global Warming chapter in their new book, the tone seemed very defensive, like someone unused to criticism. In sum, Levitt, like most scientists, is exposed to a very narrow set of criticisms, ones that most laypeople could not counter to be sure (must get one's standard errors correct), but that's really no different than the fact that most people (me included) could not write prose for the New Yorker with their syntax errors. In most ways he is unexposed to real criticism and acts accordingly.
Wednesday, October 28, 2009
Obama's Alpha Delusion
From the WSJ:
The Obama administration launched a clean-energy blitz Tuesday, with President Barack Obama sweeping into this Central Florida hamlet to unveil $3.4 billion in stimulus grants for advanced electricity-grid projects
This PR parade relies on the idea that this administration, if not Obama himself, gets into details, and chooses the right cutting edge technologies and methods. Look at Obama above, with his sleeves rolled up, giving pointers to an appreciative bunch of field managers (perhaps the NEA can get to work on some Soviet Realism in this context). In this case, Obama merely has to allocate some of our money to a select list of projects that are aligned with the buzzwords 'clean energy', and we get the increasing returns to scale that Paul Krugman won his Nobel Prize for (too bad Ann Krueger didn't win a Nobel for showing the same 'infant industry' argument has been a pretext to protect inefficient industries for over 200 years).
It never occured to any of these guys that there aren't any magic solutions to our energy problem. They act as if we only tried to develop batteries, we could have ten times the power. See this video from Zocalo, and at the end of the critical discussion about the oil industry an audience member earnestly asks: "can't we develop energy out of water?" as if the only reason we use oil is because the Rich Uncle Pennybags character from the Monopoly Game is not letting us do anything else. The electric car predates the internal combustion engine. My laptop and cell phone routinely run out of energy, highlighting the high reward waiting for the next battery innovation. There has been and continues to be research, and incentives, to increase the efficiencies of batteries.
Obama hates being compared to socialists, so I'll refrain and compare him to a communist. In the state published hagiography, Divine Stories About the Dear Leader, Kim Jong-Il is presented as someone excellent at golf, pistol shooting, technology, and battlefield courage. He's basically better than everyone at everything. For a communist state that belief is necessary, otherwise their system is too centralized.
Obama and his experts are presumably more efficient than the market at allocating more resources to productive technologies. The idea that since the market won't provide funds, perhaps the informed expected return on battery investment is truly low, seems absurd: how could selfish oafs who run business know better than an articulate, caring, public servant? It's The Secret writ large: think it true, and it becomes so. No wonder it's a popular idea: would that it were true.
Unfortunately, the bien pensants who adore Obama (or really, adore that they adore Obama), see his value add being multifaceted micromanagement. There are countless $3.4B special investment targets to do, each one with dreams of cold-fusion, high-speed trains, and the end to the achievement gap. Most people think that 'good smart people' are better in almost every way than your average businessman, and most people think they vote for such people, thus these politicians should be directing activities the way a coach directs a football team.
Alas, the value of extreme intelligence and knowledge of detail, does not scale at the managerial level. It runs out of benefit to a ruler, because they cannot and should not try to micromanage things. Thus, the best developer of a new technology is often a lousy director for a state or large corporation, and the best managers are often not the best developers. Indeed, a key advantage of those who are smart—but not too smart—is they know they don't know more than everyone. The Barak Obamas and Paul Krugmans, having excelled at Harvard or MIT, can more easily think they actually know more than everyone else, leading to the classic Fatal Conceit of planners everywhere.
The idea that the only feasible alpha for a leader of a large collective, is to enforce rules and get out of the way, is simply preposterous for those who think the Invisible Hand is merely a theory used by conservatives to excuse their indifference. This reflects a failure to appreciate the complex, homeostatic mechanisms of self interested agents within a free market, and the infinite number of ways top-down rules are worked around when applied to the masses. As Hayek noted, the biggest flaw with the free market is that it wasn't designed, it emerged spontaneously, which causes people to dismiss its value. Thus, they have 1000 page plans like our health care bill, or ideas about new committees that will assess issues intelligently and disinterestedly.
Tuesday, October 27, 2009
What Does Something Really Cost?
Any large institution has problems allocating 'fixed costs'. That is, I remember going over budgets, and certain functions, like headquarters, the CEO's salary, Treasury, was a 'fixed cost'. More problematically, you had many sunk costs, businesses we were exiting, that were obviously losing money (otherwise we wouldn't be exiting them). Each individual unit could be making money, but once we paid for the fixed costs and costs of the businesses we were winding down, the corporation was underperforming on a Return-On-Equity basis. As a corporation, you have to allocate all your costs to see if you are doing right by your shareholders.
For a government, it is much worse, because they do not have a balance sheet. Very few in government know how much anything costs. They just pass the law, and look at the marginal expenditure. Like the Social Security Trust fund, they operate using rules that would be illegal if done in the private sector. The market is a heartless thing, but at its core it makes sure that when people get more out of what they put in (profits), they do more; if they get less (losses), the do less or not at all. That's efficiency.
So when a private study reported that each Amtrak passenger costs taxpayers(ie, over the passenger payment) $32, four times Amtrak's estimate of $8. Oops. Further, the subsidy ranged from $4 (Boston to Washington DC) to $462(LA to New Orleans). Clearly, the higher costing routes seem ripe for exit, but it would be naive to think that is being considered.
Ludwig von Mises said the main reason capitalism would outperform socialism, is because in socialism no one knows how much anything costs, which makes it impossible to allocate resources efficiently. The Amtrak study highlights they don't know, and they don't even think it matters.
Encouraging a frat party atmosphere?
From today's WSJ:
Shocking.
A former Anheuser-Busch Cos. executive who for years served as the company's public voice against critics has sued the beer giant for discrimination, saying she was paid less than male executives and that the company encouraged a "frat party" atmosphere.
Shocking.
Sunday, October 25, 2009
McDonald's Alpha Deception
I would estimate 90% of all alpha is misrepresented. Anyone in charge of a business line making money, is usually too embarrassed by the straightforward nature of their advantage to admit it, so they have to point out some nuance that makes absolutely no difference. Thus, every market maker, making money off order flow, will swear they are adding value 'reading the tape' or trading like a turtle, or some other such nonsense. Finance is probably the worst, because there's so little alpha and so much branding and 'sticky money', that truth-telling is a strictly dominated strategy. If you ask your average financial executive to explain what he does, chances are he won't tell you even if he knows. Further, many are actually clueless. They don't know their job is to provide the appearance of a method to the whims of the main decision-maker, that they fit the right diversity box, or their husband is a senator. Admitting the truth would be too depressing, and the mind is very good at protecting its self image.
Thus, it's fun to see market leader McDonald's brought down to the level of a Jim Cramer. I like McDonald's: it's clean, I like the burgers and fries, my kids enjoy their play areas and have a fairly nutritious lunch (hamburger with apple slices and milk). But their burgers tend to lose adult taste tests against Burger King. Why? McDonald's burgers are primarily loaded with ketchup, which appeals to kids, where BK has more mayo, which appeals to adults. The solution might seem easy, add an option to replace ketchup with mayo. But that would make the burger choice seem much less alpha-like. A burger chain has a reputation, and they carefully project one of having super quality and care, or something outside the box like a square shape, or flame broiling. Heaven forbid they state, these are hamburgers, not steak. They are cooked by people who have trouble remembering to wash their hands after using the bathroom (thus the prominent signs), let alone the ordinal ranking of rare, medium, and well-done. A multinational corporation can't produce a medium rare burger without generating a class action E. coli lawsuit, and a well-done piece of ground beef is about as nuanced (yet still enjoyable), as an ice-cold light beer.
But that's like a finance professor saying all investment analysts can't predict the market. A thriving industry goes on, acting as if they have alpha in every 'buy' recommendation, every burger. Thus, the newest McDonald's creation are their new Angus burgers. They have...lots of mayonnaise. Too much in fact. So, even though they know this is the true 'secret sauce' in the adult burger battle, they emphasize the Angus dimension, and then overload the key ingredient. I prefer the more predictable double quarter pounder with no pickle.
Taleb Confronts His Critics
The ever amusing Nassim Taleb has penned yet another response to his critics. He simply oozes defensiveness, which combined with his arrogance, strong opinions, and popularity, makes him incredibly fun to write about (if you haven't been accused of WEB VANDALISM by NNT, you are missing out).
So, here's his new summary, which he notes parenthetically, "I have had to repeat continuously" (perhaps his emphatic reassertion is not compelling?). Anyway, he notes that "theories fail most in the tails; some domains are more vulnerable to tail events." I agree. Newton's theories don't work at the Plank length, cosmology has trouble explaining the first minute of the universe, and evolutionary biologists have trouble explaining the Cambrian explosion 500MM years ago. Explaining most of the data is easier, and generally more important.
Now, one may protest, it's not a new point. However, Taleb argues that "nobody has examined this problem in the history of thought", highlighting the common problem of autodidact philosophers, that they tend to be too dismissive of the scientific literature. He notes what he is not talking about includes most everything that is directly related to extreme events and uncertainty: falsification, power laws, Hume's problem of induction, Knightian Uncertainty, Austrian Uncertainty, integrating fat tails into models, etc. He's aware of these arguments, but claims his idea is different
His big twist: that rare events can't be estimated, because they are rare, especially, when they are rare and have large impacts. Well, I would argue this issue is addressed in the literature he notes is unrelated to this point, as long disquisitions on the difficulty in estimating an event like WWI, or standard errors on order statistics, seems like the same subject to me. He can say these earlier discussions are flawed, but they address his key point. The more he explains himself, the more he sounds like some passage from Knight, Keynes, Hume, Minsky, etc. once you translate his neologisms (historia, ludic fallacy). I would argue Taleb is much less clear than these writers, because he's trying so hard to make the old sound new. Saying it's really new doesn't make it so.
He ends with a strong plea to not be wedded to a theory, to look at the facts and avoid fitting them into a preconceived theory. I'm sure the vast 'preconceived, untestable theory' crowd has a lot of soul searching to do. Yet, I think he's the best example of their kind. He has a theory: that he's saying something really 1)new, 2)true, and 3)important. He says many things, often contradictory, but he never manages more than 2 of those attributes in any assertion.
Thursday, October 22, 2009
The Fed's Latest on Risk Management
The Federal Reserve houses probably the brightest, most thoughtful bureaucrats in government. Many like to get published and participate in academic conferences, so their selfish attempts to raise their status in the economics guild has positive externalities.
I bet most of your average risk manager's job is not about measuring, monitoring and reporting, for decision-making, but for appearances: to investors, regulators, analysts. Most risk managers don't get invited to the 54th floor corporate board room until regulators appear, then magically they are praised to high heaven. Often this includes obtusely patronizing remarks about how smart they are by senior management in the face of these outsiders, as if, hiring some cast-offs from Los Alamos clearly means there are no problems here! You can't fool people who know lots of math! Bottom line: there are lots of clueless, but high IQ/Education risk managers out there in every firm, which clearly should comfort no one.
The Fed's latest opus, done in conjunction with the G-12ish central banks, has created the Senior Supervisors Group Issues Report on Risk Management Practices, in the process highlighting that risks are still with us. Nothing it says is really untrue, but nothing is new to anyone with a little experience in the field. The specifics are very relevant to the subprime crisis, as if no-money-down mortgage pool CDOs are incipient. It's kind of like the airplane regulations: as if anyone could get away with 9/11 with mere box cutters today (the prospect of imminent death would create an avalanche of crazed vigilantes; pre 9/11 everyone thought it would be a layover in Cincinnati).
We learn that
I doubt there's a banker with rank of VP or above who does not understand that now. Hindsight's pretty good everywhere.
Then there's the
Always has been, always will be. A full-time risk manager is like a figure skater. The best are well compensated, the other 99% are obscurities with status and power about the same as you average LAN administrator.
Alas, this is optimal! Consider a full time risk manager is paid to keep bankers honest, from taking too much risk. They are preventing people from taking on new business, because the risk of default is not 2%, but 4%. Now, such a warning is very difficult to validate, the power of any test to validate is beyond your average risk manager's business life in his current occupation. Thus, as quantitative as risk management is in theory, in practice it is very non quantitative, because the big risks presented by crises happen so infrequently.
This invites a lot of posturing. Many top risk managers are ex-regulators, esteemed academics, or have fancy degrees. When you can't measure the output, you measure inputs. The bottom line is and always has been the degree to which those directly affecting revenue, the business line managers, accurately amortize the expected losses of any capital investment. Their long run success depends on that outcome, which is often binary: heads they win, tails they are fired.
Of course, at the top, these executive's success is less dependent on actual success, because these people are managing people who manage businesses, and so, like Robert Rubin, you can make $100MM without actually knowing he had a lot of mortgage paper on his balance sheet (details!). That's an independent issue, why the managers of managers get paid so much (I'm think there's room for improvement here). For the business line, the guy actually creating the business, at least 3 layers below the CEO, his risk management is essential.
In contrast, the full time risk manager is a dilettante. He does cursory reviews of perhaps 20+ different business lines and so is clueless to the real issues, because anyone spending 1/20th the time on something you do that actually has alpha, and delivers profits (ie,is actually valuable, and so not obvious), cannot be understood with such minimal focus. Yet, he's the guy you show to regulators, or investors, when talking about risk management. Reality is very decentralized, and its a convenient fiction to think that risk can be centralized and managed by the Board, or someone not working in the business day in and day out. Clearly this problem is only worsened if we think about delegating risk management to regulators, who are even further removed.
I think it's a fiction to think one can meaningfully present the risk of any collective via a concise table of numbers. A senior executive should emphasize they prioritize the validation of expected loss forecasts within each business line (at KeyCorp, we had 130 lines of business, and many of these were composites), by obligor (counterparty) rating and collateral (eg, secured by property, or unsecured?). They should note new activities have extra layers of cushion applied to these expected loss estimates. They should then present a set of examples of how risk is broken down in a particular business line (say, indirect auto lending), including actual and expected loss rates by as much granularity as possible (crosstabbing by 5 risk grades, 3 collateral types). They should also highlight any changes to the methodology, because innovation involves change, and data will not be available; that is understandable. The changes should be based on some kind of theory or analogue, or story. This helps the outsider understand why they are doing this, and how it can be validate (ie, why shouldn't people make a down payment? Because house prices always rise!).
They should then invite the analyst to ask for another example, based on the business or product of their choosing, which would imply each line was ready to present their risk. The request would then entail someone from, say, Media Lending, to come up, and explain how they slice up risk, how they validate their loss forecasts by the granularity they present (including lines, loans, and letters of credit), and how pricing, revenue sharing with cash management, and costs of funds, relate (this can highlight conflicts of interest). The presentation should be amenable to a 30 minute presentation; if that is not possible, then clearly they do not have it under control. By letting the outsider choose, they can be confident their questions would be answered similarly if they did this on another business line.
They should note that no one gets a bonus for revenue generated the prior year, but rather, as that revenue is amortized over the life of its duration. If that's a day, fine, but no one walks away making a bonus off capitalized revenue that has not yet occurred.
Risk in any diverse financial organization cannot be summed up by a third party into a scalar. The essence of risk is like the essence of productivity: the parochial knowledge, processes, and incentives of very diverse activities. This is necessarily a detail oriented issue, and the big risks are bad assumptions, not bad math. That is, it was not copulas, or correlations, that screwed up subprime, but the assumption housing prices, in aggregate, would not fall. That's a bad assumption, based on understandable but flawed logic. You cannot appreciate these bad assumptions merely by adding up all their flawed implications and comparing them to total bank capital.
I bet most of your average risk manager's job is not about measuring, monitoring and reporting, for decision-making, but for appearances: to investors, regulators, analysts. Most risk managers don't get invited to the 54th floor corporate board room until regulators appear, then magically they are praised to high heaven. Often this includes obtusely patronizing remarks about how smart they are by senior management in the face of these outsiders, as if, hiring some cast-offs from Los Alamos clearly means there are no problems here! You can't fool people who know lots of math! Bottom line: there are lots of clueless, but high IQ/Education risk managers out there in every firm, which clearly should comfort no one.
The Fed's latest opus, done in conjunction with the G-12ish central banks, has created the Senior Supervisors Group Issues Report on Risk Management Practices, in the process highlighting that risks are still with us. Nothing it says is really untrue, but nothing is new to anyone with a little experience in the field. The specifics are very relevant to the subprime crisis, as if no-money-down mortgage pool CDOs are incipient. It's kind of like the airplane regulations: as if anyone could get away with 9/11 with mere box cutters today (the prospect of imminent death would create an avalanche of crazed vigilantes; pre 9/11 everyone thought it would be a layover in Cincinnati).
We learn that
Some firms’ business models also relied on excessive leverage.ORLY? Would have been nice to read about that pre 2006.
Firms also failed to realize that two important sources of funding, securities lending and money market funds, could impose further demands on firm liquidity during periods of stress.
I doubt there's a banker with rank of VP or above who does not understand that now. Hindsight's pretty good everywhere.
Then there's the
the stature and influence of revenue producers clearly exceeded those of risk management and control functions.
Always has been, always will be. A full-time risk manager is like a figure skater. The best are well compensated, the other 99% are obscurities with status and power about the same as you average LAN administrator.
Alas, this is optimal! Consider a full time risk manager is paid to keep bankers honest, from taking too much risk. They are preventing people from taking on new business, because the risk of default is not 2%, but 4%. Now, such a warning is very difficult to validate, the power of any test to validate is beyond your average risk manager's business life in his current occupation. Thus, as quantitative as risk management is in theory, in practice it is very non quantitative, because the big risks presented by crises happen so infrequently.
This invites a lot of posturing. Many top risk managers are ex-regulators, esteemed academics, or have fancy degrees. When you can't measure the output, you measure inputs. The bottom line is and always has been the degree to which those directly affecting revenue, the business line managers, accurately amortize the expected losses of any capital investment. Their long run success depends on that outcome, which is often binary: heads they win, tails they are fired.
Of course, at the top, these executive's success is less dependent on actual success, because these people are managing people who manage businesses, and so, like Robert Rubin, you can make $100MM without actually knowing he had a lot of mortgage paper on his balance sheet (details!). That's an independent issue, why the managers of managers get paid so much (I'm think there's room for improvement here). For the business line, the guy actually creating the business, at least 3 layers below the CEO, his risk management is essential.
In contrast, the full time risk manager is a dilettante. He does cursory reviews of perhaps 20+ different business lines and so is clueless to the real issues, because anyone spending 1/20th the time on something you do that actually has alpha, and delivers profits (ie,is actually valuable, and so not obvious), cannot be understood with such minimal focus. Yet, he's the guy you show to regulators, or investors, when talking about risk management. Reality is very decentralized, and its a convenient fiction to think that risk can be centralized and managed by the Board, or someone not working in the business day in and day out. Clearly this problem is only worsened if we think about delegating risk management to regulators, who are even further removed.
I think it's a fiction to think one can meaningfully present the risk of any collective via a concise table of numbers. A senior executive should emphasize they prioritize the validation of expected loss forecasts within each business line (at KeyCorp, we had 130 lines of business, and many of these were composites), by obligor (counterparty) rating and collateral (eg, secured by property, or unsecured?). They should note new activities have extra layers of cushion applied to these expected loss estimates. They should then present a set of examples of how risk is broken down in a particular business line (say, indirect auto lending), including actual and expected loss rates by as much granularity as possible (crosstabbing by 5 risk grades, 3 collateral types). They should also highlight any changes to the methodology, because innovation involves change, and data will not be available; that is understandable. The changes should be based on some kind of theory or analogue, or story. This helps the outsider understand why they are doing this, and how it can be validate (ie, why shouldn't people make a down payment? Because house prices always rise!).
They should then invite the analyst to ask for another example, based on the business or product of their choosing, which would imply each line was ready to present their risk. The request would then entail someone from, say, Media Lending, to come up, and explain how they slice up risk, how they validate their loss forecasts by the granularity they present (including lines, loans, and letters of credit), and how pricing, revenue sharing with cash management, and costs of funds, relate (this can highlight conflicts of interest). The presentation should be amenable to a 30 minute presentation; if that is not possible, then clearly they do not have it under control. By letting the outsider choose, they can be confident their questions would be answered similarly if they did this on another business line.
They should note that no one gets a bonus for revenue generated the prior year, but rather, as that revenue is amortized over the life of its duration. If that's a day, fine, but no one walks away making a bonus off capitalized revenue that has not yet occurred.
Risk in any diverse financial organization cannot be summed up by a third party into a scalar. The essence of risk is like the essence of productivity: the parochial knowledge, processes, and incentives of very diverse activities. This is necessarily a detail oriented issue, and the big risks are bad assumptions, not bad math. That is, it was not copulas, or correlations, that screwed up subprime, but the assumption housing prices, in aggregate, would not fall. That's a bad assumption, based on understandable but flawed logic. You cannot appreciate these bad assumptions merely by adding up all their flawed implications and comparing them to total bank capital.
Wednesday, October 21, 2009
Is Levitt a Global Warming Denier?
Freakonomics was a highly popular book that appealed to both liberals and conservatives. Therefore, it carefully avoided polarizing topics, and instead uncovered the shocking truth about sumo wrestlers and other issues that are worthy of a standard 20/20 television show. Fun stuff, not what I would call economics (see the more esteemed economist Ariel Rubinstein for support).
So, this time they figured they would slay some fallacies in the Global Warming debate. They bend over backward to apply good faith to Global Warming proponents, and agree with many of it's propositions(it is not a singular hypothesis), yet try to have fun with some issues that appear ripe for debunking (eg, noting that horses generate more pollution than oil as an energy source). Unfortunately, the Global Warming Community does not approve of their shenanigans. They have too start action, now, and these issues hurt the cause. Levitt seems to like being against conventional wisdom only on areas where there are very few opinions, so he and his coauthor weaken their case by protesting too much, trying to have it both ways (I love the Weitzman argument that since a catastrophe could happen, we should spend trillions of dollars on it--it can be applied to anything, and indeed, he has used it to explain the equity premium puzzle).
Unfortunately, alternative energy sources that are currently most viable, like cleaner coal, or nuclear, are not popular with the Global Warming crowd. Even windmills, and solar, are coming into opposition for their noise or eye pollution. The only thing they really like are pie-in-the-sky battery research, and conservation. I think this highlights that most of this debate is not about Global Warming, but more power to regulate, because it adds another busy body to approve all sorts of things (like when I have to get permission from my city when I replace my old back patio with a new one).
The Global Warming debate is like many Big Issues. They are multifaceted, so debunking a point hardly makes a difference because most people's opinion has several several pillars. Given it will take my lifetime to provide any conclusive data one way or another, I don't expect this one to subside.
So, this time they figured they would slay some fallacies in the Global Warming debate. They bend over backward to apply good faith to Global Warming proponents, and agree with many of it's propositions(it is not a singular hypothesis), yet try to have fun with some issues that appear ripe for debunking (eg, noting that horses generate more pollution than oil as an energy source). Unfortunately, the Global Warming Community does not approve of their shenanigans. They have too start action, now, and these issues hurt the cause. Levitt seems to like being against conventional wisdom only on areas where there are very few opinions, so he and his coauthor weaken their case by protesting too much, trying to have it both ways (I love the Weitzman argument that since a catastrophe could happen, we should spend trillions of dollars on it--it can be applied to anything, and indeed, he has used it to explain the equity premium puzzle).
Unfortunately, alternative energy sources that are currently most viable, like cleaner coal, or nuclear, are not popular with the Global Warming crowd. Even windmills, and solar, are coming into opposition for their noise or eye pollution. The only thing they really like are pie-in-the-sky battery research, and conservation. I think this highlights that most of this debate is not about Global Warming, but more power to regulate, because it adds another busy body to approve all sorts of things (like when I have to get permission from my city when I replace my old back patio with a new one).
The Global Warming debate is like many Big Issues. They are multifaceted, so debunking a point hardly makes a difference because most people's opinion has several several pillars. Given it will take my lifetime to provide any conclusive data one way or another, I don't expect this one to subside.
Tuesday, October 20, 2009
2000th Anniversary of Teutoberg Battle
In 9 AD, my ancestors kicked some Roman butt, allowing them to avoid civilization for a couple more centuries. After all, what have the Romans ever done for us? The Battle of the Teutoburg Forest (called the Varian disaster by Roman historians) took place in A.D. 9 when an alliance of Germanic tribes led by Hermann the German defeated three Roman legions led by Publius Quinctilius Varus. There's a Hermann the German statue here in Minnesota.
There's a really cool video on the battle here, part of the History Channel's most awesome Decisive Battles series.
Monday, October 19, 2009
Focus on the Ordinary
A WSJ article notes that people are intrigued by those peculiar people who have HIV, but not AIDS. In cancer research, five Nobel prizes have been won by researchers who studied tumor viruses (3 for a chicken virus! the Rous virus). This started in 1911 when Peyton Rous discovered a virus that was found to cause tumors in chickens. The hope is that these special cases highlights the essence of the puzzle, and being a virus, leads to an easy isolation. Unfortunately, this thread has not proven very successful. Viral cancers are rare and not very relevant for human cancer, which makes sense when you consider cancer is not contagious. With a few exceptions cancer survival rates remain much of what they were in the 1950s. The Mayo Clinic reports that for cancers diagnosed from 1974 to 1976, the five-year survival rate was 50 percent, now 65 percent. This is mainly because of early detection, as opposed to any great new drugs.
A common idea is that outliers are more important than averages, as this is the them of Taleb's Black Swan, or Gladwell's Outliers. Predicting one great outlier is worth predicting many ordinary outcomes, so on one hand it seems like an optimal focus. Also, the outliers should highlight the essence of something. A stock that has risen 10 fold, or a great athlete, supposedly lays bare the essence of its greatness.
But I think we forget how biased our view is on exceptional events and people. We watch sports and learn about Usain Bolt, a most unusual man. Or my kids read the Guiness Book of World Records, containing stories about 1200 lb men and giant frogs. News is biased towards the exceptional, it takes no effort to emphasize it. In fact, it takes effort to see the ordinary. It's too bad people think of heroes as those who, for a brief moment, offered their life in some battle or harrowing situation, compared to the much more common heroism of providing one's family, not complaining, and being charitable to friends and neighbors, for decades.
A problem is that for any extremum is caused by its unique intrinsic characteristics and also random chance. We hope that we can more clearly see the intrinsic qualities associated with, say, a rising stock, by looking at sthose stocks that went up 10-fold last year. Yet, any extremum is probably a large random error. We simply can't predict the future very well, say an R2 of only 10%. Thus, a 100% stock return contains, on average, a 90% error. Yet, conditional up rising this much, this is only the mean error, its standard error is at least as large, meaning, any singular 100% return is probably all randomness. In this case, more analysis is worthless, like trying to explain a lottery number. The explanatory characteristics are irrelevant for these cases.
I remember when I worked at Moody's working on default models. The CFA types liked to do case studies of famous financial catastrophes, and worked through all the sub-accounts of say, Boston Chicken, or Enron. Unfortunately, these were really poor archetypes, and we could found that complimenting these analyses with our default model made both look less relevant. The cases that made for a great narrative usually involved a good degree of fraud, which is difficult to detect in real time. In the end, the group that sold training seminars on credit analysis continued their program of exceptional case studies, our models simply grinded out boring probabilities independent of these examples, and they remain independent areas of interest.
I find it much more fruitful to look at averages, mainly between groupings of interesting explanatory variables, and their correlation with the desiderata--not the reverse. That is, instead of looking at the top 10 stocks from last year, look at things like the top decile of p/e ratios, which predict stock returns at a much more modest level. We know these things are inversely correlated with returns (the value effect), so then the question is, how can I take advantage of this? If I combine this with cashflow/assets, or momentum, how does this work? Sure, it may at best add a couple percent to your annualized return, but at least its feasible. Needless to say, the current focus leads to an excess focus on highly volatile stocks, which is why I argue that these risky stocks have lower returns on average than their more boring counterparts.
Extremums can be informative, but they tend to dominate our information set anyway because they lend themselves to interesting narratives. Sure it would be best to know the big events if you had a time machine, but living in the present, if you really want to predict, focus on those things that are potentially predictive, which generally means looking at how averages relate to averages, as opposed to how outliers related to averages. The latter is mainly selection bias and random error.
A common idea is that outliers are more important than averages, as this is the them of Taleb's Black Swan, or Gladwell's Outliers. Predicting one great outlier is worth predicting many ordinary outcomes, so on one hand it seems like an optimal focus. Also, the outliers should highlight the essence of something. A stock that has risen 10 fold, or a great athlete, supposedly lays bare the essence of its greatness.
But I think we forget how biased our view is on exceptional events and people. We watch sports and learn about Usain Bolt, a most unusual man. Or my kids read the Guiness Book of World Records, containing stories about 1200 lb men and giant frogs. News is biased towards the exceptional, it takes no effort to emphasize it. In fact, it takes effort to see the ordinary. It's too bad people think of heroes as those who, for a brief moment, offered their life in some battle or harrowing situation, compared to the much more common heroism of providing one's family, not complaining, and being charitable to friends and neighbors, for decades.
A problem is that for any extremum is caused by its unique intrinsic characteristics and also random chance. We hope that we can more clearly see the intrinsic qualities associated with, say, a rising stock, by looking at sthose stocks that went up 10-fold last year. Yet, any extremum is probably a large random error. We simply can't predict the future very well, say an R2 of only 10%. Thus, a 100% stock return contains, on average, a 90% error. Yet, conditional up rising this much, this is only the mean error, its standard error is at least as large, meaning, any singular 100% return is probably all randomness. In this case, more analysis is worthless, like trying to explain a lottery number. The explanatory characteristics are irrelevant for these cases.
I remember when I worked at Moody's working on default models. The CFA types liked to do case studies of famous financial catastrophes, and worked through all the sub-accounts of say, Boston Chicken, or Enron. Unfortunately, these were really poor archetypes, and we could found that complimenting these analyses with our default model made both look less relevant. The cases that made for a great narrative usually involved a good degree of fraud, which is difficult to detect in real time. In the end, the group that sold training seminars on credit analysis continued their program of exceptional case studies, our models simply grinded out boring probabilities independent of these examples, and they remain independent areas of interest.
I find it much more fruitful to look at averages, mainly between groupings of interesting explanatory variables, and their correlation with the desiderata--not the reverse. That is, instead of looking at the top 10 stocks from last year, look at things like the top decile of p/e ratios, which predict stock returns at a much more modest level. We know these things are inversely correlated with returns (the value effect), so then the question is, how can I take advantage of this? If I combine this with cashflow/assets, or momentum, how does this work? Sure, it may at best add a couple percent to your annualized return, but at least its feasible. Needless to say, the current focus leads to an excess focus on highly volatile stocks, which is why I argue that these risky stocks have lower returns on average than their more boring counterparts.
Extremums can be informative, but they tend to dominate our information set anyway because they lend themselves to interesting narratives. Sure it would be best to know the big events if you had a time machine, but living in the present, if you really want to predict, focus on those things that are potentially predictive, which generally means looking at how averages relate to averages, as opposed to how outliers related to averages. The latter is mainly selection bias and random error.
Subscribe to:
Posts (Atom)