29 May 2013

I Need a Date

This post is too funny to not pass on. It's also useful for those teaching or doing baby-stats. All intro texts (R or stats or psychometrics or ...) cover the same ground, so to speak, but lack the cute setup.

On a serious note: measure of central tendency is handled, most often, in the first or second chapter; thence largely ignored. But if your field is resides in non-natural processes, i.e. anything man made or man interceded, it really does matter. As the saying goes, "figures don't lie, but liars figure." What you choose as your measure of interest/importance can make a big difference to your policy recommendation (yes, policy decisions, 99.44% of the time, is why stats get done). Since your audience is mostly suits (politicians, MBAs, CxOs, and such) who are innumerate, you can pull the wool over their eyes with ease. Or more often, the Suits have a policy decision, for which you are charged with finding a quantitative slam-dunk defense. Or, just as often, you're in the opposition, and you need to make a Wilson Burger. I give you Supply Side Economics and DSM-V (or DSM-5; I guess they figure that altering the name slightly will confuse the critics?).

25 May 2013

Losing At the Casino

For some time, but until now unfollowed, I've harbored the notion that MCMC (Markov Chain Monte Carlo) modeling might be suspect in The Great Recession. Turns out, I'm not the only one.

Skinny Tails
Monte Carlo simulation is a convenient statistical tool by which an analyst can make inferences over the probability of rare events. Basically, the method takes a model with its estimated parameters and simulates possible sample paths under some distributional assumptions on the errors. The technique is particularly useful for understanding the tail risk of distributions. The method has been widely adopted by banks, investors, and financial advisors. And, because the models failed so horribly in the current downturn, the method is now under attack.
Ya think????

Felix Solomon's take
So the lesson here, I think, is mostly that stock-market tails are fat. But there is a sub-lesson, too, which is that Monte Carlo simulations can be very dangerous, if they're implemented by people who don't know what they're doing. Including the quants at Moody's.

A 2007 critique (who knew anyone refused the Flavour Aid?)
The amount of derivatives in play is in the hundreds of trillions of dollars all bet on Monte Carlo Simulation. Wow!
It sure seems like a hell of a lot is riding on the Monte Carlo Simulation of CDOs as well as the parameters and assumption that feed into the model. If and when a Credit Event occurs that rifles through multiple CDO tranches, the guarantors will be about as well capitalized to handle the guarantees as is Madame Merriweather's Mudhut Malaysia. Do I smell a government bailout coming up?
Ya think???

Practical magic?
The downside of Monte Carlo is that it is more trusted than historical data. This misplaced trust is rooted in the idea that if a person has no historical data, then the Monte Carlo forecast can be anything that is believed plausible. For example, the phrase: "Contingency was determined by running Monte Carlo" is equivalent to "PI was determined by using Monte Carlo." Monte Carlo could render reasonable results for each value. They both will not be precise answers because Monte Carlo is an approximation, and the results for both could be completely wrong.

The issue with such methods is that they view human actions as Brownian motion (really!). While reasonable for physical processes, not so much for finance, where the rules are made up by humans and can be changed at a whim. That's why we got our Great Recession: humans changed the rules faster than the data streams could reveal the changes. The analysts made their decisions under the assumptions that Tomorrow will look pretty much like Today, which in its turn looks pretty much like Yesterday; and, as with physical processes, that the platform was stable during measurement. The fact that median house price had come unstuck from median income wasn't relevant to the modellers or their models; to look in that direction wasn't in the self-interest of those engaged in the game. Alas, if you look at the data through a toilet paper tube, you'll not see that hyena licking his chops for dinner. You're dinner.

And, as past is prologue, comes reports that brokerage houses are luring retail investors into options trading. There's nothing more fun than playing with nitroglycerine whilst bungee jumping.
At Ameritrade, which has been the most aggressive, derivatives trades accounted for about 40 percent of all customer trades last year -- more than double what it was just five years ago. A vast majority of those trades were in options.

The growth has been a big help for the online brokers at a time when stock trading has fallen. The commission on the average options trade is more than twice that on the average stock trade, according to TD Ameritrade's former treasurer, Michael Chochon.
[H]e saw investors taking up options trading and "blowing up" on an almost daily basis. He said Ameritrade carefully tracked the risks its customers were taking but did not warn them until they were close to losing it all, if then.

If that sounds a bit like the Banksters grinding ever more, and expensive, fees from common depositors to make up for the loss of all those wonderful profits on subprime and such, well "yes I said yes I will Yes" (that's from a famous dirty book, btw).

24 May 2013

No Child Left Behind

It was always a puzzle: why would Dubya promote making kids smarter? It is well known that smarter folks lean leftward, while the dumb and gun-totin' and God-addled lean rightward. Dubya couldn't really be serious, he was just looking for an excuse to defund inner-city programs deemed too efficient at leveling the playing field.

But, of course, the right leaning always claim that smart folks don't really exist, or are just pointless elites. And, may be so. But comes today a study which demonstrates that smarter means having a better bullshit filter. Dubya's worst nightmare. He and his Koch snorting buddies can't be happy.
There may be ways to improve one's ability to filter out unnecessary information, Tadin said, but he isn't sure what such an exercise would look like. "Most so-called IQ enhancing exercises focus on how you process things that matter, not how to better suppress those that don't," he said.

Well, not really. The filter bit goes by the name, critical thinking. Here's the wiki. For those old enough, or willing to read some history, there's the In Box Test. There's no obvious Wiki article, which tells you something. Here's one victim's tale, so I guess it's still around.

Needless to say, but I will anyway, IQ measurement is the bete noire of psychometricians, arguably the first quants. It poses much the same problem faced by financial quants: there is no absolute zero; it's all relative. And, as with the financial quants, it's mostly a zero sum game.

15 May 2013

Crab Apples

Much bytes have been devoted here to Apple's woes of late. And to whether, or not, these woes have any macro implications. Evidence is mounting that there are.

About a year ago, some opined that prepaids are cheaper, but:
Prepaid phone plans, where you pay the full price for a cellphone and then pay lower monthly rates without a contract, seem to offer what most budget-conscious people want. So why haven't they really caught on?

So, today we find that the tide has turned.
NPD said 32% of smartphone units moved in the quarter were prepaid, up from 21% in Q1 of last year.

Thousands of dollars a year to make phone calls? Of course not. These smartphones do much more, but that more can, generously, be characterized as self-distraction. Do people addicted to Angry Birds have a life beyond their particular version of self-abuse? Apple's (and what passes for tech these days) problem is that it has, with purpose, steered away from productive tech to entertainment. Toys. Mattel knows how to do that on a continuing basis. Apple hasn't shown that it can create a new toy from scratch. Google Glass may be worthless as a productive widget, or even as an entertaining toy, but it was created from scratch. Steve never did that.

Claiming Race

(What follows started out as a comment to a blog post, but growed like topsy, so it's ended up here. I've edited it, and added some additional context so that it stands alone.)

Here is the blog post on "claims inflation"
Here is the paper by the blog author
Here is the 2005 paper referenced in the author's paper

What caught my eye was the word 'inflation', since, these days, there's nearly always some agenda hidden in the bushes. And given that the topic was being explored by the insurance industry, well, like a multi-car pileup, I had to look. The blog author is a regular on R-bloggers, and is generally math-y and stat-y, so I was expecting to see a treatment that was data-y, math-y, and stat-y. Yet somehow above and beyond what one gets from BLS (and given that the author is EU based, UK/OECD/ECB/etc. stat offices). Much to my surprise, both the author's paper and the foundational paper are mostly handwaving verbiage. As my psychometric friends say, not much face validity.

And, why now to broach the subject? Unless one has been sleeping under a rock, it's clear that macro-inflation is really macro-*de*flation, thanks to the idiots running monetary policy in the UK and ECB. Even Japan, which had been the poster child for Prosperity Through Austerity failure for *two decades*, have finally gotten some smarts.

The motivating study opines:
Perhaps unsurprisingly, subjective judgement was the most common source of inflation assumptions for every class.

Well, I for one, am surprised. Actuaries, at least in the US, are drilled in maths and stats and data to a faretheewell, so that winging it bears on incompetence. It's not as if insurance is some newfangled enterprise with no history. Lloyds goes back to 1774, before we kicked the Brits out, for crying out loud. Although, one hears that most actuaries spend their day transferring standardized table data into standardized macro-laden Excel speadsheets. (If that sounds like you've heard that recently, well it is a Jonah-ish story.) Not that I'm being accusatory, only that's what one hears, sometimes.

One should read this paper albeit it being UK written, with a critical and jaundiced eye. If this is the state of the art in actuarial practice (and financial services broadly), then The Great Recession is unusual only for its rarity.

A number of problems:
- there are three sources of inflation; cost push, wage push, and demand pull
- holders of wealth (and insurance companies are numero uno) seek deflation, thus increasing their wealth with no risk or effort or intelligence
- all such *flation indexes are, of necessity, founded in sample surveys, and thus contain error. few, even professional economists, bother to read the details behind the BLS indexes. doing so reveals that the CIs are as wide as the Atlantic Ocean (New York to London). some even straddle zero! measuring *flation is really a fool's errand, be that as it may...
- injecting technological change to balance today's price of widget X over its predecessor is a dicey endeavor: today one spends $2,000/annum on "telephone" service where it was nearer to $200 a decade or so ago. is this service worth an order of magnitude more? is this a case of hyper-inflation? is this a case of stupid consumers over-paying for a basic service? how, if at all, is today's "telephone" service superior to yesterday's? certainly, call quality hasn't improved. is this inflation, or a new "good"? should any algorithm seeking to model claim inflation include such goods? is there any value add for higher priced robotic surgery, for example; or is such just inflation? if one uses macro-inflation indexes, such bending of the underlying data is not trivial.
- interest rates really do include a *flation premium, and thus it matters only if one's specific field of insurance does deviate widely from the norm

The result is: holders of wealth portray *in*flation as the norm, and thus seek law/regulation to hold them harmless, regardless of actual events. Until post-WWII in the US, the norm was *de*flation from the founding of the country, in fact. And this was due to lack of specie in circulation to keep up with the expansion of the economy. With fiat currencies, there is a proclivity by wealth holders to have them treated as specie, which is to say, *de*flation is preferred by wealth holders (as it always was), who have better access to government to influence decision making. The German/EU "Prosperity Through Austerity" cabal is a prime source. Japan has followed PTA precepts for more than 2 decades, and it hasn't worked out well; at least at the macro level.

The study of *flation is a topic of, officially, Political Economics, and holders of wealth take a position which is wholly Political and not much Economics: according to their rubric, all inflation is wage push or demand pull. Which is to say: unionism and Big Government dropping moolah from the sky; if only both would go away, Eden would flourish. This blindness to reality is what has fostered their continued harping that wages are too high and "Bernanke is causing hyper-inflation, and it's just around the corner!!!" Hasn't happened, and won't (at least not due to Bernanke), since all that moolah central banks are tossing around is going to current large holders of wealth, and not The Common Man. Only if, and when, The Common Man gets more moolah in his pocket will The Common Man have greater (what economists used to call) Real Demand for goods and services. This week's release of import/export prices by the US government has them down, again.

What all this has to do with claims inflation? In the near term (and the medium term if the Smart People in the ECB follow their Japanese colleagues), especially in the EU countries so enamored of Prosperity Through Austerity, *de*flation will be the event. The only certain source of *in*flation is resource failure (aka, cost push); all else is up to human folly. Well, turning corn into fuel had a signficant effect on the price of tortillas; again, human folly. Changes (increases, by assumption) in technology in medical can be viewed as resource scarcity: more expensive procedures which didn't exist earlier drive up costs (even if they're promoted as cost saving, of course) in the near to medium term. Robotic surgery is prime example; some in medicine are now looking more closely at the claims of superiority. Finally.

From Gesmann's paper:
The RAND Corporation published a detailed review of the dramatic increase in claims frequency and severity of medical malpractice claims in the US in the early 1970s [9]. Its model suggested that the single most powerful predictor of claims frequency and severity is urbanisation.

If ever there were a confounded bit of econometrica, this would be it. Urbanization, per se, is not meaningful; too general to have much use. What matters are the underlying factors of malpractice: nature of procedures (cities provide a wider gamut, thus a higher risk "index" overall), availability of patients for risky procedures (ditto), availability of lawyers (ditto), and so on. If one were to follow the implied wisdom of the conclusion of the study, insurers would only have country doctors as clients.

The major driver of insurance profitability is clever contracting and thus limiting payouts. Followed by interest rates. With the Greenspan/Bernanke monetary lever to motivate economic recovery, bond earnings have diminished. To the extent that insurers by law, regulation, or preference hold primarily/only bonds as investment instruments, they will forgo non-premium income (and would need sharp minded investment offices to manage an equity portfolio, hopefully not A London Whale) and rely more than historically on premiums, raising same. This will drive out marginal clients, incrementally, thus reducing the pool, and putting yet more pressure on premiums. Rinse, repeat. In other words, even the 1% need ObamaCare; in extremis even they won't be able to afford insurance which covers only them. Who'll cover the amortization of all those MRI machines, and the like? I suppose the Koch brothers can buy one for themselves.

Historically, *flation really is global, and is thus reflected in returns on investment. To the extent that technological change is drawing out payback periods, especially for "new" tech, (and by simple arithmetic) thus lowering real rates of return, having investment income to subsidize premiums may be coming to an end. (Aside: real returns on physical capital are what matter, and should governments raise risk-free interest rates to satisfy financial services companies, this will siphon off yet more moolah from real investment. It's a zero-sum game, I'm afraid.) To the extent that the political Right hacks away at the middle classes, thus shrinking the size of the insured pool (insurance, by definition, is pooled risk) *flation will be the by product of competing forces; on one hand insurers continue to evict marginal clients (thus reducing payouts; perhaps greater than lost premiums), while on the other socialized insurance (to whatever extent) pays for those evicted. Socializing costs, while privatizing profits.

And there is, by the bye, little evidence that long-term interest rates must needs be higher than short-term rates since long-term projects are, by definition, riskier. They're not. In fact, The Great Recession was motivated by fiddling in the short-term rate venue. The risk of a project rests in its physical reality (if any), not financial engineering. The real interest rate is solely determined by technological change, as it is embodied in new, superior, plant and equipment. To the extent that technological change turns out to be log on time, rather than (non-)linear, coupon clipping will survive only by force of law. The economics and engineering determine the real outcome.

06 May 2013

The Missing Linked

LinkedIn fell into its potty on Friday, and the post-mortems proliferate. Is there anything to be learned, beyond the certainty that LinkedIn is wildly overpriced at the moment?

From the quants' point of view (those that labour on Wall Street and The City), LinkedIn is yet another social media problem child. All of the social media companies, private and public, face the production issue. They don't supply any; they're ephemeral toys. They all provide an outlet for the ADHD addled (and some data indicates that fully fledged adults are the major users, not kiddies). They may even be a cause of mental degradation; how many Gen Ω's can do long division? Just as Word (and other GUI word processors; name one?) is known to redirect effort from substantive content to the pixie dust of formating, social media, as a business, is a net sink.

Just as housing is not an investment, from a macro-production point of view, neither are the social media companies producers of valuable (in the sense of being assigned a value) output. Just as Mattel makes Barbie, and sells rather a few, sales of Barbies is a zero-sum game, again, from the macro view. Sell more Barbies, means sell fewer Beany Babies (are they still around?). Some consumer goods, food for example, have intrinsic economic value. Most, alas, do not. Moreover, for those "advanced" economies which eschew production of physical goods (USofA, I'm talking to you), in favour of "services", the problem gets stickier. Is an economy which ceases to produce food in favour of social networks sustainable? From the point of view of the micro-economist, the question isn't even relevant; the value of a social network company is a function of the moolah that can be extracted from users who use the network. The micro-economist is devoid of policy. The macro-economist, on the other hand, is concerned with the long-term prospects of an entire economy; policy is what matters. The micro-economist, particularly those employed by corporations, is also concerned with policy, but only in a near-term sense. His job is to argue for policy which socializes cost while privatizing profit; all wrapped up in "... twenty seven eight-by-ten colour glossy photographs with circles and arrows and a paragraph on the back of each one explaining what each one was to be used as evidence...".

Why, might one think, do the Right Wingnuts continually bray about "debasing the currency" whenever Keynesian efforts are proposed; particularly now in the continuing aftermath of their wonderful Great Recession? Economies which are based on physical production are consumer-facing (by and large), while the service economies (most of the West, these days) are corporate-facing. That is to say: the high wage employment in services (where BA and more is required) provide those services not to the Barbie buyers, but to companies for internal uses. The financial services sector, progenitor of The Great Recession, is but the most egregious example. The value of services is notoriously difficult to quantify, since there's no there, there (as Gertrude Stein observed about Oakland). The valuation of services is super-glued to the value of the currency in which the services are conducted. It should come as little surprise that the financialization of US corporations has been in process for years. GE makes as good a poster child as any.
In 2007 GE Capital's profit made up 55% of the company's total.

But IT generally is in the same boat, so to speak; do I hear "Iceberg ahead"? Little of what IT does is directly sold to human consumers; we have no Barbie. We're chained to corporations.

With the exception, you saw this coming, right? Social media computing. How has this happened? And is the course of events a benign or nefarious precursor to the future? Social computing is the result of the same impulse as garage bands (and garage companies) of the past: a cheap way to fill time. The difference being that becoming a social media "developer" can be done for under $1,000. As the saying goes, water finds its own level, and lazy people take the path of least resistance. Why learn the hard bits of computer science, when you can use brute force, crudely applied, to make a toy which might make you a gazillionaire? Why learn how to extend your skills to be a utility infielder, when you can take a shot at being Tiger Woods (the previous model)?
Like everything else in technology, the cost of starting a startup has decreased dramatically. Now it's so low that it has disappeared into the noise. The main cost of starting a Web-based startup is food and rent. Which means it doesn't cost much more to start a company than to be a total slacker.
-- Paul Graham/2005
(Note the date!!!)

There's a lot of talented, smart folks betting on 13 Red. Only a vanishing fraction will win the bet. From a wagering point of view, which figures an expected value as the multiplicative product of ultimate value and probability of gaining that value, it might make sense. It is well known that Powerball and The Big Game (now called Mega Millions) are sold far more heavily as the expected value of a ticket approaches the ticket price. This is actually, from a quant's point of view, an intelligent wager. So long as the moolah one spends on those tickets isn't diverted from necessary expenses. Aye, matey, thar's the rub. It's much the same as flying in commercial aircraft: the probability of falling out of the sky is small, but the results, should the plane plummet, are life-changing.

The more pernicious, and less obvious, side effect of garage band computing is that it diminishes the quality of developers. Face it: if you're interest is in spewing out code, since that's the nature of what you can do with your $1,000 laptop (and have beer pong parties instead of going to some class on linear algebra), rather than building solid data foundations, you'll spew out code. Thus, the Gen Ω crew are throwbacks to their granddaddies, just using a laptop rather than a glass house. But they're both primitives. The Goths of the 21st century.

Which brings us to valuing what the garage band developers do. The very notion that Zuckerberg, or even Jobs, is somehow a modern Edison is poppycock. I just made up that assertion. Let's see if there's a record of anyone making it, shall we ...
Here, all three on one list, with Edison the clear leader. Zuck has no business being on the list, and Jobs isn't close to being half an Edison. For those who don't follow the link, these numbers are the results of a survey of about 1K Gen Ω's.

Well, at least I didn't find any reputable observer asserting the either was more significant than Edison. Or not. The reports don't say how many Gen Ω's voted Jobs or Zuck first! Think about that for a second.

And, here is a really scary one:
... who invented the Internet? He gave about 20 possible candidates but suggested that, depending on the next 100 years, it's possible that people will believe the answer is Mark Zuckerberg.

I reiterate: the P/E of Mattel is 19.72; Facebook is 577.78. QED.