26 April 2014

Guys and Dolls

When I was just a little boy, I asked my mother what will I be? Will I be pretty... Well, wrong musical. It was actually written for a Hitchcock film ("The Man Who Knew Too Much")! Who knew Hitchcock did songs? Not all that much.

In any case, I'm going more the Nathan Detroit, "Guys and Dolls", route. Some way to get Nate into the stream of consciousness. You may know that Nate Silver left the NYT for other parts, ESPN for the moment, so the Times have built an in-house version of Nate.

This is where the link goes from the RA piece.

What's cute is the RA piece describes the process as "...a similar poll-aggregation methodology to that used by Silver". There's been very few times that Nate/538 have been accurately characterized. Most often Nate is lauded as Saint Statistician, and while he's been good at what he does, he ain't a math stat. Having made his first splash in baseball, and given that baseball zealots were among the first groups to co-opt the word "statistics" to describe any numeric display, I suppose it was inevitable.

Kind of like what RM supporters have to deal with: any bytes stuffed into a relational engine is a "database", just because the coders say it is.

25 April 2014

The Problem With Slavery

This essay has been simmering for quite some time. Recent events and others' essays have proven too much. It will out.

Let's start with the head exploding op-ed page of today's dead trees NYT: in the left column is the Right Wing Brooks and in the right column is the Left Wing Krugman, whilst in the center is one Evan Mandery. The Wingers have different, surprise, takes on Piketty, and Mandery takes on the furtherance of privilege ensconced by the Roberts courtiers.

Not very surprisingly, I've mulled writing about Piketty, but most of what I have to say has been said by the mainstream pundits. The privilege of legacy, strengthened by Roberts, was about to emerge from these fingers, until Mandery did the work. I wouldn't have gotten into the Times, not a known pundit of course.

Another piece which is part of the stew is this discussion of Simpson's Paradox. I was moved to comment, not to question Simpson one way or the other, but rather choice of data. Not my most literate comment on record. The UCB data is built into the R distribution, so has been used many times. The piece, however, committed Type III error, and my error was in not being clear about that. Too often, stats/quants are applied to human events as if those events were existentially equivalent to events in the natural world. The natural world, chemical or physical or animal, obeys God's/Nature's rules and the participants don't have the option to amend the rules to suit. Human participants can and do change the rules to suit. Quant modeling, with the notable exception of those now labeled Behavioral Economist, remains wedded to the notion that the rules of engagement are external. They're not. As I've stated oft many times: when policy conflicts with data, data loses. We see this in the Right Wing reaction to Piketty's book, Roberts' increasingly transparent perfidy, and in the insistence in using the UCB data to make inferences about college students. Mandery's piece eviscerates one of the forks in that road.

The notion of black swans is instructive. To beings of the time, the asteroid slamming into the Yucatan was a black swan, an entirely unpredicted catastrophic event. Were an asteroid of that size to put the Earth in its crosshairs today, we'd have plenty of notice. And, if Bruce Willis isn't making some other movie, we would deal with it. On the other hand, the Great Recession isn't really a black swan event. Not least because some of us predicted it and said why. But more to this missive's point, the housing collapse happened because some humans decided to change the rules of mortgaging and securitization. God didn't do that, nor did Nature (for those not of the Deist persuasion).

So, why the title of this missive? How do we travel from the cites to slavery? Follow the Yellow Brick Road.

Aside from the moral issue, why is slavery a problem? After all, those who would rather that working folk get no more than subsistence wages (which is the essence of slavery, don't you know) quote from The Bible. And up through the middle of the 19th century USofA, slavery made some economic sense. It's not often discussed, but at the time of the American Revolution and for some decades hence most hard currency in the colonies came as the result of Southern plantation exports to Britain and Europe.

Exports, get it? With goods nearly wholly from pre-industrial agriculture. In capital lite production, slavery (and its first cousin, indentured servitude) is perfectly sensible. The goods are not to be consumed by the labor, or even for the most part, within the producing economy. What China does today with rather primitive capitalism, the USofA did in the early 19th century: make stuff, mostly from the ground, to ship to Europe to get hard currency. Did I mention that the head of Foxconn, when the suicides became such a distraction (oh why won't they live like slaves quietly?), announced that he was going to replace as many humans as possible with robots. Since Foxconn didn't, and still mostly doesn't, depend on domestic consumers, this makes perfect sense.

For modern uses of capital, all things compute related in particular, universal slavery is a problem just because it kills the consumer goose needed to lay all them golden eggs. While soft currency nations like China can engage in what amounts to wage arbitrage, the gag ultimately (and since the world isn't linear, 'ultimately' is closer than it appears in the mirror) collapses since there are no longer consumers for the burgeoning output. Piketty makes the point that returns on capital have outstripped overall growth, and thus powered concentration. There will come an inflection point, already passed in my estimation with regard to compute related capital, where capital ceases to have value since value is grounded in consumption, said consumption provided by a consumption (middle) class. And without consumption, there's no return on capital. Kind of like petro; you can't eat it or drink it or wear it. Its value is derivative. One capitalist can gain, momentarily, advantage by replacing humans with robots. But, as your Good Mother used to say to you, "what would the world be like if everybody behaved like you?" Bad grammar aside, the answer of course, is collapse. Capital requires consumers, and the mass of them work for wages. Get over it. But capitalists, and the quants and economists who churn the data for them, are basically knuckleheads. The obvious idea that the well goes dry if everyone pumps it furiously is quite beyond them. The micro(math)-economists are heavily to blame for this, in that they've implemented the naive` notion of dog-eat-dog which they base on Adam Smith (the real one), which he mostly disavows within his text. That is to say, he wasn't all that happy with capitalists.

There isn't any new continent, laden with resources as North America was in 1800, to exploit. What you see is what there is. Well, in a diminishing sort of way. There's a summer replacement TV series (CBS, if I recall) which I've only seen the promos for, "Under The Dome". The conceit is that some town wakes up one morning to find itself sealed in by a transparent bubble. A closed system. Well, the world economy is just that. We're likely, save great political will on the part of the hoi polloi (most of whom worship at the feet of God and guns), in a world of a few masters and many slaves. Just don't get uppity. The problem for capital is going to be ever diminishing returns: there's no New World to export from to the Old World. Mercantilism worked well for the Brits for a couple of hundred years, as well as various other European monarchies, not so much today.

Recent reports are that "some" want to replace the Almighty Dollar as reserve currency with some other, likely bundle, currency(ies). The problem for China, should we assume that the renminbi displaces the Almighty Dollar, is that it instantly loses the ability to attack the ROW. Now, it has to depend on its domestic market to absorb output. The best it can hope for is to get back alien renminbi for its exported output, of which there currently isn't all that much. The shoe will be firmly on the other foot. The days of collecting somebody else's reserve currency are over. All those Foxconn slaves become an instant problem, since their output no longer brings in hard currency from ROW, and they don't earn enough to clear burgeoning output. Oops.

Sometimes it makes more sense to be the mouse that shuts its damn mouth.

19 April 2014

Thread Bare Existence

Another day of news, another thread not to be ignored. And, as usual, it isn't a golden thread. Well, for those that have lots of gold, perhaps it is.

Who knows when it came into the lexicon, although thanks to movies most of us assume it was WWII Germans, but "Ve ver onhly fallowing ohdas!!" has become the catchphrase for acquiescence to evil. Nevermind that doing so oftens yields substantial income to those who do the bleating.

Insurance? Most folks don't really know what it is: pooled risk. And ever since, insurance companies have been looking for ways to exploit the many to benefit the few. It's called market segmentation. According to Adam Smith's (the real one) thesis, this can't happen. The reality is that if only to stop bad people with oligarchs is for good people to have oligarchs. This is countervailing power. In the insurance business, here in the USofA, this countervailing power is vested in the states; insurance commissioners, and such titles. Divide and conquer at its best. Unless the insurance lobby wants something really badly, in which case they complain about fractured regulation, and demand that Washington pass (more lenient than the most comatose state) overriding law.
Such actions, he said, can be discriminatory, because studies show that lower-income consumers do less comparison shopping than wealthier consumers. Also, he said, auto insurance rates should be based on risk factors, not shopping patterns.

"It's unfair," Mr. Hunter said. "Auto insurance isn't a luxury. It's mandatory."

Quants working over their Excel sheets looking for ever more obscure ways to exploit the many to benefit the few.

Kind of makes you wonder why Geico would use a *pig* as a spokesbeast?

Not content to take from the many clients, the Fortune X00 are happy to ride on the backs of the few who provide open source software. "Hey, it's free!!" Well, it's not. stealing Open Source may be a tad too strong. But may be not. While java and linux have gotten substantial support from IBM, and some others, most corporations just suck up open source like pigs at a never empty trough. So, of course, OpenSSL finally gave them their comeuppance.
Although any programmer may work on OpenSSL code, only a few regularly do, said Ben Laurie, a Google engineer based in Britain who donates time to OpenSSL on nights and weekends. This is a problem, he said, adding that the companies and government agencies that use OpenSSL code have benefited from it but give back little in return.

"OpenSSL is completely unfunded," Mr. Laurie said. "It's used by companies who make a lot of money, but almost none of the companies who use it contribute anything at all."

The Wife likes to describe this as The Hippy Way: "What's mine is mine. What's yours is mine." Lots of practice has she. So do corporations. While linux has a strong infrastructure, not least Linus, few other open source projects do. Frankly, I was surprised to discover that OpenSSL, given its keystone nature in the innterTubes, is so poorly treated. Once again, self-absorption trumps self-awareness.
[T]he OpenSSL project has operated on a shoestring annual budget of $2,000 in donations -- most from individuals -- which is just enough for volunteers to cover their electric bills.

I imagine that the CTO/CIOs at all those Fortune X00 companies earned some part of their bonuses through the expedient of slurping up OpenSSL, uninspected, rather than building/buying a vetted proprietary package.

Which brings us back, once again, to Thomas Piketty (who really is French, but with a name I would never conjure as such). I expect that Republican cloak rooms around the country will be loading up on Freedom Fries whenever he comes to town.
At the book's center is Mr. Piketty's contention -- contrary to the influential theory developed by Simon Kuznets in the 1950s and '60s -- that mature capitalist economies do not inevitably evolve toward greater economic equality. Instead, Mr. Piketty contends, the data reveals a deeper historical tendency for the rate of return on capital to outstrip the overall rate of economic growth, leading to greater and greater concentrations of wealth at the very top.

You can look up Kuznets on the Wiki. Basically, he was a Pollyanna apologist for monopolists. The flaws in his thesis were, mostly, two: that resource endowments, per capita, would continue; and that the social contract in the immediate post WWII period was permanent. Both assumptions were wildly false, as Ronald Reagan proved. Adam Smith (the real one) recognized, and condemned, the power of capital. Really. The Wiki quote (it's cited, so don't argue):
Again and again, Smith warned of the collusive nature of business interests, which may form cabals or monopolies, fixing the highest price "which can be squeezed out of the buyers". Smith also warned that a business-dominated political system would allow a conspiracy of businesses and industry against consumers, with the former scheming to influence politics and legislation. Smith states that the interest of manufacturers and merchants "...in any particular branch of trade or manufactures, is always in some respects different from, and even opposite to, that of the public...The proposal of any new law or regulation of commerce which comes from this order, ought always to be listened to with great precaution, and ought never be adopted till after having been long and carefully examined, not only with the most scrupulous, but with the most suspicious attention."

And there's where I believe Piketty's been looking at the past a tad too much, although the historical data are sufficient to cement the 1%'s place in the future. The issue is "return on capital". While it's been true that Olde Style capital got us to where we are, New Style capital has a problem. We see that problem in the hoarding, rather than investing, of earnings. Were there bounteous avenues to money making to be had, they would be. The modern investor, to use a term from one of them I recently saw, looks for "asset light" companies. Think about that for a second.

The modern uses of physical capital face a grim present and future: the outlays are just too great for all but a few monopolists, while the consumer product lifecycles grow shorter. The days of making an open hearth furnance, and reaping the returns for decades, are well past. In such a circumstance, deflation is the hoarder's friend. While we got here through a version of oligarchy, the future will be one of resource poverty and stasis.

Have a nice day.

10 April 2014

Freud May Get Really Rich [update]

That didn't take long. Yesterday, it was reported that Colbert was the odds on favourite to replace Letterman, now today CBS announces it. Colbert, the real one, is a flaming left wing pansy. Colbert, of the Report, is a flaming right wing Nazi. Some number of his fans are going to be disappointed. Either that or he plays both ends against the middle, and his psych bill is going to be through the roof.

[update]
Letterman was 46 when The Late Show With... began. Colbert will be 50, if Letterman continues to the end of his contract. While NBC went with Fallon (39) and ABC Kimmel (36), although his show didn't start in the 11:35 slot at first. Perhaps old folks should pursue late night emcee as the pre-retirement vocation, after all.

07 April 2014

About Your Mule...

There's that old, and continuing, adage: "you can't beat Mr. Market". Well, baloney. You sure can beat Mr. Market like a rented mule. For some definition of 'you', of course.

If you've billions in moolah, it does get a little harder, from one point of view; yet even easier from another. Warren Buffett is a case in point. The piece exposes him as merely mortal the last few years. May be. Then again, may be not.

Mr. Market is kind of, sort of, Dr. Jekyll and Mr. Hyde. If you've got $100,000 in moolah, then turning it into $1,000,000 is certainly feasible. If you've got billions, turning that into tens of billions, not so easy. Or may be it is.

There are only two factors which cause capital gains in stocks: real events, and real events. (If that sounds familiar, recall this old canard: "there's the three killer apps of the PC; word processing, spreadsheets, and word processing". Still true today, which is why phones and tablets have inhaled the consumer space.) So, Mr. Market is either equities (narrowing focus a bit) across the board, or a particular corporation. For equities, Ukraine is the current event driving his angst. Earlier, it was the tsunami of moolah from The East (more on this later). For the ABC Corporation, it's... who knows?
Mr. Buffett's talents are widely known. But despite his celebrated past performance, his returns since the beginning of 2009 have been disappointing.

He's too big to hit a grand slam.

What's happened since 2009? A so-so stimulus, followed by a number of QEs. And, let's not forget, The Giant Pool of Money still looking for 10%, risk-free returns (the expanding Chinese housing bubble is a chicken come home to roost). In all, Mr. Market, in the large (Mr. Hyde) is floating on helium. In such a situation, Mr. Buffett can't do well, since productive value becomes unstuck from fiduciary value.

Could it be that the Fed's pushing a string? The idea with the QE exercises was to make physical investment desirable, by lowering the cost of money to do so. More investment in physical production, yields more employment, which yields more income, which yields more demand, and around we go. If only we give the carousel enough oomph to turn on its own. Hasn't worked out that way.

On the other hand, the others with moolah piles of Mr. Buffett's sort, can do quite well. Here's where quants can earn their keep (if you assert that money manipulation is 'work') since the tactic is to track gross money flows and move ahead of the lemmings. Since Mr. Buffett chooses to make long-term bets on, often, whole companies, moving in and out of positions is difficult, if he attempted that. Paulson, with a little help from being inside, did just that. So did Jamie Dimon.

What quants aren't very good at, since it's outside their purview, is understanding catalyzing events of individual companies, the Dr. Jekylls of this tale. This is where the lonely $100,000 has the advantage. Did deep into ABC Corporation, tech or biotech these days, and you'll find some type of "innovation" about to be birthed in the next X months or Y years. Buy up a bunch of ABC, and wait for the catalyst. It helps a whole lot to understand the venue in which ABC operates, of course. Buffett's billions are too big to get away with that; the ABC's of the world tend to be, at most, small-cap outfits. The Baker Bros. funds have managed to pull it off with some consistency in bio.

The string pushing hasn't worked out as planned.
According to a report by Moody's Investors Service, American companies outside the financial industry were sitting on a combined $1.64 trillion of cash by the end of 2013. And tech giants like Apple, Google and Microsoft had the most.

That's a really big number. And doesn't include the Too Big to Fail cabal. Is it any wonder that the QEs have had little effect? Other than the ABC Corporations converting old high cost debt into new lower cost debt.

Speaking of Big Numbers, Krugman shares the page with a piece taking the Big Data meme to task. Two doses of reinforcement.

Krugman:
Many people understand that a falling price level is a bad thing; nobody wants to turn into Japan, which has struggled with deflation since the 1990s. What's less understood is that there isn't a red line at zero: an economy with 0.5 percent inflation is going to have many of the same problems as an economy with 0.5 percent deflation. That's why the I.M.F. warned that "lowflation" is putting Europe at risk of Japanese-style stagnation, even though literal deflation hasn't happened (yet).

Here, I still find he's too Pollyanna about the whole thing; there are folks who'd love a Japan. The cash hoarding by the .1% is clearly a move to get deflation up and running; they've run out of ways to productively convert fiduciary capital to real capital, and having no income generating skills, "[We] have always depended on the kindness of strangers." Deflation, while it might not get them 10%/annum, is totally risk free returns. Who can turn that down? "Turning Japanese I think [we're] turning Japanese I really think so" (The Vapors, 1980)

In my dead trees version, more than half the page (less Krugman's full length column) is devoted to a take down of Big Data. Again, comforting to know that some who can finagle a piece on the NYT have figured it out.
The first thing to note is that although big data is very good at detecting correlations, especially subtle correlations that an analysis of smaller data sets might miss, it never tells us which correlations are meaningful. A big data analysis might reveal, for instance, that from 2006 to 2011 the United States murder rate was well correlated with the market share of Internet Explorer: Both went down sharply.

Makes sense to me!

But how to explain this non-sequitor?
FINALLY, big data is at its best when analyzing things that are extremely common, but often falls short when analyzing things that are less common.

Fact is, it's the outliers that Big Data allows us to find, as the murders and IE example (OK, it's not likely causative), or that Target customer that found out his daughter was preggers. Now, their examples are a bit off target, so far as I'm concerned. The issue with Big Data has always been a cost/benefit one: collecting enough of the right data to find black swans early enough to profit is difficult to calculate, especially if you don't know which particular, very valuable, swan you're expecting. That's part of what makes them black. The Great Recession was a black swan to many, simply because they'd ignored the data which really mattered. As lemmings, they all ran over the cliff. Not so black for a few who paid attention. Big Data didn't help them, rather Econ 101.
About a year after Pole created his pregnancy-prediction model, a man walked into a Target outside Minneapolis and demanded to see the manager. He was clutching coupons that had been sent to his daughter, and he was angry, according to an employee who participated in the conversation.

"My daughter got this in the mail!" he said. "She's still in high school, and you're sending her coupons for baby clothes and cribs? Are you trying to encourage her to get pregnant?"

Dancing belly to belly.

05 April 2014

Attack of the Vampire Squid [update]

With all the boohooing from Titans of Industry, and if you believed them, you'd think that Obambi has been the apocalypse of Satan. Well, not so much:
Floyd Norris
After-tax corporate profits in President Obama's five years in office have averaged 9.3 percent of G.D.P. That is a full two percentage points higher than the 7.2 percent averages under Lyndon B. Johnson and George W. Bush, previously the presidents with the highest ratios of corporate profits.

[update]
A reader comment, and my reposte, led me to wonder what lefties had actually said that Roberts was moderate. The first search came up with this piece from 2007, and I stopped looking (if you follow the link, it's a 2006 paper):
And in an academic piece written as Justice Alito joined Justice Roberts on the Court, No Exit? The Roberts Court and the Future of Election Law, I concluded:

Making predictions is always dangerous, and the conclusions I reach should be taken in the tentative spirit in which they are made. My best guess is that a decade from now, we may well face a set of election law rules that differ a great deal from today's rules. It may be that in 2016, individuals, corporations, and unions will be free to give as much money as they want to any candidate or group, subject to the filing of disclosure reports.

Off by a few years, was he.

03 April 2014

Fiber Art

A long ago friend decided to go into the fiber art business, after being a bureaucrat and educator. I guess there's art to be found in fiber. Just not in North Carolina. That happened nearly three years ago. And the innterTubes cartel is still at it.

It was, therefore, with some amusement that I read up this Times story on how innterTubes pioneers are moving house to have really (by US standards, anyway) high speed broadband. The piece goes on and on about how behind the curve the US is, mostly in fact due to monopolistic rent-seeking private capital, with regard to innterTubes service.
"I just returned from Stockholm where fiber connections are cheap and as available as running water," said Susan Crawford, a visiting professor at Harvard Law School and author of "Captive Audience: The Telecom Industry & Monopoly Power in the New Gilded Age." As a result, she said, developers there have "a digital sandbox to play in," which means they are more likely to develop the next generation of software and hardware.

Ah, it must be all that midnight sun that has addled the psyches of Swedes to make them hate the Free Market. Fact is, the Damn Gummint (DoD and NSF, mostly) created the backbone. And it was academics, mostly, that figured out the tech. Only later, once the heavy lifting was done, did the Daddy Warbucks folks say, "gimme, gimme".

What's truly remarkable, and must be due to denseness on the part of the writer and/or editors:
It's why Brad Kalinoski and Tinatsu Wallace moved from Los Angeles to Wilson, N.C. The husband-and-wife team co-own Exodus FX, a company that provides special effects for commercials, television and feature films like "The Black Swan" and "Captain America."

The reason that bit is remarkable: North Carolina has paid off the ISPs by banning municipal broadband service, fiber generally. I guess irony is lost in the Times. This is a state by state list.
"In New York, I pay four times as much as someone in Stockholm would pay for a connection that is 17 times as slow on the download and 167 times as slow on upload," [Crawford] said. "Most of us are paying enormous rents for second-class service."

Well, corporations are people, too. And they need big houses to hold all those big ideas they have.