30 August 2013

Nice Threads

Back in 1955, one Sloan Wilson wrote a book "The Man in the Gray Flannel Suit". I don't know whether I've read it; in 1955 I wasn't reading at adult level. By the 1960's, when I was, it was certainly well known; likely as a prescient omen. The counter-culture and all that. Other books from the 50's: "The Hidden Persuaders" (1957), "The Power Elite" (1956). Here's a factoid: in surfing the Wiki for other forgotten books, I tried 'keeping up with the joneses', expecting to find something from about the same period; post WWII, in particular. But Nooooooo. Turns out that's the title of a comic strip from 1913!! I'll bet you thought McMansions and two car garages were a recent phenomenon. Apparently the rot has been simmering for a century.

Anyway, IBM's been running a series of ads on the TV about 'social biz'. You can YouTube for some of them, but I didn't find a link for the one that set me off, in which we see a series of interviewees and a narrator bemoaning the fact that too many hires just don't work out. Use social/psych methods to weed out the undesirables. This from what is, superficially, a high tech outfit (it isn't, and never was, but that's a much longer Other Story).

That's the potatoes and onions. Here's the meat for the stew.

Yesterday it was reported:
"Every day, they are learning how brilliant [Snowden] was," said a former U.S. official with knowledge of the case. "This is why you don't hire brilliant people for jobs like this. You hire smart people. Brilliant people get you in trouble."

I suppose that means brilliance comes with a conscience. Einstein was a pacifist.

29 August 2013

The In Crowd

"I'm In with the In Crowd"
- Dobie Gray (the first to cut it, but who's he?)
- Ramsey Lewis (a bit Herb Alpert)
- Petula Clark (the first version I heard)
- Mamas and Papas (totally forgot they cut it)
- Brian Ferry (didn't know about)
- Alvin and the Chipmunks (eeeeeeeeeeeewwwww!!)

All by way of introducing the latest to cut a version:
Larry Ellison!!! Yea, the crowd goes wild!! He's got rhythm!! Twerking Ginny Rometty up on stage!!

While my beloved Organic Normal Form™ on multi-processor/SSD meme has yet to go viral, Larry and his frenemies continue to look for speed. In odd places. DRAM is staying cheap, but it's never going to be supportive of really big datastores (unless/until we get 128 bit processing). So what's up with the latest, soon to be announced, in-memory from Larry?

Following one of the links, reminds of something I noted when Larry bought Sun: he did it to steal IBM's mainframe business, a place Oracle hasn't had a lot of success; not for java or MySql (both of which he's managed to mangle, badly).
"Oracle thinks that by controlling the total stack--applications, database, middleware, down to the silicon itself--it gives them advantages they can't get by using commodity products like x86," [Nathan Brookwood, principal analyst at Insight 64] said.

Larry wants to party like it's 1979, and he's living in Armonk.

Leavin' On a Jet Plane

Has anyone noticed that Boeing continues to make a considerable amount of money shifting the 737? Today's announcement:
7:46AM Boeing statement on WestJet's intent to purchase 65 737 MAX airplanes (BA) 103.27 : Co is "delighted" that WestJet has entered into a letter of intent to purchase 65 737 MAX airplanes, consisting of 40 737 MAX 8s and 25 737 MAX 7s. The pending order is valued at $6.3 bln at current list prices.

Could it be that the laws of physics are so well known, not about to change any time soon, and determine that a nearly 50 year old airframe is the best Newton has to offer? Is it any wonder that The Best and The Brightest go on to use their quant skills to play dice with the economy? Could it be that there are limits to what we can know? We can't know that which Mother Nature has not done. We really can't create our own reality. We still require air, water, and carbon. Gravity will always be with us. Sub-atomic physics will always be probabilistic. The (natural) periodic table is complete. And so on. Unlike the 19th century, so beloved by wingnuts, there was a lot (perhaps most, in total) about Mother Nature we were still learning. And turning that learning into economic activity. As we approach (if we haven't arrived) the limits of ferreting out Mother Nature's scheme of things, how do we drive economic activity? Finding petroleum or uranium won't happen again. And, they ain't more like them. Mother Nature has showed us so.

Are we doomed to a spiral of The Great Recession? Are we doomed to a world of economic activity connived from morphing human rules of behaviour? Rather than building a wholly new widget from newly discovered Unobtainium, we play out our lives building, or gaming, ever more elaborate accounting schemes? Instead of Einstein being the hero, it's Ponzi?

Have a happy morning.

28 August 2013

This Ain't Lake Woebegon

Yet another post (and a Blogger blogger, to boot) extolling the virtues of better, cheaper education. And yet again, missing the point.

Which point is: the Eden of the post-WWII USofA, up to 1973 (OPEC embargo, and all that), wasn't driven by better, cheaper education. Yes, the GI bill did allow servicemen/women to get a college degree. But, no, that isn't what created the broad, demand supplying, middle class. That middle class was created by the egalitarian ethos which was the result of the socialist afterglow of shared sacrifice in war. It's that simple.

Unions were widely legal, and widely joined. Blue collar wages were middle class, so people bought cars and houses and sent their kids to State College. Corporate titans took out substantially less of the corporate revenue. The GDP was skewed toward actual production, rather than financial gyrations. The USofA controlled the international exchange system, both implicitly and explicitly, to a far greater extent than any time (save possibly, today) until the oil embargo.

So, as an R exercise, it's an interesting post. As prescriptive for "what's wrong with the US education process", not so much. If all kids get CS (or name your favourite) degrees, they'll turn themselves into dollar-a-day off-shore coders. Car mechanics and plumbers will make more (likely, many do now). Too many folks who know Process Q, may, in the short term, increase the demand for Process Q folks. The Flavour of The Month syndrome. But, capitalists being wily beasts (wilier than that coyote character), will read the data and find that there're scads of them available, and if the work products are dumbed down enough, they can be treated like cogs-in-the-machine (remind you of that Charlie Chaplin flick?).

Much of the direct cause of The Great Recession is little discussed, but is kind of frightening: a great many of the quants who were part and parcel of the process were refugees from math and science disciplines. Some because they couldn't get work in their area of training, and some because they chose to follow Mammon rather than God. The Great Recession demonstrated, in the background, alas, that we do produce lots of STEM folks. We just don't employee them in STEM (I, personally, don't count Bankstering as a STEM occupation). Do we really need to produce yet more Banksters To Be, just because we don't do much STEM in the capitalist nirvana?

Unlike Lake Woebegon, it doesn't do the majority any good if all children are above average; that just makes them all ... average. And infinitely interchangeable in a laissez faire world. Who knows, may be little Billy will get his BS in EE, and create the next Depression inducing derivative market. Makes a parent proud, don't you know?

23 August 2013

You Can Quote Me

Today brings a couple of juicy observations, which have been added to the Sigs/Quotes larder.

From his column today (I continue to fault him for not discussing why people continue to plunge):

In short, the main lesson of this age of bubbles -- a lesson that India, Brazil, and others are learning once again -- is that when the financial industry is set loose to do its thing, it lurches from crisis to crisis.
-- Paul Krugman/2013

And this on the NASDAQ fart:

You have a very Rube Goldberg [computer trading] system. We've just put patches on it without attacking the basic problems.
-- Gene Noser, co-founder of the brokerage firm Abel/Noser/2013

20 August 2013

Twang My Wang

Joe Nocera is generally an astute fellow, but his memory of, or research about, the world of IT is faulty. Today he looks to Wang as an earlier manifestation of all that ails BlackBerry.
He and his company stubbornly clung to the notion that the main thing people wanted from their computers was word processing...

Well, guess what? If you were around back then, or know someone who was, MultiMate (a "clone" of Wang's WP system, but standalone for the PC) and Lotus 1-2-3 were the impetus for the rise of the PC. The killer apps. Once NetWare got going, other LANs followed, and the networked PC became the reality. It was, and still is, the mantra of the PC world: "The three most important apps for the PC are word processing, spreadsheets, and word processing." Wang was right, but Wang, the company, was too expensive compared to networked PCs running MultiMate and 1-2-3. It didn't even require Windows to put the end to Wang.

Wang, along with Data General, subject of "The Soul of a New Machine" and a host of other mini-makers were mostly caught out by the transition from low level cpu design, using discrete logic parts on boards, to chips built with LSI, then VLSI, and whatever SI they call it these days. In other words, Wang, et al, were killed by the tsunami of good-enough commodity cpu and TCP/IP, as embodied by the Intel chip. And so it goes today.

The only real similarity betwixt what happened to Wang and BlackBerry is the level of specialization in their products. With the rise of the networked PC, even general purpose minis went down. A special purpose machine for word processing didn't stand much chance. If memory serves, and Wiki says so, too Wang did make some general purpose machines, even kind of big ones. But all people remember is the Word Processor. With BlackBerry, Apple turned the cellphone into a general purpose toy. The fact that corporations have chosen to take chances with less secure networking (hello, NSA!)? Who knows how that will end up. On the other hand, Mussolini had a prescription for governance where corporations and The State are (co?)equal partners.

18 August 2013

A Fish Out of Water

Robert Shiller is, at least by topography, a saltwater economist, being at Yale and all. And from most of his writings, as well. Today he seems to have gone all Social Darwinist! He recounts how he and his colleagues "innovated" some real estate indexes, and thus continued the march of capitalist innovation.

I beg to differ. In a piece from last year is a reference to some analysis of invention/innovation in the West over the last century or so. On the whole, as outlined in that piece, the pace of real invention has slowed dramatically due to the simple fact that we, as a species (at least as expressed in our scientists and engineers), have figured out 99.9% of what there is to know about the planet and universe. The Great Inventions were driven by new discoveries of the physical world, prime among them: the periodic table. Once you know what the elements are (by about 1937 the natural elements were cataloged), and the Bohr model (1913), putting them together in required ways is a series of engineering exercises.

Thus we see the rise of Intellectual Property, as a way to segregate wealth based on algorithms since we're no longer capable of finding new physical entities. As the founders knew, a society can't survive as a democracy if individuals (human or corporate) can sequester bits of math and charge a fee to implement them. Shiller did that, and got away with it. Shiller quotes a study, which itself quotes a McKinsey (about as right wing a consultancy as exists) study:
... in free-market capitalism, "from 10,000 business ideas, 1,000 firms are founded, 100 receive venture capital, 20 go on to raise capital in an initial public offering, and two become market leaders."

First, there's no meaningful free-market in the USofA. Second, the purpose of innovation isn't to make two folks really rich at the expense of the rest of the economy; it's to foster further development and provide better living for the society. Patents, now the bane of capitalism of the sort propounded by Dr. Shiller, were explicitly time limited by the founders. Owners manipulated Congress to increase the length of "protection" afforded by patents and copyrights over the years. Mickey Mouse is owned by Disney essentially in perpetuity. Apple, at least, may not be able to enforce its patent on rounded cornered rectangle. Phew!

Is this merely descriptive, or prescriptive? Is Shiller making a case for a winner-take-all economy? For an economy where 1 in 5,000 wins, and 4,999 lose? Ayn Rand would be proud, I suppose. But it's hardly an economic structure which is sustainable. Does he realize that Edison, the exemplar of the lone inventor had, for some period of time, 5,000 employees? And that, just as corporations do today, took for his name all inventions? And that Edison attempted to monopolize electrical transmission? And so on.

To start at the beginning of his piece:
CAPITALISM is culture. To sustain it, laws and institutions are important, but the more fundamental role is played by the basic human spirit of independence and initiative.

Ah, the myth of the lone inventor toiling away in his (not likely a her) garret. Not. While it is possible for a lone coder to dream up some index on a PC, real innovation these days is coming out of those behemoth corporations by salarymen. The notion that Facebook, or some real estate index which led to the Great Recession, is the sort of activity we should emulate is folly. Having friends, or data telling you that houses are wealth rockets, isn't quite the same as building a better mousetrap. Moreover, Shiller recounts his travails in 1991. Bollocks. In today's world, hardware and software needed to implement something along those lines can be had for less than $10,000. Getting the raw data is a different matter; I don't know how that part works. But the raw data isn't, and can't be, the crux of the innovation.

I do wonder, though, whether Shiller sees the irony of claiming innovation for what he did with real estate price indexing, and what others did with real estate price indexing? I mean, isn't he just as responsible for The Great Recession? He planted the seed of the flow of "innovation" (and the finance types crowed loudly that what they were doing was "innovative", no lie) which propelled us to the crash. I wonder? Has Shiller been a true Job Creator? Depends. His various companies may have been staffed, but did such staff, if any, simply move from similar positions with similar analytics' firms? How many analysts/clerks/etc. in financial firms were made redundant when they were replaced with his company's reports and data? IOW, do his companies (and all similar) actually grow the macro-economy on net, or just re-arrange the deck chairs?

Shiller, apparently without knowing it, makes Huebner's case: real world innovation has slowed to a crawl because there are few, if any, mysteries left in the real world. We turn to monetizing pictures of dancing cats and real estate indexes because we simply have nothing better to do with our time. Yikes, if I may say so. The world my grandfather was born into in 1887, in Saratoga Springs, morphed by orders of magnitude by the time he died in 1967. The same won't be true of those born in 1937. What of today's life? Radio, check. Telephone, check. Automobile, check. TV, check (yes, look it up). Airplanes, check. Rockets, check (yes again, people in them came later, of course). Nuclear age, check (implementation was an engineering exercise). Semiconductors, check (really, it wasn't Shockley).

Lasers, new (but just barely, the science predates 1937). And a few more, I suppose. But all based on science that was established. What new science has there been since 1937? And I mean NEW, not just ever more honing of existing. So, folks turn to maths, and try to sequester algorithms. Very naughty. Pythagoras' many progeny would each live very well today with the rules coders are attempting to enforce. Just imagine: a penny for each time the Theorem is implemented each day, world wide. Dat's alotta moolah.

Which brings us to the other side of the coin, Big Data. Once again, the mainstream pundits have finally caught up with me. They don't send a check, though. Here we see Shiller's notion of innovation being revealed as that man behind the curtain; a blowhard fraud. Not that I haven't been singing that tune for some time, of course.

Data is good, "Have SQL/R, Will Travel" is my motto. But, as the mainstream is now figuring out, digging in Big Data means that the needles found must be really precious needles for the exercise to be worthwhile. Not so much. Reality bites.
Christened by the World Economic Forum as "the new oil" and "a new asset class," these vast loads of data have been likened to transformative innovations like the steam locomotive, electricity grids, steel, air-conditioning and the radio.

Riiiiiiiight!! A pile of dung, with an undigested diamond somewhere encased. This is stuff NSA is mucking through. Your tax dollars at work.
The rate of productivity growth, whose steady rise from the 1970s well into the 2000s has been credited to earlier phases in the computer and Internet revolutions, has actually fallen. The overall economic trends are complex, but an argument could be made that the slowdown began around 2005 -- just when Big Data began to make its appearance.

Where's my pet rock? It's around here someplace.

As mentioned before, studies done years ago determined that "office productivity" suites led to lower productivity in offices. This is particularly true of Word style GUI applications, since so much time and energy is easily siphoned off into "pretty printing" rather than analytic thought. Why work when you can play?

One theory holds that the Big Data industry is thriving more by cannibalizing existing businesses in the competition for customers than by creating fundamentally new opportunities.
Et tu, Shiller? Just as Google siphoned off adverts from newspapers, killing same, so some other vehicle will siphon off adverts from casual search. Basing one's business on advert payments is really risky; your product is ephemeral and not even, so to say, yours.

Robert J. Gordon puts the wooden stake through the heart (you've read this before, here):
Robert J. Gordon, a professor of economics at Northwestern University, said comparing Big Data to oil was promotional nonsense. "Gasoline made from oil made possible a transportation revolution as cars replaced horses and as commercial air transportation replaced railroads," he said. "If anybody thinks that personal data are comparable to real oil and real vehicles, they don't appreciate the realities of the last century."

Die Dracula, die!!

The commercial use of Big Data appears to be mostly in service to marketing and advertising, which are, by definition, non-productive activities. Non- is not the same as un-, but isn't easily defended as an increase in an economy's output. To the extent that such use creates *new* monetary demand for related goods and services, then yes. But it's a slippery slope to count these activities equivalent to motor car production, for instance. (Aside: the same nexus occurs with gambling casinos; put one in your state, and the only way to have a net gain is to entice over-the-border gamblers, *new* monetary demand, otherwise it's just your citizens shifting $$$ from movies to craps.)
Cat videos and television programs on Hulu, for example, produce pleasure for Web surfers -- so shouldn't economists find a way to value such intangible activity, whether or not it moves the needle of the gross domestic product?

Not only NO, but HELL NO (expurgated that for those with tender mercies). BLS did so, to some extent alas, with the revision to the National Income accounts. They ain't both apples.

Staying on this path will turn the USofA into a Bermuda from sea to shining sea.

According to the BEA as a percent of GDP:
Manufacturing, 1950: 27%
Manufacturing, 2012: 12%

Again, one can play ostrich and stick one's head in the sand and assert that all is well, and things are going as planned. May be not.

14 August 2013


One of the minor themes of these endeavors is that, as physical capital (which is the only version which is actually productive, let the flames begin) is assigned to "tech" sectors (as opposed to legacy uses, such as steel mills and other stable technologies), we find that payback periods shorten rapidly. The inverse, of course, is lower real return on such capital. The implication of this, of course, is that the interest rate which can be supported by the macro-economy must fall. Ultimately, fiduciary interest can only be paid out of increase in real production in an at least stable macro-economy. Otherwise, it's a zero sum game, and consumption must fall, further diminishing output and thus returns to said capital. A death spiral, so to speak.

But I've not found much hard (ish) data. Until now. This JP Morgan report, a puff piece labeled "For Institutional and professional investor Use Only | Not For retail use or Distribution" (who'd a thunk it?) makes the following observation (page 9):
Nominal U.S. equity returns of 8% equate to average annual real returns of 5.25%, after subtracting our core inflation estimate. While at first blush those real returns appear rich, they are below the historical long-term average of 6.2% dating back to 1850, a stretch that includes a mix of bull and bear markets, and takes in two world wars, the Great Depression and a secular bear market (see Exhibit 4B).

Now equity returns, as used by Morgan, aren't directly return on real investment, but on stock shares. But close enough for my purposes. The balance of the report is an ad for JP Morgan "services" so I'm not inclined to buy their doom and gloom. On the other hand, when paybacks shorten, the only way to counter that is to monopolize, and we see that in tech.

The Wiki quotes Greenspan:
Alan Greenspan testifying at the Financial Crisis Inquiry Commission in 2010 explained, "Whether it was a glut of excess intended saving, or a shortfall of investment intentions, the result was the same: a fall in global real long-term interest rates and their associated capitalization rates. Asset prices, particularly house prices, in nearly two dozen countries accordingly moved dramatically higher. U.S. house price gains were high by historical standards but no more than average compared to other countries." [my emphasis]

So, has the GPoM been soaked up? Or is it still out there?
In their July 2012 report Standard and Poors described the "fragile equilibrium that currently exists in the global corporate credit landscape." U.S. nonfinancial corporate sector NFCS firms continued to hoard a "record amount of cash" with large profitable investment-grade companies and technology and health care industries (with significant amounts of cash overseas), holding most of the wealth.

By January 2013, NFCS firms in Europe had over 1 trillion euros of cash on their balance sheets, a record high in nominal terms.

Yes, yes it is.

13 August 2013

A Drug Suicide

The University of Phoenix is for-profit, and near the top of the list by size. The Wiki link has the story. It advertises, a lot. The current campaign stresses that there are a slew of red executive-sock wearing graduates out there, just itching to give you, newly minted alum, a job. Not a word about what they'll teach, or you'll learn. For the Phoenix grad, it's who you know rather than what you know. If that sounds a bit like what Ponzi dreamed up, I won't argue. Having "taught" at a for-profit computer school (certificates, not degrees on offer), the tale of the Wiki is deja vu. Profit tends to corrupt, when there's no cop on the beat, or he looks the other way.

Which brings us to the recent collapse of Vical. As I type, its market cap is less than half of what it was on Friday. Vical is a small, yet smaller, biotech/biopharma/pharma (it's hard to tell which term means what, these days) that had two interests: vaccines for infecious diseases, and Allovectin-7 as a cancer immunotherapy. Immunotherapy has been, but may be not much longer, the sexy path in oncology. If you surf over to its Yahoo! page, you can find links to various publishing under the "Headlines" section. "Seeking Alpha" posts links to Yahoo!, so it's prominent there. Other such sites exist. SA is essentially an organized message board where the pumpers and bashers have at it 24/7/365, with the goal of influencing retail gamblers to either buy or short stocks. Every now and again, and I can't figure out their criteria, they'll generate and archive a conference call for a company, usually an earnings call.

This time, they provide a transcript on the post-mortem CC of the Phase III trial of Vical's primary compound, Allovectin-7. You can follow the breadcrumbs on A-7 development through the Yahoo! message board, SA articles, SEC filings, and clinicaltrials.gov. Vijay Samant, CEO (not the first one) does the presentation, and takes (invited only) questions.

This is where it gets interesting:
As a company and as individuals, we have invested substantial efforts and resources into the Allovectin program for nearly 20 years.

He then goes on the characterize the trial:
We believe the results are clear and conclusive, but no margin for alternative interpretation. Allovectin, simply, did not provide the expected benefits, we do not see a feasible path forward for Allovectin and we are therefore terminating the program.

What? It took 2 decades to figure out that this compound was no better than Standard of Care (SoC) chemotherapy? If you follow through the SA articles, you find anonymous (mostly) posters who had concluded that the trial would fail. Company statements had been necessarily ambiguous leading up to the results, but the point for this essay is that the initial description of the trial by Vical had the trial ending more than 2 years ago. As the date slipped, the zealots started the "the trial is extending, and the SoC (control arm) patients can't be living longer, so A-7 has to be gangbusters' good". And so it went for years.
The key information that we released today is that the trial did not meet either the primary or secondary efficacy endpoints and the program is terminated. There is absolutely no ambiguity here, okay.

Phase III trials, particularly oncology, often fail. Phase III is the last step to filing for FDA approval, so a Phase III will have the least likely chance of making a bad drug look good. There are lots of ways to do that, generally revolving around patient screening and is expected in Phase I and Phase II trials. Also, earlier trials have as few as 1/10 the number of patients, so cherry picking isn't out of the question. Phase III trials are (almost?) always double-blind, so no one involved knows enough to bias the results (if the blinding is adhered to; there have been stories about some trials, especially outside US/EU.

Note also, no statement about searching for beneficial sub-groups, and the like. Most often, when a Phase III trial fails, the company will find that left handed redheads during the full moon did much better, so they'll pursue approval (without a new trial, of course) for the left handed redhead dosed at full moon group. 2 decades, and not even a sub-group? That's abject failure.

How could a company pursue a dead end for 20 years? Simple: it's what they do. Small bios have one good idea (they think it's good, anyway), and pound on it until it either rises like the phoenix, or dies utterly. The standard meme is that the public sector is mired in do-nothing bureaucracy, while the private sector is data driven survival of the fittest. Not always. In the case of these mini/micro-cap pharmas, the managements and BoDs are in it for the sinecure. Most put most of their shares in retails hands; the pros don't bother. Rarely, GenVec is a current exception, will a board throw in the towel after total failure. It just isn't in their profit driven self interest.

Here's a bit of exchange:
Eric Schmidt - Cowen & Company

Okay. And one more question and I am not trying to be harsh, I appreciate your personal commitment to seeing Vical through, but I do have cash, it's been well over 20 years if Vical has been working on DNA-based delivery technology and still no major successes to show for. When you and the Board discuss the potential future of the company, will you include in that discussion things like basically winding down operations and just trying to monetize the existing partnerships you have and give back cash to shareholders?

Vijay Samant - President and Chief Executive Officer

The answer is no Eric, because if you go back and look this failure has nothing to do with the DNA platform, it has to do with the mechanistic approach of how this cancer mechanistic DNA based application is going to work and it didn't work. ...

Note, while I didn't hear the inflection in his voice, Samant appears to see no irony in the fact that it took 20 years to find out that the *basic mechanism* avowed for A-7 simply doesn't exist.

So, what's the moral of this sad tale? I'd say: biostats and scientists are an emasculated bunch. There is a continuing stream of press on the "fact" that it takes $1 or $2 or $4 billion to bring a drug to market. Only a team of forensic accountants can determine whether any of those numbers is true; how much of that $1 billion was management salaries, expenses, bonuses, consultants, and the like charged back? How much was wasted on deadends, just because Dr. Yes wouldn't concede that compound Q doesn't work, but he had enough juice with the Corner Office to continue? How much was due to trial designers succumbing to both the pressure of The Suits to continue and the need to pay for little Billy's braces?

My memory is foggy, but it tells me that before the liberalization of patenting, most drug development was directly funded by the government. Perhaps we need more enforcement of Phase III type gating early on. In the case of Vical, its 15 minutes of fame are about over, and won't be long remember, except for those who bet their 401(K)s on it. If the public, particularly investors in such outfits, had access to the full datasets of all pre-clinical, clinical, and trials, this sort of nonsense wouldn't happen. We do, after all, have R for everyone. There's the famous quote "given enough eyeballs, all bugs are shallow." Drug companies have been fighting opening their data for quite a long time. Patents, and all that. Trial data is different. All of it should be publicly available. Such a tack would also move to dispel distrust of FDA, too.

There'd likely be fewer jobs for techy types, and suits, since the Vicals wouldn't provide employment for 20 year stretches. But, is that a bad thing? If the USofA is supposed to be the one place on Earth where merit trumps birth or connections, shouldn't we have rules to make it so?

12 August 2013

VT Phone Home

Well, The Yellow Brick Road has been assaulted by sinkholes and mudslides of late. Should we fight on? Or capitulate to VSAM? I say... fight on.

Here's a (partial; I'm not omniscient, though I try) list of calamities.

SSD makers.
Fusion-io, Stec, and OCZ are all in deep doodoo. Fusion-io can't make its Wall Street numbers, and appears too dependent on Apple and Facebook and perhaps a few others. Stec is now part of Western Digital, so who knows. OCZ still hasn't figured out its numbers. On the other hand, there remains a slew of niche, mostly private, pure-play SSD companies. I got on the mailing list for one of them, SolidFire, and their latest mailing is for something called the SF9010, claiming to "At full scale of 100 nodes, the SF9010 is the world's largest, fastest all-SSD storage platform on the market - available for less than the cost of traditional performance disk solutions."

Samsung has announced a vertical, higher density, NAND. Samsung and Micron, possibly with Seagate and Western Digital, are likely going to out-vertically integrate production of NAND, and dependent devices. Just as Content, by some lights, controls the consumer device market (say goodbye, Apple), so too will the physical investment in hardware control the device market. Apple was happy with the rich 20%, and the strategy worked. Until it doesn't.

The knuckleheads have a head of steam, no doubt about it. But thinking people understand that an application API world is just fancy talk for siloed data. The RM sought to change that, among other things. And data without ACID is just a lump (I typo-ed that as 'limp'; where's Freud when you need him?) of bytes. Those that ignore history are doomed to repeat it. Now, the basis of the NoSql argument is that a database engine can't keep track of all that data to be searched (which is where most of this comes from). So, it's back to dumb files and smart code. So, such coders are forced to write a simple minded TPM, or ignore the whole problem.

Oddly, Postgres folks are well into Foreign Data Wrapper support. This is called Management of External Data in the 2003 SQL, but has been around as Federation since at least 1990. Progress, not exactly a pure RM/SQL engine of course, could do it then. So, there is hope.

The PC is market is dying!! The PC market is dying!! Remain calm! All will be well?

What Apple started, or more accurately continued, was the revelation that non-business computer users didn't really need a 'computer' and only had one because the applications (not the programming infrastructure) of interest ran on one. Apple transformed itself from a computer company, which built its machines (almost) wholly from bought-in parts, to a toy company with ever more powerful embedded controllers. And consumers realized that they didn't need computers because they didn't create computer programs. D'oh!! Lotus 1-2-3 was the first step down the slippery slope; if one ignores VisiCalc. Either way, IBM's notion that the Personal Computer would be used by engineering types to write programs quickly, for themselves rather than wait for the crew running the mainframe to churn out COBOL or FORTRAN, turned out to be terribly wrong. Well, and letting the BIOS loose.

The spreadsheet meme, thanks to 1-2-3, has given us naive' views of data, and the London Whale. Not to mention that much of financial services, across the board, runs on spreadsheets. More than any other factor, the spreadsheet has been the leader of the assault on the RM and SQL.

I see glimmers, ever so faint, that the realization that large format keyboard applications don't work so well on small-ish devices sans 'real' keyboards. Touch means picking, and picking means listable data. Listable data means narrow definitions (tables). And, finally, narrow tables means Organic Normal Form™. Back about 1990, Phil Johnson mentioned in passing that he'd never trust a schema that was shallow and wide; a few tables with lots of columns in each. He wanted to see lots of tables with a few columns in each. He never mentioned the RM or Codd or the like. He intuitively understood. Yes, Google will still do its thing. But operational systems really can't drop ACID; even if Date suggests that such would be OK with him.

The days of centralized data, on a passive terminal device (VT-X00, for example), are back. I continue to be amazed how slowly the Well Known Pundits are coming to this realization. But they will because they must. It's not nice to fool Mother Nature.

08 August 2013

A Quantity of Health [update]

More than a year ago, I made the point that even the 1% will need Obamacare, in due course. As most regular reader knows, quants earn their livings in various ways. In the private, for profit, world that usually means coercing the naive' into spending money on widgets they wouldn't have done save for the exercise of the quants' magic. All those ads and "suggested" products are mined from user data. The term usually applied to this endeavor is "market segmentation". When used to entice an RC Cola buyer to also buy Moon Pies, no harm done (well, to the extent that later diabetes can be disavowed).

Anyway, as in the prior essay, buying insurance (health or auto or home) isn't the same as buying that box of Twinkies or an iPhone. You're not buying a "thing" to consume. Well, so long as you're not a Right Wingnut with a warped sense of consumption. Buying insurance is paying into a pool of shared risk. The more who share in the risk, the lower the cost to all. Well, if profit isn't involved. What corporations do: socialize cost and privatize profit. Left to their own devices, they'll remove (through price arbitrage) all but the most profitable. In the prior piece, I expounded on the likelihood that health care, beyond the simplest 19th century doctoring, would cost too much even for the 1%. Just too few "consumers" to spread the fixed cost.

In the case of property/casualty insurance, we've seen the result here in the Northeast US. The insurers (including FEMA, which runs the national flood insurance program) are raising rates, and re-defining storm zones, so much that much of the coastline is too expensive to insure. What's really annoying is that the issue wasn't an issue when it was Red States that got all that implicit support from the rest of the country. But let a once-in-a-century (or two) storm happen in Blue State territory, and all hell breaks loose. "We can't afford all those freeloaders." In due time, all those Red State House members will see to it that their constituents aren't priced out. They do control the House, you know. Whether the Blue State residents get equal treatment? Doubt it.
One effect seems to be clear: Many people around the country may be considering walking away from their insurance, calling into question the premise of the legislation, which was to make the flood insurance program financially sustainable.

(Nailed it!!)
On Tuesday, the House passed a $50.7 billion relief package for Sandy. This time, 180 representatives voted against it -- 179 Republicans, one Democrat -- 56 of whom had voted for the similarly sized Katrina bill.

Now, consider that the analog for health insurance is: you die.

Well, now the mainstream press has delved into the issue, and one has come up with an historical example.
Younger and healthier workers canceled their P.P.O. plans, enrolling in cheaper H.M.O. options or dropping Harvard insurance altogether. Left with a sicker patient base, the P.P.O. raised its premiums further, which prompted the next layer of relatively healthy customers to leave.

And so on. In 1997, Blue Cross/Blue Shield withdrew its P.P.O. from the market, making it a victim of what economists call the death spiral of adverse selection.

Just what the Right Wingnuts want: kill off the sick and damaged, leaving the rich and healthy with cheap insurance. The only problem with the plan is that, in due time, first the 1% will be found to be too sick and damaged and thence the .1%. In due time only Bezos and the Koch brothers will be able to afford the insurance. One might speculate that for those, insurance per se wouldn't even exist. They'll just buy up a personal physician and hospital. Kind of like the 17th century. And all will be well.

The fact is: insurance is socialism. That the USofA is mostly for-profit now (it was not always so, by any means), turns insurance into Twinkies. With all that implies. Kill off the poor and sick and damaged. Don't wait, times a wasting. The Chinese indentured labor makes our iPhones anyway; who needs any more poor Americans?

My, my. Don't believe what I say, believe what I do. Now, the folks who did the study are described in the reporting as "...the Commonwealth Fund, which strongly supports healthcare reform...", but still how would you go about finding left wing Republicans? The closest cadre I can think of: Fairfax County Virginia, home of all those public trough lapping good-for-nothings. In any event, the study found that with respect to the 26 years of age provision,
They found that by last March, 63 percent of young adults identifying as Republicans had enrolled in a parent's health plan in the last 12 months, compared to 45 percent of those who considered themselves Democrats. About 26 percent of the 1,800 adults surveyed said they were Republicans, 28 percent said they were Democrats and the rest either said they were independent, some other party, or did not say.

And the gut shot (if you're a Right Wingnut, anyway):
"There is a stereotype that young adults believe they are 'invincible' and don't want or need health insurance," said Collins. "This survey shows that is a myth -- typical uninsured young adult is from a low- or middle-income family and works a low-wage job. In general, young adults value health insurance but cannot afford it."

"Enrollment rates of working young adults in their own employer-sponsored plans average nearly 70 percent, with cost being a principal factor cited among those who do not enroll," the report reads.

In other words, yet again, insurance is shared risk. Broaden the pool, and one lowers the individual burden. Segment the pool into ever smaller and homogeneous pieces with prices assigned based on historical risk in each sub-pool, and you drive out the "less desireables" through the simple expedient of increasing the burden on those least able to bear same. And this works fine for a while, particularly in non-health products. And has a certain righteous appeal to the Social Darwinist cabal. But with the advances in health care coming almost exclusively with (increasingly expensive) treatments, not (inexpensive) preventions (the original HMO approach didn't really work), even the 1% won't be able to carry the fixed cost burden. The problem is that when your pool shrinks too small, and your ability to sequester more of the GDP wanes, even you won't be able to afford an MRI. Or some such.

04 August 2013

Feathering The Nest

Along with the stat stuff, this endeavor is mostly about replacing code bloat with svelte relational database schemas. While part of the motivation is DRI, part is pure efficiency. Sad to say, accountants have figured this out first. Who'd a thought accountants were smarter than computer geeks?
Perhaps the biggest problem, though, was that billing by the hour incentivized long, boring projects rather than those that required specialized, valuable insight that couldn't (and shouldn't) be measured in time. Paradoxically, the billable hour encouraged Blumer and his colleagues to spend more time than necessary on routine work rather than on the more nuanced jobs.

Substitute "long, boring" with "convoluted spaghetti code".

The BLS is also in the midst of increasing the GDP number by "counting" movies. Who knew? Whether an economy which makes movies, and has a higher GDP than an economy which makes food, clothing, and shelter for its citizens, is the superior economy? We can't all be parasitic economies like the Caribbean and European money laundries.