30 September 2014

Apple is a Great Source of Iron

Does anyone else find it just the slightest ironic that Apple would abandon its "patented" solid rectangle with straight sides and flat front/back for something much more "round" in all three dimensions? And tout its thinness? And then, to top it off, provide a Stay Puft marshmallow man silicone case? Kind of like putting a Santa Claus suit on Twiggy.

I Want It All Done Now!! Do You Hear Me?!

Codd is reputed to have said, in some manner of words, that the RM means a query (whether read or write) is "done all at once". This makes sense, given that the RM is built on set theory and expressible as a linear algebra problem. Just as the regression equation can be written in matrix notation (atomic operations on multiple values), so one can view RM operations. We all know, of course, that real computers are sequential processors most of the time most places. Until now, again. HP has gone and done it.

Here, and here. One says Informix is an option, while the other says it's part of the ship. Informix has been into parallel for some time. Lots o gerbils spinning in those cages can do a lot of work in toto. No, not in Kansas. See Chapter 12, here.

Dr. Codd was right, after all. He he.

24 September 2014

Where Did I Leave My Keys??

Quants aren't just used by the Masters of the Universe to crash everybody else's immediate vicinity (Hank Greenberg's attempt to get his AIG punishment overturned is about to go to trial; chutzpah of the highest water). They're also the backbone, so to speak, of clinical trials of the Next Magic Bullet drug/device/therapy. FDA swears by p-value.

Adam Feuerstein is something of an enfant terrible to retail plungers in penny stocks; forever calling "bullshit" on their next 10 bagger. But, to his credit, he sometimes gazes at larger companies when they're working on interesting compounds. Today he's got an ADHD company (that's the stock symbol, too) to talk about. What I find interesting is the large impact that placebo makes in the study. One has to wonder what's being measured?

Welcome Back Gumby [update]

Some time back, I predicted that the flip phone would rise, Phoenix like, from the mire of smartphone muck. These 5+ inch behemoths just won't work very well. Some curved screens have appeared, but now we get word that the 6/6+ are spontaneously bending. I told you so. There are certain female tennis players who'd turn a 6+ into a crescent moon if stuffed long enough in a pair of Spanx. So to speak.

Now you know why they didn't use sapphire for that screen!!

I can't find that any of the usual suspects which tear apart these devices have specified what type of aluminum is used. As the Wiki shows, there are a raft of hardened and alloyed aluminums. It seems that "case hardening" isn't used with aluminum alloy although the effect is much the same, but "through hardening" is. Me thinks Jony and Tim got it a bit wrong.

[another update]

18 September 2014

15 September 2014

Up The Down Staircase

I'm not a fan of David Stockman, to say the least. But recent punditry on the subject of interest rates, asset prices, and bubbles led me to the obvious question: is it true, as it appears, and in the data; that corporations (that is to say, their capital allocators) are executing share repurchases in excess? Which is to say, as I have said more than a few times, that the growth in share prices is the result of intrinsically lowered returns on real investment? Which is to say, further, that the Masters of the Universe CxO types simply aren't generating any return?

So, of course, I went wandering on the innterTubes asking for the value of corporate repurchases. Alas, a Stockman piece came up first. Ever more alas, since he uses the data to make his usual wrongheaded conclusion. He hasn't actually learned much since his defense of Voodoo Economics.
Self-evidently, the corporate form of business organization is designed such that some considerable portion of net earnings should be returned to their owners each year. But a 95% rate of distribution is a giant aberration. Were this outcome to occur on the undisturbed free market, for example, it would signal an economy that is dead in the water and that participating companies face a dearth of opportunities to reinvest profits in future growth.

And, of course, that's exactly what's been going on. As the science and engineering geeks figured out some time ago, we're in the post-discovery age, of marginal incrementalism (kind of like, very unique). With real interest earnable near zero, then share prices (and bond prices, to be clear) go up to meet this rate. Or, as Vinny in those mob movies used to say, "who's gonna give ya a better deal, huh?" Lots of Chinese and Germans and such thought that Florida and Costa del Sol real estate was a sure thing. Not.

Or, as another pundit put it:
Corporate CEOs, with their massive share-buyback programs are in effect investing in the stock market rather than in expanding business opportunities at their companies. Either they expect higher returns from the market, or lower returns in their business, or some combination of both. Given their questionable track record in timing the market, this may be a cause for concern.

It's one thing, although wholly foolhardy, to pay Goodell $44 million to "oversee" the behavior of his employers, since it's just football. Quite another for the Masters of the Universe to get paid such sums and not actually do anything.

10 September 2014

Gorilla My Dreams

Lots of whining about no sapphire display in the 6s continues. This is especially true of the plungers who expected to get that cottage in Bimini off of the GT Advanced Tech stock they bought cheap, cheap a year ago. Sniff.

The likelihood that any company is foolish enough to make a pure sapphire screen for 5+ inch phone is just about 0.0 (well, may be not; no information on how the screen is built and not coming to a Best Buy near you any time soon). Why? The modulus of elasticity for sapphire is 345 GPa, while for Gorilla Glass (don't know whether all versions) is 71.5 GPa. It's called Young's modulus E (not, alas, in memory of yours truly), and it matters. It means that sapphire lacks ductility; twist it or bend it, and it shatters. This is well known. The reason it can be used in conventional watch faces is because a) they're generally 1 to 1&12; inches in diameter and b) the cases are themselves far more rigid than a 5+ inch phone. While a regular watch face, sapphire or mineral glass, can be punctured, torque stress is effectively zero in normal use. Can't say quite that for phones which spend most of their time in the ass-end pockets of pimply adolescents.

There would appear to be two ways to use sapphire for a 5+ inch phone screen. Either fabricators of sapphire figure out how to shave it off boules wafah tin, thus without wasting most of the boule in the process, and, figure out a way to laminate it to (Gorilla?) glass. Or, the engineers figure out a way to vapor deposit sapphire to glass. A quick innterTubes search yielded lots of links to deposition on sapphire, but none to deposition of sapphire. Which isn't surprising. Although... the innterTubes does show lots of papers (most for $$$) describing vapor deposition of aluminum oxide, which is the base molecule of sapphire; not crystal, though.

As to the "slicing" bit, turns out, they have; basically, boil sheets off the boule. Very cool engineering!! Hot off the presses, too.

Here is a piece on Apple patenting a laminate. A year ago.
The use of a sapphire outer surface with a glass inner surface for the display may be used in future iDevices where the two sapphire surfaces are laminated together with the glass providing support for the display and the sapphire providing scratch resistance and durability advantages.

I swear; I wrote the first paragraph before I stumbled on the Apple patent. I swear to the Great Jehovah. In any case, Apple/GT couldn't git er done for the 6s.

Anyway. The reason sapphire becomes practical in laminate, is that for such thin sheets, bending stiffness improves ductility. Somewhere in materials' science I expect there's a bunch of algebra which demonstrates that a sufficiently thin sheet of sapphire is more elastic. Durned if I could find it though, but it sure looks like Apple thinks so.

Our 7% Solution

The good folks at simple-talk asked me to take a swing at some topic, and I've sent off a draft. We'll see how it works out, but, coincidentally, Apple has had its big soiree, and the Watch and Pay were prominent. The 6 not really so much, being well, the 6th in line. Yeah, so what's new? No sapphire, either.

Well, one Michael Hiltzik of the LA Times has this to say about Pay:
The late, great social critic Neil Postman used to pose a question that had to be asked whenever a new technology appeared in the marketplace: "What is the problem to which this is the solution?"

It was a rigged question, of course, because more often than not the answer was "Nothing."

Turns out, I reached exactly the same conclusion in my musing on the requested topic. Because, more often than not...

We'll have to see how strong Steve's residual Reality Distortion Field is, since neither Watch nor Pay look to be much of a solution.

09 September 2014

Sleeping With Godzilla

Well, the Apple reveal is about over. Two points are made clear. First, the law of large numbers (of which there are a small number) is biting Apple wannabe billionaires in the tusch. It's much easier to sustain a double digit growth rate when you're the little fish in a big pond; lots of other little fishies to eat. Get to be the largest fish in the only pond, and you'll soon run out of other fishies to eat. Neither of the iPhones strikes me as a gotta have. But then, I use a mobile phone to make phone calls, not support a distraction habit. A failure on my part, I admit. These new phones aren't likely to be much more than replacement buys by current zealots.

Second, the Watch (not, stressed by some, an iWatch) is fully underwhelming. Suunto et al have had similar for decades. And the strap appears to be generic plastic (news: also leather or stainless, all for the same price???). Brightly colored, of course. But, still, plastic. It ain't what iPod was.

But, as to the title: GT Advanced Technologies which makes both sapphire tools and sapphire (the latter mostly newly), some months back announced a "deal" with Apple. At first blush, the reaction was "Apple!! Let's go buy that yacht!!" On further reflection, the "deal" turns out to be mostly a loan against deliveries, for which Apple is not obliged to make any purchases. So, over the months, the bulls and bears argued about where Apple would use all this sapphire that GT was obliged to hold capacity to deliver (but, remember, Apple wasn't obliged to buy). The 6, of course, was the bonanza. Not only larger screen (4.7 and 5.5, in turns out), but would sell by the millions. As I type, the nature of the screen of the 6 is still being argued. According to various live-blogs (the stream, of course, crashed), it's an updated Gorilla Glass (Corning), not full sapphire. The Watch is being described as sapphire, but it's teeny in size and of unknown quantity. GT's share price has been pounded, headed back to the LoD.

Don't go to bed with Godzilla. You'll likely get eaten.

05 September 2014

Do You Believe in Magic?

As mentioned a few times, I've gotten on the SolidFire mailing list, which means a blurb every few months. Today's touts the Gartner report on All Flash Arrays Magic Quadrant designation. Gartner gets paid a bunch to bloviate, which is nice work, if you can get it.

Turns out that Hitachi didn't get a mention in the report, so The Register gets itself a story. There's a link at the top to a screen grab of the quadrant, but the main story is how Gartner chooses to slice-n-dice storage.

Also of interest: Nimble, with a market cap today of $1.92 doesn't get a mention, either. They do hybrid drives, so I suppose not letting them in makes some sense, but hybrids may well end up being the winners. We'll see.

Alice In Wall Street

Financial engineers, despite their pronouncements of math/stat/coding wizardry, are really just some truffle pigs and cockroaches. The truffle pigs snuffle in the ground, generally around trees in deep parts of the forest, for the scent of the black truffle. A valuable fungus, that black truffle. Finding one, the handler has to quickly extract it from the maw of the pig. Turns out, unlike retrieving dogs, pigs like to eat their catch. Cockroaches insinuate themselves in the darkest parts of dwellings, seeking water and the occasional crumb. Few of the mainstream pundits have taken this bull by the horns. Today, Floyd Norris, reliably more aggressive then 99.44% of his brethren, lays out the story. Again, and with some rather wonderful quotes.

Not to mention: I've referenced Lewis Carroll more than once, generally repeating the phrase "... words mean what I say they mean...". Which isn't exactly the text.
Part of the regulatory challenge was laid out in 1871 by Lewis Carroll's "Through the Looking-Glass":

"When I use a word," Humpty Dumpty said in rather a scornful tone, "it means just what I choose it to mean -- neither more nor less."

"The question is," said Alice, "whether you can make words mean so many different things."

"The question is," said Humpty Dumpty, "which is to be master -- that's all."

Norris spends most of the column with the history of Lehman, and Repo 105.
Lehman taught us that there were plenty of tricks around to make balance sheets look better for one day and then revert to an undisclosed reality that was much worse.

One of the tricks was to adjust the weights of assets; as in weighted average. Of course, analogous to a Bayesian prior, the banker gets to make up the weights. And that's what Lehman (and the rest as well, of course) did.
It turned out Lehman was valuing some securities at 85 percent of face value while competitors thought they were worth 20 percent or less.

In general, the bankers do have a point: some assets are riskier than others. Home mortgages, for example, won't go rotten all at once all over the country. Will they? So the banker puts a lower risk weight when figuring the backing capital.
That made sense in theory -- some assets are clearly riskier than others and more likely to produce losses. So less capital was needed for low-risk assets, like loans to high-quality borrowers. Trusting regulators let the banks use their internal risk models to determine the weightings.

But, quoting Stanley Fischer, Fed vice chairman
"... any set of risk weights involves judgments, and human nature would rarely result in choices that made for higher risk weights."

But, here's the pig/cockroach bit:
Financial engineers invented securities that would count as equity for bank regulatory rules and balance sheets but looked like debt to the Internal Revenue Service. Unfortunately, when the pinch came, they turned out to be more like debt. Big banks are being forced to stop using such things.

Not satisfied with just putting a heavy thumb on the scales, they dispensed with the scale entirely and made up the numbers.

So, are financial engineers good for the commonweal, or only for themselves and their handlers?

As the song says, "Mama don't let your kids grow up to be banksters".