Comments, observations and thoughts from two bloggers on applied statistics, higher education and epidemiology. Joseph is an associate professor. Mark is a professional statistician and former math teacher.
Very sharp piece by Felix Salmon (one that's also extremely relevant to our technology and hype megathread) on how the Amazon narrative drives the stock price and vice versa. Among other things, Salmon effectively proves that the Amazon-as-disruptor effect -- where business journalists invariably attribute every development to Amazon -- is almost always bullshit.
The sad part is (and I in no way mean this as a knock against Salmon) there's no reporting here. Felix is simply pointing out facts that are and have been widely known for a long time. The reporters who have kept this narrative going knew (or should have known) that the numbers completely undercut it but they chose the good story over good journalism.
Amazon is a perfect case in point. For one thing, its market capitalization of $814 billion compares to net income, over the past 12 months, of $3.9 billion. That means it’s trading at a price-earnings ratio somewhere north of 200. Compare that to a ratio of about 24 for the stock market as a whole. If Amazon traded on the same multiple of earnings as everybody else, its stock would fall by roughly 90 percent.
The stock is supporting a narrative, which in turn is supporting the stock. And the narrative is, frankly, looking a bit shaky these days.
...
Amazon similarly never fails to appear in just about any story about the so-called retail apocalypse—a phenomenon that is much, much bigger and more complicated than “people used to shop at stores, but now they just order from Amazon.” After all, Amazon still accounts for only 4 percent of American retail sales. Indeed, with its Whole Foods acquisition, Amazon clearly considers in-person shopping to be entirely aligned with its mission of customer service and convenience. The retail apocalypse is, at heart, a real estate story (too many malls built in areas where no one wants to shop), much more than it is an Amazon story.
Here’s a game you can play along at home: Pick a story, more or less at random, about any company that sells consumer goods and that is doing badly or worse than expected. Then count the paragraphs before the word Amazon appears. Take this story about Procter & Gamble, Unilever, and Nestlé, for instance; the fourth paragraph has the requisite clause about “Amazon.com Inc.’s rising prowess in selling more household staples.” But there are no numbers attached to that claim, maybe because Amazon, even if it’s growing, still has less than 1 percent of household-staples sales.
And while Amazon’s entry into the fashion business is certainly noteworthy, that doesn’t automatically mean it’s “disrupting fashion retail,” as the Financial Times would have you believe. That would be hard, with just $25 billion or so in revenues—just 2.5 percent of a $1 trillion market.
Michael Hiltzik also has some worthwhile thoughts on the subject, and he even works in a Dr. No reference.
One of the concepts that features heavily in the upcoming technology book is the idea of ubiquitous explosive change. When people wonder at the pace of progress today, they normally point out some aspects of lives that have radically changed, but in the late 19th/early 20th centuries (and to a lesser extent, during the postwar era) the challenge was finding an aspect that did not.
We can quibble about start dates and stop dates, but certainly during the period between, say, 1880 and 1905, inventions and advancements
Inventions and advances were so big and coming so fast that innovations which would have been at or near the top of the list for the decade are often neglected and sometimes even forgotten. New line For example, in any other era, making the biggest improvement in decade to what was, at the time, the world's only mass medium would have been a pretty big deal.
The machine revolutionized typesetting and with it especially newspaper publishing, making it possible for a relatively small number of operators to set type for many pages on a daily basis. Before Ottmar Mergenthaler's invention of the linotype in 1884, daily newspapers were limited to eight pages.
When you see one of those pretty, quaint yet ingenious and functional turn-of-the-century aircraft, chances are it came from this guy.
Alberto Santos-Dumont; 20 July 1873 – 23 July 1932, usually referred to as simply Santos-Dumont) was a Brazilian inventor and aviation pioneer, one of the very few people to have contributed significantly to the development of both lighter-than-air and heavier-than-air aircraft.
...
In 1904, after Santos-Dumont complained to his friend Louis Cartier about the difficulty of checking his pocket watch during flight, Cartier created his first men's wristwatch, thus allowing Santos-Dumont to check his flight performance while keeping both hands on the controls. Cartier still markets a line of Santos-Dumont watches and sunglasses.
From scientific American, 1905/10/14
And from Wikipedia (circa 1900)
When I started on this post, I immediately thought of a song a friend of mine named Jerron had introduced me to, so I did a quick Google search for the title and guess what popped up...
Sometimes, when you've been working on a theory for a long time, you come across an example so apt, so overly on-the-nose, that it makes you doubt yourself.
Case in point, one of the theses of the upcoming technology book is that we have collectively bought into the idea of constantly accelerating progress to such an extent that people are (consciously or unconsciously) starting to distort the historical record in order to keep that record consistent with the myth.
Among the problems you run into trying to fit a nice
exponential curve to technological progress over the past 200 or so
years are the substantial spikes in innovation around the late
19th/early 20th centuries and the Cold War era. The causes for these
surges are complex but the short and very incomplete version is internal
combustion/electricity/lucky breaks in the first case and Cold War
cash/deferred demand/lucky breaks in the second.
The
only way to make the data fit the curve is to drop lots of important
accomplishments from the surge periods and/or greatly lower your
standards for more modern accomplishments.
As I near the end of the project, I came across this:
In its recent Equity Gilt Study, which is a massive annual report by Barclays chronicling the bank's thoughts on important topics in finance and economics, the bank focused heavily on new technologies and particularly on cryptocurrencies and artificial intelligence.
I read over this list a number of times, trying to find something I had missed. I couldn't believe that neither the airplane nor the telephone would make the list but both Bitcoin and the Bitcoin ATM did, that the creation of recorded media is ignored -- neither photography, phonographs nor motion pictures make the list -- but they find room for three companies (Netflix, Apple iTunes, YouTube) that distribute that media. Telstar didn't make the cut either.
I'm tempted to say something snide about people who rely on Barclays or Business Insider for investing insights, but the inability to have an intelligent, historically literate discussion about technology is not limited to one institution or publication.
There's a lot of interesting and important stuff to discuss in this article by Derek Thompson, both in terms of the implications of Disney's starting a streaming service and in the ways that Thompson's reporting is shaped and often distorted by the need to adhere to a standard narrative. Unfortunately, I really don't have time to delve into the first at all (maybe later) and, for the moment, I'm just going to look at one particularly egregious paragraph to illustrate the second.
No company has been more responsible for shaping the modern entertainment landscape than Walt Disney. In 1937, with Snow White and the Seven Dwarfs, its first feature film, Disney invented the family blockbuster. In 1954, with Disneyland, an anthology series hosted by Walt Disney himself, it became the first movie studio to strike out for the wild west of television. Since then, Disney’s dominance has only grown. Of the dozen films with the largest worldwide box-office take since 2010, Disney released eight.
While the dates and numbers are correct, pretty much everything else in this paragraph excluding the final sentence is wrong. It is very much the standard narrative and, not coincidentally, one which Disney would very much like you to believe, but it's not the way it happened.
With all due respect to the success and influence of snow white, saying that all Disney invented the family blockbuster is simply silly. Loads of counterexamples here, including the films of arguably the biggest star of the time, Shirley Temple.
The claim that Disney was the first studio to make a major play for television is even worse. RKO set up a television division in 1944 when the industry consisted of a handful of stations, a full decade before Disney got into the business. Through its stake in the Dumont network, Paramount had been there since 1946. Warner Bros. had been flirting with the business for years and would finally hit on a tremendously successful slate of TV Westerns shortly after the debut of Disney.
None of this is meant to take away from Disney's remarkable accomplishments in the medium, particularly the unprecedented impact of the Davy Crockett shows, but the version told here is simply bullshit. You can make the case for Walt Disney being the first major film producer in the field (assuming you're fairly specific about “major”) and certainly for the man being a television pioneer, but the standard narrative of innovation, disruption, dominance, simply doesn't fit the facts.
This sketch from the largely forgotten HBO show Likely Stories (h/t Mark Evanier) hasn't agedd that well -- these things seldom do -- but it has some fun moments and it fits nicely with some of our ongoing threads.
We previously discussed Turn of the Century scientists announcing major discoveries only to have the effect sizes later turn out to vanish entirely. This is probably a better example.
We've been over this before (repeatedly). Virtually every story you read or hear about Netflix -- whether it's a review of a new show, an interview with a star, or a report on an award -- is largely the product of a massive PR campaign. One of the key components of that campaign is the relentless carpet-bombing of LA with billboards and bus signs.
Admittedly, everyone plays this game, but not to the same degree. Among the major networks, spending seems inversely related to success. Little for CBS, more for NBC, but none of them spend that much. Certain cable networks (AMC, FX, TNT) maintain a big presence as do the pay channels (HBO et al.). The biggest players per show and perhaps overall are the streaming services with Netflix leading the pack.
I knew a huge amount of money was being spent on this, but I didn't realize it was this huge. [emphasis added]
Netflix is offering $300 million to acquire Regency Outdoor Advertising, a company that owns the familiar billboards along Sunset Strip, at Los Angeles International Airport and near the UCLA campus, Reuters reports.
The streaming TV giant plans to increase spending on marketing its original shows and movies to $2 billion this year, amid growing competition from technology companies such as Amazon.com, Facebook and Hulu, as well as from traditional media companies like Disney that are investing in their own streaming services.
Billboards are clearly part of Netflix’s promotional arsenal, with Sunset Boulevard adorned with images touting Stranger Things and The Crown. Should the transaction be completed, HBO and Showtime would seemingly need to look for other locations to promote its shows.
So why do we expect reading scores to be skyrocketing in the first place? Why do we almost universally refuse to acknowledge that scores are up at all, let alone up a fair amount? Why are we so determined to believe that kids in the past were better educated than kids today, even though the evidence says nothing of the sort? It is a mystery.
My opinion: because there is a lot of money in education and it won't be possible to "disrupt" education and redirect this money if the current system is doing well. Notice how there is always a lot of money in being a disruptive company, at least for the top management (see Uber -- it is clear that it pays better to run Uber than it does to run a traditional Taxi service).
It also moves the goalposts. If everything is falling apart then it isn't such a crisis if the disrupted industry has teething issues once they strip cash out of it to pay for the heroes who are reinventing the system.
But if current educational systems are doing well, and slowly improving through incremental change, then it is a lot harder to argue that there is a crisis in education, isn't it?
A few years ago I decided to take some time off and focus on writing. Now, with some big projects either out of the way or nearing completion, I've decided it's a good time to get back into the saddle. I miss the challenge of actually digging into real data and I don't want to let my analytic muscles atrophy.
Most of my experience has been in data mining and predictive modeling primarily with large to very large data sets working in SAS and R with the occasional detour into Python. I've also done some work with text mining and Bayesian networks and would like to explore that area further given the chance.
If you know of a position that sounds like it might be a fit, please contact me at the Gmail address consisting of my first and last name followed by WCSV, no spaces. If you know of someone who might be looking for someone, please feel free to share that contact information.
Thanks and now back to our regularly scheduled blogging.
The article does a superb job getting in the heads of the subjects. The details are sharp and informative and the quotes are often unintentionally revealing. The problem is what the piece doesn't reveal, at least not as plainly as it should.
When telling an account of misguided (and in some cases even delusional) people, there's always a bit of a balancing act between the desire to let the story tell itself and the impulse to jump in and point out important facts. The article (also from the New Yorker) on the essential oil industry that we spent quite a bit of time discussing last year did an almost perfect job maintaining this balance. Between the narrative and context, the damning conclusions were all but inescapable.
Perhaps the problem here is length. The piece is simply too short for the nuanced story it needs to tell. As a result, we get a beautifully drawn picture of a group of people pursuing a dream, but no real indication of how unrealistic and even dangerous that dream is.[emphasis added throughout]
Between mid-December and early February, bitcoin lost more than half its value, dropping from a high of nearly twenty thousand dollars to just below seven thousand. Depending on whom you asked, it was either a catastrophe—a portent of things to come—or a rare opportunity. Anthony Pompliano, a venture capitalist who is prone to posting bullish, cryptocurrency-related aphorisms on Twitter (“Bitcoin is the ultimate test of someone’s imagination”), reassured his eighty-three thousand followers that it was almost certainly the latter. “This may be the first real ‘crypto recession,’ ” he wrote. “Those that stick around will be rewarded immensely.”
...
By many counts, “literally every guy in crypto” is pretty much everyone in crypto, at least for the time being. A handful of surveys and studies estimate that women make up somewhere between four and sixteen per cent of cryptocurrency investors. Morin, during her introductory remarks, explained that she had heard the four-per-cent figure over the recent winter holidays, when bitcoin was valued at nearly twenty thousand dollars. Part of the problem, she determined, was a paucity of educational resources for women about the fundamentals, and risks, of investing in cryptocurrencies. “We have an opportunity to rebuild the financial system,” Morin said, quoting Galia Benartzi, the co-founder of Bancor, a cryptocurrency protocol; protocols, like Bitcoin or Ethereum, enable decentralized networks of computers to collaborate in maintaining a shared history of immutable transactions, known as the blockchain. “Are we going to do it with all guys again?”
If your goal is to get more women investing in products like Bitcoin, making them better educated about cryptocurrencies and their risks is the opposite of what you want to do. This is one of those points that would make itself in a more nuanced piece but which needs to be spelled out explicitly here. Pretty much any responsible expert will tell you that, in terms of investment, Bitcoin et al. are likely to crash in the fairly near future and have very little chance of recovering even a fraction of their current value (which already represents a serious drop from the peak value). To leave this detail out is like telling a story of pioneers heading west and not mentioning that their route goes through the Donner Pass.
The brevity and lack of context also means that some of the most interesting historical and cultural aspects of the story are hit upon but not delved into. For instance, there's the growing sense (almost always indicative of a dangerous bubble) that the risks of not investing exceed, perhaps by a great margin, the risk of investing, an attitude you'll seldom see more plainly stated than this [again emphasis added]:
Though the speakers emphasized, for legal reasons, that they were not offering financial advice, the general consensus on how to participate wasn’t particularly novel: buy a little bitcoin (“as much as you feel comfortable never seeing again,” Alexia Bonatsos, a venture capitalist, advised); start experimenting with different wallets (the ways, or places, to securely store one’s public and private keys, used to send and receive currency); and play the long game. Take advantage of resources, such as Linda Xie’s guides to cryptoassets and Laura Shin’s “Unchained” podcast, and ask knowledgeable friends for access to Listservs and online communities—in short, network and Google. (It doesn’t hurt to have some technical know-how; for security reasons, it’s safer to have a hardware or paper wallet than to use the more user-friendly platforms recommended by Morin and Bonatsos, like Coinbase or Robinhood.) “Think, obviously, about the risk of what you’re putting in,” Simpson said. “But really think about what is the risk of not investing, and not learning and not participating. Because I think, over a period of decades, if you invest the time, invest money, and start really participating, you will do well.”
Another interesting notion is the idea that schemes which promise the opportunity for anyone who puts up a moderate amount of money to get rich quick are somehow democratic and, more to the point, that criticizing these schemes is somehow undemocratic. If you go back and read a detailed account of the original exploits of Charles Ponzi, you'll see this was a major theme even then.
There is something utopian, and appealing, about the potential for cryptocurrency to provide an opportunity for more equitable wealth distribution.
...
Cryptocurrency is the closest thing they have to employee equity, itself a speculative asset; it’s their opportunity to be in the right place at the right time. They’re largely writers and academics, activists and artists, even some tech workers looking for a change.
In addition to the standard bullshit stories people tell themselves about implausible get rich quick schemes, cryptocurrencies bring with them all the hype and magical heuristics we would expect from Silicon Valley.
“It just can’t happen that we have another wave of technical innovation happen, and that all of society is not participating,” he said. “I think that means both men and women; I think that means, you know, people in cities and people in rural areas. This technology is so profound, on so many levels, that it feels really important to educate everyone about it.” By his account, it wasn’t just about the money: the blockchain—that ledger of permanently documented exchanges, which is distributed by participants in a given protocol’s network—has far greater implications. He suggested that other transactions could move to the blockchain, eliminating flurries of paperwork, and intermediaries, as well as increasing the digital security and privacy of all parties; he gave the examples of buying real estate and negotiating venture-capital contracts. (In 2017, women-founded companies accounted for just over four per cent of venture capital deals, and received about two per cent of that year’s venture funding, according to Fortune magazine.) And yet speculation about the possibilities of the blockchain have a tendency to turn cypherpunk. “We all use things like social capital, and love, and empathy,” Dave Morin said. “Most of those ways that we interact have not been turned into money, or haven’t been turned into a currency of any kind. It’s the first time in history that we’re taking all these things that have not been a currency in the past, and turning them into currencies that can be exchanged in various different ways.”
The potential applications of block chains, the viability of Bitcoin is a currency, the wisdom of investing in cryptocurrencies, and the vague but powerful sense that these things are portents of a wondrous New Age all get mixed up in complex and often contradictory ways. For instance, the promise that some new cryptocurrency will shoot up in value greatly undercuts the idea that it would make a good medium of exchange, but you will see these statements made side-by-side all the time.
Finally, the story basically ignores the disturbing potential social costs of the Bitcoin bubble. Of course, any time you promote shady, get rich quick investments, you are likely to drive a significant group of people into financial ruin. In this sense, promoting the bubble is a bit like telling poor people to spend more of their income on lottery tickets. There is, however, one important difference. As a rule, at least some of the money collected for a chance at mega millions goes to things like education and infrastructure. That Bitcoin investment is doing this....
A closed-down coal plant in Australia's Hunter Valley, about a two-hour drive north of Sydney, is reopening in order to provide inexpensive power for Bitcoin miners. A tech company called IOT Group has partnered with the local power company to revive the power plant and set up cryptocurrency mining operations, called a Blockchain Operations Centre, inside it. This would give the group direct access to energy at wholesale prices.
According to The Age, the Hunter Valley coal power plant was closed back in 2014. Hunter Energy plans to restart the generator in early 2019. The company understands the demands of cryptocurrency mining, and hopes to make the power plant even more attractive to tech companies by adding cleaner energy sources, such as solar power or batteries.
Cryptocurrency mining is an incredibly power-intensive process. It involves using energy hungry computers to solve complex problems, generating intense amounts of heat and using quite a bit of electricity. As a result, miners and mining companies have been on the hunt for inexpensive electricity. Operating from within a coal plant meets that requirement for sure.
I realized recently that my networking skills (which weren't that strong to begin with) have atrophied while I've been focusing on writing. As a partial remedy, I thought I'd invite all of our regulars to connect with me on the big business networking site (the one that starts with an L). Just mention you're a reader of the blog.
Thanks,
Mark
p.s. The following has nothing to do with the post. I just thought we needed a picture. (from Galaxy Magazine June 1951)
