Pumps, spoofs and boiler rooms: October 21, 2018 Snippets
As always, thanks for reading. Want Snippets delivered to your inbox, a whole day earlier? Subscribe here.
This week’s theme: exploring five common kinds of scams we see in hot markets and bubble environments. Plus welcoming Journal to the family.
Over the past few weeks of Snippets, we’ve been looking at bubbles: starting with their underlying psychology, and then last week considering the promoter as a kind of “antihero” for the innovation economy. This week, as promised, we’ve got a really fun topic to cover and I’ve been looking forward to writing it for several weeks now: we’re going to talk about scams.
Why scams? Well, for two reasons. The first reason is that at the peak of any bubble, including those where the object of speculation is some new and poorly understood technology or breakthrough, we get a lot of scams. The excitement is too high, and human nature is too predictable, for us to ever expect otherwise. These scams proliferate as the bubble inflates on the way up, but their real time to shine is that crucial moment of time right as the bubble starts to burst and begins to fall back earthward: when rising paper gains are no longer materializing the way we’re hoping, and we turn to the eagerly awaiting arms of less scrupulous promoters, con men, and scammers. This unethical activity inevitably becomes associated with the new technology in the eyes of the public, particularly in the subsequent period when fortunes are lost and popular opinion turns decisively sour. We see it today with cryptocurrency, but the same association can be found with almost any new technological bubble: whether we’re talking 1999, 1882 or 1720.
The second reason is simpler, and it’s one I think about a lot: if you want to really understand how something works, study the scammers. They’re the true students of how things work in practice, and where practice differs from theory. They have to: in order to pull off something illegal and get away with it, you have to really understand behavioural motives and market mechanics at a quite sophisticated level. So today, we’re going to look at how markets work, particularly during states of irrational exuberance, by looking at five common scams we see frequently in markets: wash trading, pump and dumps, spoofing, boiler room scams, and Ponzi schemes. (And by the way, as you can probably guess, all of these scams are rampantly found in the crypto world.)
Wash trading: We’ll start with something easy: wash trading, defined very simply, is trading with yourself. It’s putting out a buy order for some marketable security and then immediately fulfilling it yourself, or vice versa: whether you’re doing it solo, with a group, or (as it often the case today) with an army of trading bots. Why would anyone want to do this, though? After all, by buying and selling the same security over and over again, all you’re doing is incurring exchange fees: there’s no economic gain to pursue since your left hand and right hand are trading with one another. But consider this: by wash trading between an army of anonymous bots, you can create the impression of trading volume. Furthermore, if your wash trading army is meaningful enough in size, you can actually influence price: by trading with yourself at slightly higher and higher prices, for instance, you might convince the market that a rising price is occurring genuinely, rather than due to your manipulation. But again, why would you want to do that? Well…
Pump and dumps: Remember when we wrote about mimicry two weeks ago? Well, here’s a chance to put the lessons of mimicry into practice. If I’ve successfully begun wash trading on a market and have convinced enough people on a market that a) an increase of trading volume, and b) an increase in price are both genuine market activity, then what do you think the other market participants will do? Well, if they see other people buying, then they’re going to want to buy some too! This is one of the purest expressions of the difference between theory in practice: in theory, when people see prices rising for no discernible reason, you might expect them to sell and make a profit; in reality, they see prices rising for reasons they don’t understand, and they go buy more. “Pump and dump” scams exploit this behavior: I can buy up a chunk of stock for $10 a share, start wash trading it with myself and coax the price up to $10.50, then $11, then $12. Once the crowd joins in the fun, I’ll initially help them out, then pull back a bit: let them push it organically up to $13, 14, 15 as herd behavior and mimicry override any rational sellers cashing out. And then that’s exactly what I do: before the enthusiasm runs out, I’ll cash out my own shares at the inflated price, pocketing whatever profit I’ve made minus my trading costs. Those who bought at the top are left holding the bag: that’s why we call them “bag holders”.
Spoofing: While pump and dump scams are quite brazen and require a lot of muscle to pull off, spoofing is a bit more subtle. Just like pump and dumps, they also work by creating the illusion of either positive or negative market sentiment, and then selling against that fake sentiment. The way they work is by taking advantage of the fact that markets do not work instantaneously: when you submit a buy or sell order, there is a delay before buyers and sellers can successfully get matched, and a trade can successfully get made. If you’re able to act very quickly, you can “spoof” the market by entering large buy or sell orders — large enough to strongly influence market sentiment — and then withdrawing those orders before they ever get actually fulfilled. In the meantime, you’ve traded on the other side of that sentiment: so, for instance, if you spoofed a very large buy order, you take advantage by selling your real stock at a higher price than you otherwise would’ve gotten; similarly, if you’d spoofed a very large sell order, you can take advantage by buying up stock at a cheaper price than you should’ve. Spoofing was very common in the early days of high frequency trading, when physical distance and length of fibre optic cables between you and your exchange became crucially exploited resources — Michael Lewis’s book Flash Boys tells that story brilliantly if you’re interested.
Boiler Room scams: These are my favourite, just out of sheer brazenness. In a classic Boiler Room scam, whenever there’s some bubble going on that has captured the hearts and minds of the naive retail public (think the dot com bubble, when everyone became a day trader and briefly became paper rich), there a lot of genuine retail demand for brokers who can buy and sell stock for customers. Let’s say you want to take advantage of this retail enthusiasm, and you’re pretty sure that we’re near the top of a bubble. What you’ll do is set up a retail shop that accepts buy orders from customers, takes their money, and then never actually buy the stock: you keep the cash, and you let the customer think that the stock is purchased and in their account. Once the stock crashes, which you’re confident will happen eventually, then you buy the stock for the customer, making their account whole; you pocket the difference. The more prices fall, the more money you can skim as profit. But what happens if prices don’t fall, or if enough customers want to cash out when the market is high? You’ll be in huge trouble! Well, here’s where the “boiler room” part comes in: you’ve made sure that your money, your employees and your whole operation is somewhere where the customer can’t find you: if too many customers try to cash out their stock at prices that would ruin you, you simply pack up shop and disappear. The reason why the scam works so well is that the scenario in which you have to run away with the money is the one where your customers are paying the least attention: it’s when prices are rising, and your customers all think they’re geniuses: they’re probably not cashing out in large numbers; in all likelihood, they’re buying more.
Which leads us to our last category for today…
Ponzi schemes: Odds are good that out of all of these, Ponzi schemes are the most culturally familiar to most of you. Ponzi schemes, and their closely related and “distributed” cousin the Pyramid scheme, both employ the same basic tactic: advertise a financial product (say, a share in a mutual fund, or shares in “a new Bitcoin mining initiative”) that pays out some very large amount as a dividend: say ten percent a month. (That’s 120% per year, which is completely unreasonable to expect for anything legal, but ‘ten percent a month’ sounds just possible enough that some percentage of the retail buying public will fall for it — especially when we’re dealing with something new and exciting, like a crypto mining project or whatever.) Initially, people will be skeptical, but they’ll get a lot more excited once they see that the first investors are actually getting paid real dividends, in cash! Wow, it must actually be a legit project! What’s happening, of course, is that these aren’t real dividends: they’re being paid out of the investors’ principal. But word travels fast: new investors join on, providing fresh cash with which the existing investors can keep getting paid their hefty profits, month after month: new investors make the existing ones whole, perpetuating the euphoria for longer and longer. The reason why this scam works is similar to the reason boiler room scams work: the more money that your investors are “making”, whether in paper gains in the boiler room scam or often in real dividends in a successfully-executed Ponzi, the less they’re going to pay attention to how it’s being made. Again, in a rational world, we’d do the opposite: higher returns should garner more skepticism, not less, in an efficient market world. But that’s not what people are like: when we see ourselves getting rich on paper, we don’t think “I might be a sucker”; we think “I’m a genius! Everyone else is a sucker for not getting in on this.” We’re always the hero of our own story, in our heads: successful Ponzi scammers like Bernie Madoff, at the end of the day, are people who understand how to exploit this vulnerability.
Next week, we’ll look at some real-life versions of these scams in the cryptocurrency world, before we dig into one of the largest systematic market manipulations of the industry that’s now beginning to come crashing down: Tether.
Paul Allen passed away earlier this week, and the technology world paused for a moment to mourn, thank and reflect on his 65 years here on earth. Without young Paul Allen’s initiative, there would have been no Microsoft, and without older Paul Allen’s generosity, millions would be worse off.
What I loved about Paul Allen | Bill Gates
Personal reflection on Paul Allen’s passing | Steve Sinofsky
Paul Allen thought like a hacker and never stopped dreaming | Klint Finley, Wired
With philanthropic efforts, Paul Allen leaves behind a legacy | Alex Fang, Forbes
We’re not the only ones to point this out, but no reflection on Paul Allen would be complete without mentioning that he could really play guitar. Even Quincy Jones mentioned as much in that epic interview earlier this year.
Another must-read piece this week is this one in the Wall Street Journal, exposing a key miscalculation of how much attention Facebook users were actually paying to videos.
The piece matters not only because Facebook was got a key metric wrong (and, apparently, failed to disclose or correct these metrics for quite some time), but also because quite a number of media companies used this incorrect data as cause to fire large numbers of their writing staff and replace them with video teams. The “pivot to video”, as it has become derisively known, has not only contributed to the gutting and decimation of the media industry (as real writers and editors get let go in favor of increasingly bite-size and expensive video content), but it was also based partially on false information. One thing this story makes painfully clear: there has never been a greater need for a new business model for journalism that actually works. Because the current status quo is a disaster, and this story helps illustrate why.
Was the media’s big “pivot to video” all based on a lie? | Maya Kosoff, Vanity Fair
Conceptual breakthroughs:
Unbundling and rebundling, forever:
“The physical cloud”: how robots and drones will change retail forever | Christopher Mims, WSJ
Other reading from around the Internet:
Unflattening Hobbes | Venkatesh Rao, Ribbonfarm
Ben Bajarin on Apple’s secret silicon sauce | Apple 3.0 Podcast
Brands & bone saws | Scott Galloway
An oral history of the Stone Pony, the legendary rock club of Asbury Park, NJ | Nick Corasaniti, NYT
“Polio-like” childhood muscle-weakening disease reappears | Karen Weintraub, Scientific American
And finally, in this week’s shocking but also oddly intriguing news:
“Devastating” scandal rocks wine world as 23 master sommeliers stripped of their titles | CBC
In this week’s news and notes from Social Capital, we have a new family member that we’re delighted to introduce: Journal.
Meet Journal | Samiur Rahman, Journal
Have you ever wondered why the built-in Search function on your Mac is so woefully inadequate at finding anything you’re actually looking for? Or why there aren’t any good integrations to tie together your work life across all of your open tabs, from Google Docs to Slack to Evernote to Dropbox and more? Or as a general question: why isn’t there a Google-like search and organize product for all of our personal stuff that we actually care about? Well, one reason is that hyperlinks, which help give the web structure and organization, don’t exist for all our personal files and projects: the link between, say, four related items — an email about a meeting, a calendar invite to that meeting, a Slack thread preparing for that meeting, and the Dropbox-ed slide deck prepared for the meeting — is a conceptual link. They’re all about the same thing, and any user can understand that. But a computer doesn’t understand that. Not without us manually labeling every email, thread, file and folder by topic, date and subtext, which we’re obviously not going to do. At least, not until now — because now there’s Journal.
Introduced by Journal’s founder & CEO Samiur Rahman: “We are building a new kind of Journal. You write notes in it, save interesting links, and drop in important documents and messages for later. When you need something, ask Journal, and it will actually help you find it.” It seems simple enough, until you stop to ponder what it actually takes for software to develop a sophisticated enough conceptual model of your work and life to genuinely understand it the way Google understands the open web. Samiur explains: “Beneath the surface, the Journal platform has two unique attributes that will enable more flexible and personalized product experiences than we’ve seen from existing knowledge management services:
- A state-of-the-art machine learning and natural language processing model that conceptually understands people’s digital information across formats
- An architecture and UX that understands and shows information as distinct types (e.g. files, tasks, articles, messages, products, and more).”
Journal comes in three forms today: a Mac app, a web app, and a Chrome extension. With each of them, you get a really elegantly designed dashboard that puts all your stuff within a keyboard shortcut away, and will quickly become your default action for navigating through your work life. Having extensively tested the Chrome extension myself, I can confirm that it does indeed work as magically as you’d hope. As with many of the best products, ordinary users will never truly appreciate how much sophisticated machine learning effort and design chops went into its product development — it just works, magically, and you stop thinking about it quite quickly. It simply becomes a part of your life.
You can sign up to gain early access to Journal here, and while you’re there you should sign up for Machine Learnings, their superlative machine learning-focused blog and newsletter, as well as Noteworthy, the Journal HQ blog.
Have a great week,
Alex & the team from Social Capital
