I actually used to write book arbitrage software, and this person has managed to completely misunderstand it.
I'm not arguing here for its societal value, just that dan appears to have misunderstood some fundamental bits about how it works.
> Robot 1 offers Book A for $10. It probably doesn't have it in stock, but it knows where it can buy it for, say, $9, drop-ship it to the customer without ever seeing the acutal product, and make a buck on the arbitrage. Robot 2 notices this and lists the same book for $11, hoping to do the same thing. Robot 1 notices Robot 2's listing, doesn't know or care that Robot 2 is another arbitrage trader, and reprices its listing for the book to $12. Robot 2 notices this and goes to $13, and away we go.
This makes no sense; the fundamental rule of amazon bots is that consumers buying used books almost always buy the lowest-priced used book available. So the bot war would send the price down, not up.
Why in the world would bot 2 offer the book for $11? It would know that nobody would buy it when there was one already listed for $10.
The silly-high-priced books that I saw when I was working for the book company were rare long-tail books. One of the discoveries that many people have made is that for rare books, whether or not you understand what the heck it is, some % of them will get bought even if they're listed for silly high prices. Enough people will pay thousands of dollars (not an exaggeration) for a used copy of a book that they can't find elsewhere that stocking the long tail of books can in fact be very valuable.
That means that, when you find lots of rare books, your incentive is to throw out some crazy-high price for each, because some % of them will, in fact, get paid, making the whole enterprise worthwhile.
edit: Arbitrage on the same market Does Not Work, in my experience, and to just claim that it does with no evidence doesn't make it any more likely to do so.
>Why in the world would bot 2 offer the book for $11? It would know that nobody would buy it when there was one already listed for $10.
The author covers that point with this quote:
>In this case, buyers could easily find the "original" products for themselves, and pay less... but the search results are cluttered with millions-sold arbitrage super-stores
Our Oriental Heritage, by Will Durant currently has about 56 used copies for sale on Amazon. It's not especially rare; I bought it to fill out a half-set I bought from an antique store for $20. But the two highest prices are $225.00 and $588.99 when there are seven listings under $3.00.
That's just one example that I noticed the other day. Presumably, the cheaper copies are either poor condition or from high-volume stores. The middle prices seem to be in better condition, from lower-volume sellers. The higher prices are from sellers with many different levels of volume. Presumably, they're hoping for the inattentive-buyer market.
My guess would be that they listed the book a long time ago and lost track of it or that their pricing bot has a bug. Or they're humans (one dedicated person can do a surprising amount of book sales, we discovered).
Our experience was that there is no inattentive-buyer market. Those high-priced books will not sell.
edit: I'd add that we didn't like to stock books with low prices like that. There's no margin in the <$5+s&h books you see on Amazon. Instead, we had a big pulper and we'd pulp our 10m copies of "Patriot Games" and sell them to recyclers. My understanding is that my former employers have moved to start making recycled paper themselves and selling that these days.
You wrote the bots, and I wrote the code that (tries) to keep them from messing with Amazon (as an intern).
People on Amazon don't just buy the lowest priced item: They buy from whoever won the "buy box": The big "add to cart"/"buy now 1 clock" buttons. You seem to be talking about selling used books, where the market is a little different than "new" items.
Price is obviously the biggest determination of who wins, but you can be disqualified because you don't have a high enough ranking, and other factors related to customer satisfaction.
So an established merchant may win the buy box even if they're 50 cents higher. And that's exactly where this scheme comes into play, as soon as you've got two of these bots playing off each other and no other sellers qualified to win the buy box.
We were also obsessed with our customer rating, that just didn't happen to be my job. It was tough to get happy customers to come back and give us 5-star ratings; people with bad experiences would come by and rate but those with perfect experiences never would. Or people who got a book in perfect condition, on time, would rate us 3-stars. What the heck did they expect?
Obviously, shipping on-time and good product is crucial too. Again, mostly just not my job.
So we were definitely working on the customer satisfaction angle as well, there just wasn't much that I could do about it, so I talk less about it.
that doesn't work on amazon, any book with an ISBN will get condensed into one listing. You can get a little tricky with it, but not enough to allow intra-market arbitrage.
I understand that he's claiming it exists, and it might with other products, but my experience was that it does not with books.
Buyers are willing to pay a very small premium for feedback score. Furthermore, one of the bots will know that it has a lower feedback score, immediately killing any hypothetical positive feedback loop.
> This makes no sense; the fundamental rule of amazon bots is that consumers buying used books almost always buy the lowest-priced used book available. So the bot war would send the price down, not up
Consumers often take into account seller reputation. If a book is $20 at a seller with stellar feedback from a large number of customers, and $18 at a seller with very little feedback, I'll probably buy from the $20 seller.
What happened in the case with the two bots on Amazon was that one bot was pricing for a seller with a stellar reputation based on a lot of feedback. It was finding the low price from sellers whose reputations were not as good, and setting the price enough above that to make a profit reselling from the low seller if someone ordered.
The low seller's bot was using a strategy of undercutting the lowest other seller.
The amount the first bot was adding was more than the amount the second bot was undercutting by, so the net result was the price zoomed up.
> That means that, when you find lots of rare books, your incentive is to throw out some crazy-high price for each, because some % of them will, in fact, get paid, making the whole enterprise worthwhile.
For the rare books, I expect the price would be limited both by value and by how much time it would take to find another copy somewhere (say, a university library two towns away) and how much the buyer's time is worth.
Domain squatting I thought tends to be more about trademarks and trying to confuse people.
This has been studied and documented in detail, about a year ago - five minutes on google will find it. Basically, someone tracked the two prices for a single title rising in lock step - just the thing you say makes no sense.
My own book, which is available new from Amazon for around $12, has listed for several hundred thousand dollars from a reseller.
Just because something happens sometimes doesn't mean it's par for the course. That guy selling your book for $100k? He's probably not selling your book at all. It's hard to believe nobody would have worked out that this is not the most efficient scheme.
> Why in the world would bot 2 offer the book for $11?
Dan's explanation:
"The price-setting robot sets its price far enough above that of the other vendor that if someone buys, the robot-owner can buy the book from the other vendor, send it to the customer, and still make money."
To make money - out of nothing except a listing - only requires a buyer, either misguided, malicious or unfortunate. This might even work in large numbers by pure chance. Or it might not, but someone is trying.
> consumers buying used books almost always buy the lowest-priced
There it is, a bet on the complement of "almost always", at an insignificant cost.
> The price-setting robot sets its price far enough above that of the other vendor that if someone buys, the robot-owner can buy the book from the other vendor, send it to the customer, and still make money.
Yes, this is what book arbitrage is, I hope I'm aware of the concept! It's what I did for a few years.
> To make money - out of nothing except a listing - only requires a buyer, either misguided, malicious or unfortunate.
Or none of the above! If they buy it on Amazon for $10, but it was only $5 on AbeBooks, and they never would have found it there, they're a perfectly happy customer.
> There it is, a bet on the complement of "almost always", at an insignificant cost.
I'm suggesting that, as a person with several years professional experience doing so, nobody makes money on Amazon by being anything except the lowest-priced used book available.
I'm further suggesting that the bot wars that happen (often!) are negative feedback loops, causing prices to drop, not the other way around, and he has provided neither examples nor a reasonable explanation of why bots would fight to be lower on the list.
Being first on the list of used books is everything when selling on Amazon, and if there are bots going the opposite way, they won't last long. Low cost ≠ free.
I did something similar in the early 2000s. I sold public-domain books (as in created by government agencies) as pdf on cd on eBay.
The cost varied between $1-$10 with $5 shipping, and would cost me $1.20 out the door. I had everything automated with a robotic disc publisher/printer and stamp printer that I would work maybe one hour a week and bring in a couple hundred dollars each week.
It was definitely a great source of side income. The main market was people who didn't search google, since they could take the title of the auction, put it in google, and usually find the PDF right there.
There's still people doing the same out there but now the market's flooded and there's much less profit margin available.
Depending on the size of the dataset, you may have been providing a valuable service to those on dialup (however unwittingly). I remember how even a download of 50-100MB would take dedication to keep everyone off the phone for hours and pray that there were no interruptions upstream. Paying $10-15 for a guaranteed CD of what I wanted would have been well within reason.
Money laundering seems like an interesting explanation, but there are far better and less risky ways to launder money. Cash businesses, or even bitcoin are superior to "selling" an app that could be taken down at any time, that requires credit cards on file to purchase, and whose transactions Apple or a credit card company could freeze, reverse, or cancel at will.
It would be fairly trivial for Apple, in cooperation with a law enforcement agency, to source the transactions, map the patterns, and triangulate to likely points of origin. And it would probably arouse a lot of suspicion if an app of such obviously bullshit quality suddenly started raking in millions of dollars in sales -- thereby making sales of the app inefficient and risky as a laundry-integration method.
I suppose the whole thing relies on iTunes gift cards and credits, rather than credit cards, but why overcomplicate things beyond there? Buying and reselling gift cards at a moderate discount would launder your money just fine, while limiting your exposure to third-party risk.
Apropos of nothing, the dissection of the high-priced crapware reminded me of "I Am Rich," the infamous and short-lived iOS app from 2008. It cost $999.99, and all it did was display a glowing red gem and a tagline on your screen. Evidently, at least eight to ten people bought a copy before Apple swung the ban hammer.
Apropos of nothing, the dissection of the high-priced crapware reminded me of "I Am Rich," the infamous and short-lived iOS app from 2008. It cost $999.99, and all it did was display a glowing red gem and a tagline on your screen. Evidently, at least eight to ten people bought a copy before Apple swung the ban hammer.
I never understood why Apple banned that application. There was no fraud involved. The application did exactly what the description said it does (show a glowing red gem for 999.99) and it didn't violate any of the (quite arbitrary at that time) Apple app store policies, or requirements.
If you want to put your douchiness on display by paying so much cash for no other reason than to yell out to the world that you can afford to throw 1000$ out of the window it shouldn't be up to Apple to put a stop to it.
I assume it was a matter of fear about bad publicity on Apple's side. Nevertheless I thought pulling this app was one of their more innane decisions.
The official explanation is that a few people complained to Apple that they'd been "tricked" by the app. (Because somehow, paying $9.99 for a glowing gem is a more rational purchase.)
There was nothing technically deceptive about the app or its marketing. It was banned largely out of what I presume was a "broken window theory" approach to store curation: let stupid apps live, and more stupid apps will show up. I am not sure if I agree with the call that was made, but I understand the rationale behind it.
Sure, but hindsight is 20/20. Apple was making this call back in 2008, when the app store was still really new, and a lot of developers were waiting in the wings to see if it was going to be worth building for.
Navigating the quality/quantity straits was by no means an easy task back then.
The tabloid press (in the UK at least) used the story as evidence of how the App Store was pointless, overpriced and going nowhere. Given that it was early days for the iPhone, I don't think it was the kind of PR Apple wanted, so canned it.
AFAIK, they had to refund anyone who had complained, but they probably kept the money from anyone who didn't. (We're talking about maybe 10 transactions, of which 2 or 3 complained.)
"One explanation for this could be money laundering."
Reminds me of something a few months or so back (in 4chan/g/) when Samsung had some sort of a campaign that gave it's users free android market credits. I don't remember how you had to authenticate yourself, but it was easy to fabricate and people did it multiple times over. As a result, people made really bare android apps to buy from themselves to turn that credit into real money.
You authenticated by entering the serial number of the Galaxy Note 2. Many people obtained them en masse from youtube users unboxing the phone who didn't bother covering it. Others, I think, just brute forced the serials to enter. Not sure if that worked.
I;m interested if google caught them or not. Surely they'd notice a spike in expensive app purchases using gift cards (if they measure things like that).
I would assume it's purchased from stolen Apple IDs and/or credit cards. That seems a whole lot more straightforward and likely than someone handing out dirty money to a bunch of people with legit Apple accounts.
this sounds like a plausible approach, although I wonder for how long such a run would work given that very soon, the seller account would be shutdown if 99% of purchases were made by credit cards reported stolen or accruing reverse charges quickly. I could see that work briefly, but the seller account would long be gone now, after a single run even, I'd assume.
unless, this is just nothing large at play, just a random dude stealing cards here and there, less than 100k a year type of criminal activity.
"One explanation for this could be money laundering"
Selling some useless crap for stupid prices is one of the easiest laundering techniques beloved of many people trying to account for a sudden influx of wealth. Oh ? The suitcase of cash, well it's like this, I sold some old jeans, my wrecked chevvy and a busted tennis raquet to some nameless guy down the [pub|bowling alley|policemans annual ball] and he gave me like $250,000 in this suitcase for them.
I'm sure some of these apps are people having a laugh and then getting an even bigger laugh when someone actually pays the stupid price tag they put on. However, at least some of this has a distinct whiff of laundry detergent and it's always useful to have some schmucks innocently purchasing thrown in for good measure to delay investigators as they follow dead ends to people with more money than sense.
If you're doing it properly, then yeah having a paper trail to justify it helps. Currently some of the drug dealers in the UK are using fixed odds betting machines in betting shops, they put money in, make safeish bets, lose some money, maybe win some then check out however much they've got and move onto the next one. Keeps a trail as you're on CCTV doing something entirely legitimate.
"Selling some useless crap for stupid prices is one of the easiest laundering techniques"
If the goal is to do something completely and utterly ineffective, sure. If you were under suspicion and justified an influx of cash through specious claims, the money would be confiscated as proceeds of crime and you'd be on the hook to prove otherwise. People can argue whether that's fair or not, but the whole reason real money laundering exists is to integrate shady money into legitimate trades and enterprise, and this is a very complex, very involved endeavor: You can't simply manufacture a ridiculous explanation and call it a day, or you might as well simply hand it over when they come calling and put your hands out.
Even though most proceeds of crime legislation uses balance of probablity as opposed to beyond reasonable doubt, the practice is generally for you to be able to show some explanation where it come from which is non-criminal in order to keep the cash. Immoral (such as selling crap to people paying stupid prices for something which is worthless) is not enough to have the cash taken off you. There are lots of stories of people being duped but no actual law being broken; you pay me $50,000 for my burned out chevvy, that's your problem. I have a way to explain $50,000 out of thin air. Any junk yard will sell me a wrecked car for $100 and it's just a profitable business transaction for me to sell it on. Plus, you don't care because I gave you $55,000 15 minutes before the "transaction" so you made a profit when you handed me back $50,000!
In practice to get to a confiscation hearing, there is an investigation which means following the transaction trail and the more complex and convoluted that is made with dead ends and people who disappear (but leave a small trail of evidence they were around at some point to prove they are real), the more time and money authorities expend and the longer it goes on with me hanging onto the money (even if you seized some of it already). The fact I'm selling crap and the transactions make it look like I'm mining gold in a moron mine is enough for a good lawyer to talk me out a seizure or have some bargain made with the prosecutor (which I can always delay paying anyway if I'm in Scotland it seems : http://www.dailyrecord.co.uk/news/crime/revealed-75m-dirty-m.... I'm sure other places have similar stories.)
How legislation is supposed to work (as portrayed in TV/Film Justice) and how it actually works in real life is very different. If it worked TV Style, every 419 scamming prince would be in cuffs when he went to the Western Union office to collect his processing fees. Instead they drive around Lagos in BMWs laughing at all the mugus they scammed with lame-ass stories.
Robot book publishing is really interesting. I think it peaked sometime around 2010-2011 when a lot of "robot books" were available on Amazon. Read more:
Wikipedia had to go through a push around that time to excise any references to such books from Wikipedia itself, because some had crept in, mostly by unsuspecting users who believed they had found an actual third-party book on an obscure subject, and didn't notice it was suspiciously similar to the Wikipedia article. It defeats the purpose of citing sources, of course, if Wikipedia "bootstraps" its own sources by citing print-outs of previous versions of Wikipedia articles as the source.
This does leave a more subtle problem than literal printouts of Wikipedia articles, which is harder to solve. Sometimes journalists and even scholars consult a Wikipedia article for background, and then just repeat something similar uncritically, without independently verifying whether the Wikipedia article is correct. If the Wikipedia article then at some point ends up citing that writing as a source, a hidden reference-loop is created, with Wikipedia effectively "laundering" a self-reference (unintentionally) through a third party repeating something they read on Wikipedia. (Not a phenomenon unique to Wikipedia admittedly.)
When I was building the search for a big ecommerce books site (approximately 20 million products), it was a nightmare trying to filter the print-on-demand products. At the time they were creating new "publishers" on an almost weekly-basis, we were constantly having to tweak our blacklists. The big problem was caused by the titles of all these books being stuffed full of keywords, so for almost any conceivable search they would show up quite prominently.
There were some technological solutions we implemented, but it was a major annoyance all the same.
Arbitrage adds market making value. Just like trading desks that make it so I dont have to (physically) find a person who has apple shares in order to buy some. By making a well (or better) run market, prices can converge to their true value. The seller of the shares can potentially sell to more buyers, therefore finding the highest price possible. The buyer can potentially buy from more sellers therefore finding the lowest price possible. Between the two we get the "most agreeable" pricing .
This is why selling CDs of free things on ebay converged down to the price of a CD + Shipping . Similarly its why (at least for now) bitcoins ought to hang out around the cost of electricity + the capital cost of ASICs ... If the price is much lower, dont mine, if the price is much higher its cheaper to buy electricity and ASICs than to buy bitcoins.
I'd hazard a guess that everyone here understands how arbitrage is supposed to work.
But how valuable is market making, really? Is it always a good thing that "well-run" markets tend towards the extremes of having razor-thin producer excess or razor-thin consumer excess? Because that's what seems to happen in big/equilibrium markets. One party has a slightly stronger bargaining position and the process of market-making teases that out and translates it into "one person wins big, one person wins just-barely-more-than-nothing." In the process, some of that excess winds up in the hands of middlemen, so to argue that "it's a feature, not a bug" you have to show not only that the value of driving one party's margins to 0 is positive, but that it exceeds the sum of the value captured by middlemen in the process.
Sometimes we can throw up our hands and say "eh, you win some you lose some," but other times the result is so unspeakably horrific that this attitude goes from "unproductive" to "evil." Example: labor markets. Is it a feature or a bug that unskilled laborers don't manage to capture any of their producer excess, and that the equilibrium wage would lie just north of what what was required to make work preferable to starvation and homelessness (less, actually, because some people will be willing to invest savings/debt/daddy's$$ in acquiring experience = differentiation = freedom from the crushing competition)?
Look, I don't know of any system that I'm convinced is better. But that doesn't make the market perfect, so let's not pretend that its shit doesn't stink. In particular, is it possible to structure a market in some way to split the excess? Are there any big/equilibrium markets people know of that we could look to for inspiration?
As someone who does not have a lot of money, this article's title intrigued me. However, I was disappointed to discover that the article did not actually provide ways to get money for nothing. Overall, I was overall not very satisfied with the experience of reading this article, mostly because it did not provide me with ways of obtaining money for nothing, even though the title suggested it would. I would not recommend the article to anybody who is seeking to find ways to obtain money for nothing, as the article does not contain such information, and it would be a disappointment for them as well.
OT, but related: One way to eliminate the lower price competitors is to "buy" their stuff, but with a card that has no money/credit on it. It'll be 10 days before Amazon allows the item to re-listed.
Had this happen to me three times in a row on the same item. Why do arbitrage when you can simply nuke any lower-priced competition at no cost?
In my case, every other item for sale, regardless of seller, was at least $150 higher. (I eventually sold the item on Craigslist for cash.)
Now look at them yo-yo's that's the way you do it you code the app on the iPhone store.
That ain't workin' that's the way you do it Money for nothin' and chicks for free
The money laundering hypothesis appears a lot, however I'm not sure I buy it -- the complexity and cost of getting many people involved, with many accounts and payment methods (themselves all tracked), all to then pay 30% to Apple on top, seems like a very convoluted and inefficient method. Worse, the goal of money laundering is to do it as subtly and legitimate-looking as possible, and putting garbage apps at ridiculous prices doesn't seem to meet that requirement.
It seems more likely that someone was trying to game the charts: Some categories need surprisingly little revenue to top, and each of these apps started at ridiculous prices, dropping to the normal $0.99 a month later. Some very naive person may have thought "investing" in buying only a few copies of their app (getting back 70% of it, or even more if they bought discounted iTunes cards) would get them enough exposure that it would pay back (whether through future purchases of the app, or as some sort of backstop to demonstrate their excellence as an iOS developer, etc). This is, it's worth noting, exactly how many books get on best seller charts, particularly in the smaller-market, easily manipulated categories. ^1
Just as a point of reference, I implemented anti-money laundering systems for a banking conglomerate, so I have been deeply involved in analyzing money laundering schemes.
I should add that I would love to make money for "nothing". To make the next Flappy Bird. Don't we all? To be very negative about that has the unpleasant stench of jealousy.
Perhaps these apps are created by developers from other countries who are not as familiar with USD and software value and are just trying to publish something on the App Store.
I could see a young college student in Pakistan creating an app that does something totally random and just throwing an equally random price on it just for the experience.
Absolutely possible. The linked article claims, however, that these apps also saw success on the charts. Now that isn't actually linked to any citation of said charts, so whether it's true or not is open to question, but my comment was purely based upon that.
> It seems more likely that someone was trying to game the charts: Some categories need surprisingly little revenue to top, and each of these apps started at ridiculous prices, dropping to the normal $0.99 a month later.
Hang on... so the top paid apps are based on revenue? I always assumed it was based on the number of downloads.
There is a top paid, and a top grossing, overall and by category, on both iTunes and the Play Store. I don't know what the metric for "top paid" is (though comparing downloads/ratings and price/IAP, it, too, seems to be by gross and not counts, or some combination of the two), but top grossing would very much benefit from this scheme.
On the smaller scale (e.g. not billion dollar drug cartels), the primary mechanisms remain cash-based service business fronts and casinos.
The more novel vehicles that took off in the 1990s was through insurance, where reporting was significantly more lax and the payouts were less tied to to the inputs. Enormous new regulations as a reuslt of 9/11 (and anti-terrorism financing) put dramatic new controls, such that large, third-party payments, or irregular situations will always get flagged and submitted to the government(s) for review.
I'm not arguing here for its societal value, just that dan appears to have misunderstood some fundamental bits about how it works.
> Robot 1 offers Book A for $10. It probably doesn't have it in stock, but it knows where it can buy it for, say, $9, drop-ship it to the customer without ever seeing the acutal product, and make a buck on the arbitrage. Robot 2 notices this and lists the same book for $11, hoping to do the same thing. Robot 1 notices Robot 2's listing, doesn't know or care that Robot 2 is another arbitrage trader, and reprices its listing for the book to $12. Robot 2 notices this and goes to $13, and away we go.
This makes no sense; the fundamental rule of amazon bots is that consumers buying used books almost always buy the lowest-priced used book available. So the bot war would send the price down, not up.
Why in the world would bot 2 offer the book for $11? It would know that nobody would buy it when there was one already listed for $10.
The silly-high-priced books that I saw when I was working for the book company were rare long-tail books. One of the discoveries that many people have made is that for rare books, whether or not you understand what the heck it is, some % of them will get bought even if they're listed for silly high prices. Enough people will pay thousands of dollars (not an exaggeration) for a used copy of a book that they can't find elsewhere that stocking the long tail of books can in fact be very valuable.
That means that, when you find lots of rare books, your incentive is to throw out some crazy-high price for each, because some % of them will, in fact, get paid, making the whole enterprise worthwhile.
edit: Arbitrage on the same market Does Not Work, in my experience, and to just claim that it does with no evidence doesn't make it any more likely to do so.