I've heard from a friend who used to work in Houston refineries, that it's not unusual for refinery workers (who do carry mobile phones, by the way) to call their brokers immediately when they hear an explosion or see an unexpected plume of smoke from a neighboring refinery.
Also smoking on the job is apparently much more common at refineries than I would have expected.
Edit: this comment reflects the state of the industry 20 years ago, second hand. Take it for what it's worth.
I once read an argument (maybe here on HN, I'm not sure) that if insider trading were legal, Enron wouldn't have taken anybody by surprise. I'm not sure I buy the argument that it should consequently be legal, but it's an interesting idea.
On the other hand, consider how rare the Enron case seems to be. Since it's implosion, how many other companies were based on total fraud? Theranos is all that comes to mind and it wasn't even public.
It seems like the system mostly prevents massive Enron style fraud.
heaps of public companies turn out to be frauds, there is an entire industry out there of people spotting fraudulent stocks and shorting them (see Muddy Waters, Citron research et al)
Sino-Forest has probably been the largest and most well known (as well as Valeant and their fake pharmacies) along with a bunch of Chinese small-medium-cap reverse IPO's like CCME
Total fraud isn't the only sort of corporate fraud though. Legalized insider trading would really open the floodgates for financially incentivized 'whistle-blowing' of any sort of fraud (as well as basically anything else that could conceivably move the market...)
That's an interesting idea! Finding the 'victim' in insider trading is complicated as well because it is typically someone who would have made a trade but didn't, so how do you know who they are?
There was a Planet Money podcast about this [0] that if I remember correctly said it wasn't even explicitly illegal in the US until relatively recently. It was considered theft to take insider knowledge, but not illegal to trade on it.
My friend's explanation was that if the really flammable stuff gets out, it's already going to be on fire anyway. He described a technician tossing his cigar butt in a puddle of diesel fuel to make the point. The fuel didn't catch fire.
Diesel isn't particularly volatile, it's difficult to ignite even with an open flame. Nothing like gasoline/petrol (which also would probably pass his cigar butt test, but vapors from which could certainly ignite from the match he used to light his cigar.)
Not the OP. Two thoughts: Working in an environment that could explode at any minute requires some relief to combat stress. Kind of like a war zone. OR Smoking doesn't make sense, so of course people who choose to smoke would do so where it doesn't make sense.
So in your world, if you’re willing to be X amount self-destructive or accept X risk, you’re also willing to accept X + Y risk or self destruction regardless of how large Y is. Got it.
Which also doesn't hold. I'm willing to invest my money, but I'm not sticking it into penny stocks. Your reasoning ignores any sort of reasonable boundary people have wrt risk.
Which shouldn't have quote because it wasn't what I said, and isn't even a fair reading of what I said.
Correlated implies a statistical relationship, ruling out phrases like "will take any risk regardless" and requiring phrases more like "may tend to" and such.
And then I even said that I "expect" the correlation, so it's probably reasonable to read the usage as more conversational than scientific.
It's pretty typical addict behavior. Next time you go to a gas station, look at the ground. Unless they sweep regularly, the ground next to the pumps will have a great deal of cigarette butts on it.
I'm not surprised in the least. The industry is full of stories that started out as rumors but have ended up proven as true about the lengths they will go for more or better information.
I've heard an apocryphal story that I'm still waiting for truth on. It was of a soybean prop trading firm quant who reverse engineered a cross-town rival's trading strategy, allowing them to front run their strategy. The way that they did it was taking a number from the firm's name, and using it as a random number seed, and then using that seed to find the randomized order sizes would result from their iceberg order algorithm. They would use those sizes to match to l2 passive limit order flow to try to identify activity coming from the firm. Once they discovered the direction the firm was going, they'd place a huge market order and suck up all of the liquidity that the competitor was trying to acquire passively, effectively moving the market before their competitor could fill their position.
Stuff like this still blows me away, but it's just crazy enough to be true.
I heard that story from a quant friend. However, the way I heard the story, the targeted firm found out about their rival's attempt to game their strategy, and faked a bunch of activity. The rival's algorithm acted on the activity and placed a large market order...in the wrong direction.
From a place of total ignorance, why would telegraphing a particular false strategy to your competitor in order to get them to respond in a way that is profitable to you be illegal?
Because you’re not just telegraphing it to your competitor, you’re moving the entire market and affecting the bid/ask and price transparency for all its participants. That affects liquidity, derivatives, a lot of downstream negatives.
That said, there are narrow instances where you can place less-than-real orders in a market [0]... it just has to be a market that is nearly and completely illiquid.
Sorry, should have been more specific. As I understand it, no trades were made--they just spoofed whatever activity their competitor was looking at, so it wasn't illegal (but then again, I was hearing the story second-hand).
I'm lost as to what is meant here by "number". Are you saying the target firm was randomizing its iceberg order's component order sizes using the same seed? If yes, how could the target firm's usage of that seed be known by the "attacker" here?
The target firm had a number in their name, kinda like a16z or Office365. The attacking quant knew the target was using iceberg orders which split up and randomize parts of large orders. Attacking quant took a guess at an RNG seed by guessing they used the number in the target firm's name, and then somehow fit the resulting random number stream to order chunk sizes, and was able to confirm they used that number as a seed. From then on, using that information, they would identify (I'm assuming probabilitistically) when the target firm was executing a large iceberg order, and then front run the remainder of their order.
It sounds ridiculous on its face, but the industry is known for going to extremes, and that sort of problem solving isn't unheard of in other domains (like cryptography for example). I'm way out of my depth with understanding it, and it could be complete bullshit, but it isn't outside the realm of what I've seen proven in the past, which is why I'm hoping someone someday will confirm it.
It would be amateurish if it were cryptography for sure. But I would never expect that sort of attack vector as a quant. Most quants are worried about alpha loss, not having your order flow fingerprinted.
don't think this is fraud. Oil refineries aren't lying about their oil production to the general public. It's the hedge funds' fault for relying on their location data.
I'm not sure about American law, but most of our laws have a term mentioning "... in Good Faith."
These two words are pretty spectacularly effective in disarming you from all sorts of trickery.
It is actually illegal to beat an opponent's bot in HFT. It is regulated. One is not supposed to release a strategy on the market, which goal is "to beat an opponent's bot". There've been precedents of that (i.e. bot of a pension fund figured out and beaten by algorithmic trading company bot) and the resulting trades were reversed, fines imposed by regulators.
Source. 4 years of R&D at HFT company that traded around $10^10 during that time :)
This won't meet the elements of fraud in the US because while you could argue that the performative activity was a deliberate representation (in contrast to a false claim where I tell you directly about a great deal on a bridge that doesn't really exist), you'll have trouble establishing the detrimental reliance aspect. If you see misleading activity and then jump to an incorrect conclusion instead of verifying its truth or falsehood then that's on you.
Good point, but I'm not sure if fake market behavior (along with the direct exhortations to participation) directly maps onto fake real world behavior with no solicitation.
As a (somewhat ridiculous) counter example, suppose I know you, a trader, love tap dancing and I, also a trader, happen to be a skilled tap dancer. On Monday I show up outside your office and start tapping up a storm. You are unable to resist the sound of your favorite thing and rush out of your office to enjoy my performance. I return at the same hour Tuesday, Wednesday, and Thursday, and by then you are out on the sidewalk expecting me when I show up. Friday you leave your office as usual but you wait for me to show in vain, for I am in my office and taking advantage of your absence to make a big trade against you.
Have I defrauded you? You believed that I would show up to entertain you instead of competing against you in the market, and frankly I trained you to think so only to enrich myself at your expense.
I don’t know why people think it’s ok to intend to commit fraud, intend to swindle someone, do it. Then have a long song and dance about how the fraud committed wasn’t actually fraud
Let’s make it simple you intend to commit murder, you buy the tools, draw up a plan, then get arrested. Guess what you are up for with zero interaction with the victim
> Researching how to dispose of a body is only research. Intent is the key word here...
Intent is a tricky thing to prove, it requires knowing the state of someone's mind. Until we have mind-reading devices, there is no way to know someone's state of mind with complete certainty, so we necessarily have to rely on circumstantial evidence, including what someone said, wrote, and yes, researched in the past.
Researching a grotesque topic is by no means proof that you committed a grotesque act, but it's definitely a nugget of evidence.
Rhetorically, What is the intent here ? The intent was explained as to distract to extract an advantage. The story was told with us knowing both sides view. I have no idea who you think you are fooling.
But these guys wouldnt be broadcasting this to everyone - just the people who were idiotic enough to look at the spoofed data. They are conducting acts (albeit crazy ones) in a private setting. Technically, thats really what they are doing, and it would be hard to prove otherwise.
They are being spied upon by having their location data sold be carriers. They did not sell their location themselves. For all intents and purposes, they are conducting acts in a private setting that is being violated upon.
I think realistically, the refinery wouldn't have a reason to release the information that they saw a red flag. The stock wouldn't be affected, there'd just be an internal scramble/panic attack.
The refinery wouldn't necessarily notice it's been the stage of a fake disaster. The scheme uses a channel (cell phone location feeds) the refinery is unlikely to watch.
Are they placing detection equipment at/near the target locations to monitor for cell phone signatures in the vicinity, or are they gathering from mobile apps who resell location data to aggregation firms?
You can buy location data from companies that buy it from carriers, some weather companies, and some advertising companies. If you're an oil trader, you almost certainly have the resources required to accept all these feeds and clean + cross ref them for a couple hundred locations.
Yea that wasn't clear. The article made it sound like they're aggregating data from various sources that collect location data (probably from those app metric providers like Tune and others). So if someone ran an open source phone with F-Droid or just installed zero apps with metrics (and Google isn't one of their sources), they wouldn't ever show up in this data set.
I'm guessing location data comes from the Apple/Google location APIs that can map peoples aprox locations based on their proximity to Wi-Fi points (I think they register Wi-Fi points and then the first GPS signal the phone gets after; like when someone walks outside, and use a set of those to get approximate triangulation for indoor Wi-Fi APs).
I've envisioned an application like the following for construction sites and industrial installations:
There would be a "supervisor" app with a schematic 3D model of the site, with dots or tags indicating the location of smartphones of workers, who have installed a satellite app.
There would be messaging between the satellite apps and the supervisor app, and a worker with a satellite app could stream video or post pictures to their tag, which could then be accessed from the supervisor app.
There could also be a VR/AR version of this. No video stream would take over the POV of the VR/AR users. Rather, the streams would be available as pop-up "screens" from the location tags on the schematic model.
For bonus points, interface this with CAD/CAM, so that not only the schematic model can be imported, but the design details can be accessed and drilled-down, and relayed to the workers on-site.
The purpose would be to provide answers and clarifications to workers very quickly, as well as providing contextual information localized in the 3D space of the design to supervisors, architects, and engineers.
You're loosely describing BIM (Building information modeling)[1], but you definitely bring up some novel points that aren't covered by existing BIM softwares, to my knowledge.
Yeah, this is pretty much what a plethora of punchlisting platforms do currently, but in the form of annotated photos captured on an iPad and automatically tagged to the site plan rather than realtime video feeds (which is both more practical and preferable from a liability standpoint, as the GC and architect both have incentive to maintain a paper trail of who approved X change, etc.) The AR angle certainly sounds cool but even aside from the HW not being there yet, I'm not sure how it would add much value.
photos captured on an iPad and automatically tagged to the site plan rather than realtime video feeds (which is both more practical and preferable from a liability standpoint, as the GC and architect both have incentive to maintain a paper trail of who approved X change, etc.)
There's no reason why the realtime video feed and the voice communication between the architect and GC couldn't be tagged and timestamped as well.
The AR angle certainly sounds cool but even aside from the HW not being there yet, I'm not sure how it would add much value.
For plans/models which have some 3D complexity, being able to crane your neck and move your POV around might be valuable for understanding exactly what's going on. This might be more applicable to building the 1st iteration of something like a large jetliner than to a building, though I've heard of situations in the building of stadiums where this would've been applicable.
The point is really to save the project time by letting the workers ask clarifying questions quickly, while enabling engineers and architects to understand the situation quickly, all while communicating in realtime. The turnaround for answering such questions probably couldn't get faster than that.
I'm confused, throughout the article it mentions contractors being in oil facilities. But it also says "Orbital receives location data without any personal identification information and aggregates it."
So how do they know these people are contractors? Or are they just measuring the raw number of people in the oil facilities?
Seemingly a contractor would be a person who is not typically at the oil field but stays for an entire work shift. Other people would be on tours or probably execs.
They are buying this data off ad providers and getting an ad id and geo coords, this is what leads me to believe they are just going off volume of users in the location
If you have lots of location data say in Houston, and then suddenly in a refinery that's not in their normal day-to-day location data you could assume it's a contractor being called into investigate an on-site problem. or if you detect increased activity from devices that don't have a history on-site. Not 100%, but then they're not going for 100% either.
The discoveries through big-data analysis and visualizations can be endlessly counter-intuitive, mindblowingly unpredictable but also dynamic and powerful, even visceral when you have massively voluminous data that can be fed into systems like Tableau where you can actively tweak data queries, aggregations, transformations and visualization types.
The result can often be something akin to: "wow, who could have ever predicted that this data would extrapolate out to show such a unique visualization or trend"
The article doesn't specify, they could be just looking at patterns of "usual" vs "unusual" ad ids showing up.
However, re-identification of "anonymous" ad ids is extremely common, and they could easily be using a service that gives them, for instance, a home address and email account associated with the phone via re-identification methods.
Contractors might have specialized skills related to refineries, so you could look for phones that spend small-to-medium amounts of time at refineries belonging to different companies. You'd probably catch some inspectors and the food truck guy and whatnot, but for the most part it should only be contractors moving around between companies like that.
You don't really need to know they are contractors, they are looking for a jump in the number of people onsite because access is controlled. A major turnaround or outage will bring in hundreds of extra contractors for days at a time.
They could get gps traces for each unique phone within the confines of the plant but labelled by random uuid. Then you can tell if a new one shows up, but with only locations inside the refinery you can't really say more.
Or the weather app is selling their location data to one or a hundred data brokers every 5 minutes. Mobile monetization via third party SDKs is a big thing. Likely 100k different SDKs to choose from (a number that came from a data aggregator).
Add in mobile operators, etc, and recombine the anonymous ids with the help of LiveRamp et al and it’s just a simple data problem.
Satellite companies also track shale drilling rigs. There are a few 'stages' in the lifecycle of drilling rig, with a focus on "DUP" rigs -- drilled but uncompleted -- which can be brought online on short notice(). The various stages of development require crews of different experience or expertise. I was told that some/a company is able figure out the phone numbers of crews accustomed to (I think) bringing DUP wells fully online. Then they are able to monitor as they move around the permian or bakken and assess how many wells are brought online. I'm not sure that they know phone number (333)2920-2920 is Frank Jones the Expert, but they do know that the arrival of (333)2920-2920 and a cluster of other numbers is associated with a DUP well turning on production shortly after.
()I'm not very familiar with the shale fracking process, so don't jump on me for technical inaccuracy.
Also there are many tongue-in-cheek or naive comments about refineries feigning problems, or knowledge of refinery issues being licenses to print money in futures. It's not even close.
Many of the important refineries in the US are monitored 24/7 by third party companies (eg. Genscape) who stare at refineries with thermal cameras. These immediately notice potential problems, as flaring occurs, or some unit dramatically changes temp (gets hotter or cooler), updating subscribers in moments. There are similar services that scrape all the public-warning systems that refineries must use to inform residents when they have emissions or problems.
Finally, while no doubt there are times when refinery upsets (particularly for major gasoline-producing refineries) occur there is an immediate, clear, exploitable move in futures markets [in fact recently], there are also many times when that knowledge would have been swamped by some other force. *I recall reading a few months ago that someone analysed a literal insider traders' trade history and realized they weren't able to exploit to nearly the level you'd expect. It just turns out that markets are absurdly complicated.
And the next step in this race will be oil refineries buying a bunch of put options, then intentionally getting a bunch of strangers with smartphones into their refineries to trick hedge funds that there is something wrong, then watching those put options skyrocket. Would that be insider trading?
A clever person could defeat this entire system in a day.
If you look at the graph, currently we're talking about a fluctuation of about 7 devices.
All you'd have to do is set up a handful of devices (or emulate them and spoof the GPS).. download some sketchy apps, or look for apps that have ties to this company, to bump the location information up. Then turn the devices on or off, so that there's never any change.
If I was the oil company I'd probably set up a shell company to buy the data feed as well so I could get it perfect.
If you're the oil company you probably don't care one bit if a trader is making money on the price of oil, because it isn't coming out of the oil companies profits. It's coming from other traders.
Almost single major oil company & refinery has a team of traders that directly participate (or use the investment banks / futures brokers) in the crude oil and refined product futures / options markets. This is both for hedging and speculation. If a refinery knows when it's going down for (un)planned, it lets it's traders know and they do trade on that info. In commodities markets, this is legal and does not constitute insider trading.
They still don't want their competitors to know how well they're doing or their suppliers, who might increase their prices if they know the company is doing well.
I wonder if there's companies who sell ELD info similar to other examples listed in here about carriers and apps selling user info. Having ELD info from truckers around refineries and drill sites would give you a pretty good insight as to what's going on.
I think many people will agree with me that this kind of data collection is creepy and unfortunate, but if it's the world we live in then we should try to use it for good. I'm struggling to think of a way that a non or for-profit could use aggregate data to predict events in a way that would generally be considered a net positive to society. Does anything come to mind?
I don't think either of your proposed options are legal/possible. afaik this geo data isn't sold to a single party, and is collected through so many various channels purchasing them all would be quite expensive.
As far as making your own cell tower, I would love to know if this is even something a non telco company can do. I doubt creating any blocks for specific content would be possible or legal as well. How would you block https traffic, and how would you determine who is an employee and who is not?
This tracking is a result of allowing visitors to bring personal devices on prem. As long as they are personal devices administered by their owner I don't know what could be done to mitigate this.
The problem still exists if all phones are put into a faraday cage when they arrive at the site. Instead of tracking users at the plant they just track users who 'go offline' at the drop off site.
Getting employees to stop bringing personal devices is the only way to actually quell this fear. Even the US military has an issue with Strava run logs leaking data about the movement patterns in their bases.
This is done to simulate popular launches (eg. hire people to wait in line when your new store or restaurant opens, to give the impression that the new place is hot) or for faking headcount metrics for investors (eg. hire people for a day when the investors come, so they can gawk at the open office full of people).
It's ethically dubious, but if it's done for a day or two, it's not that expensive compared to actually hiring that many people on a full time basis.
They are traders not oil refinery managers. If a trader has the knowledge that a refinery is having a problem minutes or even seconds before other traders they can use that to their advantage and maybe make profit in the market.
There are different zones. In Zone 0 or Zone 1 area you can't have any of that. Also there are special phones that can be used in those zones. Those things are basically bricks.
Probably because no one has ever actually documented sparks caused by mobile phone RF emissions, at least any in current use. That was essentially a bunch of pseudoscientific FUD.
Power switches with mechanical contacts on older phones might be a different story, but again, there's no smoking gun (or flaming refinery) that points to this having ever been an issue.
My thought was that it was always older style cellphones with detachable batteries, and careless user dropping phone, detaching battery, and -that- caused a spark, not RF emissions.
I remember a mythbusters episode where they filled an outbuilding with dozens of different cellphones and lots and lots of open cannisters of gasoline and other petrochemicals, then called them all repeatedly without incident.
I'm sure there are parts of a refinery where even static electricity could be enough to set something off but I'd assume they were not areas that people typically worked.
Pretty sure the field teams aren't allowed them. Maybe in the foreman's trailer? Otherwise, as far as I understand, they can use them back in the cafeterias and locker rooms.
Aside: Mobile phone spark risk at petrol stations is a myth: http://i.stuff.co.nz/technology/digital-living/3909468/Petro... - Dr Kruszelnicki said the forecourt warnings were more about service station owners "covering their arses from a legal point of view".
Of course, a refinery needs to be more careful of batteries.
Also smoking on the job is apparently much more common at refineries than I would have expected.
Edit: this comment reflects the state of the industry 20 years ago, second hand. Take it for what it's worth.