What Your Phone is Telling Wall Street (wsj.com) 108
Your phone knows where you shop, where you work and where you sleep. Hedge funds are very interested in such data, so they are buying it. From a report: When Tesla Chief Executive Elon Musk said the car maker would work around the clock to boost production of its Model 3 sedan, the number crunchers at Thasos Group decided to watch. They circled Tesla's 370 acres in Fremont, Calif., on an online map, creating a digital corral to isolate smartphone location signals that emanated from within it. Thasos, which leases databases of trillions of geographic coordinates collected by smartphone apps, set its computers to find the pings created at Tesla's factory, then shared the data with its hedge-fund clients [Editor's note: the link may be paywalled; alternative source], showing the overnight shift swelled 30% from June to October.
Last month, many on Wall Street were surprised when Tesla disclosed a rare quarterly profit, the result of Model 3 production that had nearly doubled in three months. Shares shot up 9.1% the next day. Thasos is at the vanguard of companies trying to help traders get ahead of stock moves like that using so-called alternative data. Such suppliers might examine mine slag heaps from outer space, analyze credit-card spending data or sort through construction permits. Thasos's specialty is spewing out of your smartphone.
Thasos gets data from about 1,000 apps, many of which need to know a phone's location to be effective, like those providing weather forecasts, driving directions or the whereabouts of the nearest ATM. Smartphone users, wittingly or not, share their location when they use such apps. Before Thasos gets the data, suppliers scrub it of personally identifiable information, Mr. Skibiski said. It is just time-stamped strings of longitude and latitude. But with more than 100 million phones providing such coordinates, Thasos says it can paint detailed pictures of the ebb and flow of people, and thus their money.
Last month, many on Wall Street were surprised when Tesla disclosed a rare quarterly profit, the result of Model 3 production that had nearly doubled in three months. Shares shot up 9.1% the next day. Thasos is at the vanguard of companies trying to help traders get ahead of stock moves like that using so-called alternative data. Such suppliers might examine mine slag heaps from outer space, analyze credit-card spending data or sort through construction permits. Thasos's specialty is spewing out of your smartphone.
Thasos gets data from about 1,000 apps, many of which need to know a phone's location to be effective, like those providing weather forecasts, driving directions or the whereabouts of the nearest ATM. Smartphone users, wittingly or not, share their location when they use such apps. Before Thasos gets the data, suppliers scrub it of personally identifiable information, Mr. Skibiski said. It is just time-stamped strings of longitude and latitude. But with more than 100 million phones providing such coordinates, Thasos says it can paint detailed pictures of the ebb and flow of people, and thus their money.
Re: (Score:2)
Next up: Driving around the area with a carfull of smartphones to skew the data.
Re: (Score:3)
Re: (Score:3, Insightful)
We are not most people.
Re:easy as hell to avoid (Score:4, Interesting)
We are not most people.
But we still are too dumb to RTFS.
They are NOT using location data from apps. They are just counting cell phone transmissions.
They are just trying to estimate the aggregate number of people in the factory, to see if they are really working late to ship product.
Similar tactics have been used in the past. For instance, when satellite images first went on-line, hedge funds developed software to count cars in mall parking lots. This put them weeks ahead of other investors that were waiting for the Fed to release data based on retail surveys.
Another tactic is to photograph cargo ships entering and leaving port. If they sit high in the water, that means many of the containers are going back empty, and the trade deficit is more likely to widen, which means a weaker dollar.
Re: (Score:3)
One I remember being famous back in the day was Iomega had released the Zip Drive, and online communities tracked the number of cars in the parking lot on the weekend.
Certain companies do things to obscure their energy use to try to keep that from leaking information. It gets complicated...
Re: easy as hell to avoid (Score:3)
"But we still are too dumb to RTFS."
We're not dumb - the fucking story is paywalled. And unfortunately search result quality on Big Brother Google has for some time been downgraded to show only semi-official fake news sources. So a quick search turned up no liberated copy of the article.
Anyways, the both summary says they are buying databases of surveillance data from smartphone malware/apps. "Thasos gets data from about 1,000 apps, many of which need to know a phone's location to be effective".
Certainly so
Re: easy as hell to avoid (Score:2)
Most people know how to click "No" when the app asks for permissions. Most people just don't care.
Re: (Score:1)
Facebook apps should already be uninstalled or disabled.
Re: (Score:2)
That's not a problem with the app, and changing the app won't make "most people" any less stupid and gullible.
Disinformation... (Score:5, Interesting)
Re: Disinformation... (Score:2)
They didnâ(TM)t use location data from apps. Phones connect to telecommunication masts, so itâ(TM)s just a matter of counting the number of phones by analyzing how many are connected to the masts in a certain area
Re: (Score:2)
Re: (Score:3)
That used to be true. Towers are far more precise now than they used to be. Still might be worse than GPS (it depends), but good enough for this purpose.
Re: (Score:2)
There's no way tower-based locations are as good as GPS, at least when GPS is available. (Inside buildings and under foliage might be a different question.) There just aren't enough towers to triangulate a phone's location using mobile network signals as well as a phone can triangulate its own location using GPS signals -- those are optimized for location performance, whereas mobile network signals are usually optimized for some data throughput measure.
Re: (Score:3)
The issue is GPS is permission based and the phone must be able to "see" the birds, and have network connectivity
"aren't enough towers to triangulate a phone's location"
The phone is constantly polling cells providing sending signal strength. Even if there's not sufficient signal to use, the network shares that information in order to make a handoff unnoticed. Triangulation of analogue signal was used extensively in WW2 with very accurate results.
These days history is in the logs, and it's trivial to
Re: (Score:2)
You don't even need multiple towers. There are many antennas on each tower, each of which has different sensitivities in different arcs. Towers now pass off from directional antenna to directional antenna.
Re: (Score:2)
Cell towers do have multiple sectors, but I find it hard to believe that the channel state information from two of them (because any given phone will not have a line of sight to three sectors) is enough to determine the phone's location. The environment is full of weird reflectors and obstructions that make the signal not just fall off with a power of the distance.
Re: (Score:2)
There just aren't enough towers to triangulate a phone's location using mobile network signals as well as a phone can triangulate its own location using GPS signals
Wait for 5G to be widely deployed, the number of "towers" involved is insane. The planning looks more like a wireless mesh network than a traditional mobile network (lots of small installations on traffic lights/light poles rather than large installations in high spots).
Re: (Score:3)
b) They snooped on Tesla because there was a specific reason to do so. If Ford was making similar claims about increasing hours, they'd have a reason to snoop on Ford.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
You get GPS spoofers for mobile phones. They intercept the system call to retrieve the GPS location and provide a user supplied coordinate. I relocated to the Mongolian desert out of boredom and started getting Chinese SMS messages from manufacturing companies looking for new business.
Re: (Score:2)
and does not send your location and your data to whoever is paying for it.
And strict clear laws for how data collection can be done and used.
Yes, the data is there, regardless what you do,
even with a dumb-phone your operator will know which tower you are connected to.
Thats why (Score:4, Interesting)
What can't be seen from space can now be tracked per worker per shift from the ground.
Who works in an office, who works on the mil/gov production line.
The skilled workers that get the mil production line working after it stops.
What areas of a mil production line still have the most visits by experts.
Wonder what a nations top police, city government officials, mil contractors and mil do all day? Map it out and find out
Re: (Score:3)
Re: (Score:2)
The smarter nations don't allow their security cleared workers to bring consumer hardware in with them.
Yeah, and just this year it took some random person looking at publicly published fitbit data to point out the error of the US government's ways.
Remind me again how the proactive smart people banned the data uploads before this became an issue?
Re: (Score:2)
That would be the security support offered by the NSA and GCHQ to ensure nothing was detectable from any interesting site?
The next problem is the factories that make mil equipment.
Then the police, governments, contractors.
In the old days every worker was security cleared and the secure building was safe from look down, had no windows.
Now everyone interesting wonders around with global tracking devices for their shift.
Better quality walls to block t
Re: (Score:2)
Similar stories (Score:5, Interesting)
The FedEX guy picking up the packages from our company in 1990s said he noted down the number of packages being shipped on the quarter ending evening. He always bought the stock if it showed significant jump over the previous quarter.
For all that talk about these techniques, the day before Tesla announced its third quarter results, real old fashioned bean counting led a very notable short (named Left, some company called Citron) to reverse course and announce publicly he was no longer shorting Tesla. S3 partners calculated that announcement made 1 billion loss for the shorts.
So, yes, novel methods are being discovered. But it is rounding error compared to old fashioned standard research.
It back fires as often as it hits pay dirt. (Score:3)
Very funny to read about some photo showing one car with hood open, and these shorts speculating, "they are fixing something in the engine". (The frunk is empty, it is a small storage compartment and you open the hood to put the car in tow mode to load it on a carrier. )
Re: It back fires as often as it hits pay dirt. (Score:1)
Where I live a certain large store went out of business. Now that store's equally large parking lot is full of brand new Teslas. Hundreds of them. I have no idea if the stock is stagnant, and/or being replenished.
Re: (Score:2)
Pick any car model that sells about 250,000 units a year. Subaru Legacy, Hyundai Sonata ... How many new cars you see in their dealership lots. Is the Tesla lot bigger or smaller?
Re: (Score:3)
Doesn't Tesla build to order though? Dealerships stock unsold vehicles but you currently need to be on a wait list for a Tesla vehicle, there's currently no way to buy a Tesla off the shelf, so shouldn't Tesla be shipping them as they come off the production line?
Re: (Score:2)
All it means is their delivery infrastructure hasn't scaled up properly to meet the new production rates.
Re: (Score:2)
All part of growing up. The production cost per vehicle dropped 30% from Q2 to Q3, due to training and experience of the workers. The started with such poor training and planning. Well, they will eventually sort it all out. Driving a BEV is like having gasoline at 80 cents a gall
Re: (Score:2)
It collects orders, and uses some AI to predict the sales, and builds them in batches. It misfires, it under predicted the demand for AWD, and they had unsold RWD. Looks like it over predicted the performance version and they are clearing the inventory with price reductions and incentives. They have a huge 3750$ tax cut reduction coming on Dec 31. That will help them clear all the inventory. Al
Build To Stock vs Built To Order (Score:2)
No, Tesla does not build to order.
Sure they do. They also do some amount of Build To Stock. Doing one does not preclude doing the other, even with the same product going down the same assembly line. My company does some similar things. We mostly BTO but when we have good reason to suspect a customer will order more of a specific product we often will BTS some extras units so we can save money on setup costs and buy materials in bulk. We are taking a calculated risk by doing so but it usually works out in our favor. Tesla if they are s
Re: (Score:2)
"Alternative data" (Score:2)
Next quarter (Score:2)
Tesla announces even more production, buys thousands of flip phones and randomly turns them on and off inside the factory all day long.
Re: (Score:2)
That's illegal stock manipulation.
Re: (Score:2)
b) It could be both trolling surveillance AND illegal stock manipulation.
Fuck 'em. (Score:2)
I don't know about you but as far as I'm concerned, these Thanos guys can eat shit.
How anononymous is the data? (Score:2)
Big data suppliers consistently ignore their published policies and often break the law and nothing serious happens to them. Just naming Facebook and Twitter proves the point.
If you take this story at face value you deserve to be hoodwinked and manipulated.
Parallel Construction (Score:1)
Could this just be clever "parallel construction" to cover insider trading? Buy some satellite data or phone location data and use it to prove what you already know.