Oil traders tried to use Apple mobility data to predict market recovery, left 'disappointe...

Jump to First Reply
Posted:
in General Discussion edited July 2020
Oil traders looking for ways to estimate fuel demand were disappointed by Apple's mobility data, a report claims

Credit: Apple
Credit: Apple


Apple in mid-April unveiled a new Mobility Trends Report tool that visualized anonymized Apple Maps data requests. At the time, traders thought they had found a clue to guess at the speed of fuel demand recovery from the coronavirus pandemic.

But the data didn't translate into activity in gasoline demand, Reuters reported. That's because the data was based on search information and not miles traveled.

At the time of release, and even as Apple added new regions to the tool, the Cupertino tech giant stressed the data privacy of the feature. That's despite the fact that the tool captures data for about 100 million iPhones in the U.S.

In lieu of using Apple's data, the traders have moved onto other sources of information -- such as TomTom or U.S. Energy Information Administration reports.

Comments

  • Reply 1 of 6
    Eric_WVGGeric_wvgg Posts: 975member
    Horace Dediu of Asymco, one of the very few tech analysts I have any respect for, published maps comparing state-reported cases of COVID-19 to these same trends. If there is any correlation… well it sure seems spotty.

    http://www.asymco.com/2020/05/11/covid-19-us-dashboard-may-11-including-mobility-data/

    (before anyone chimes in saying "well many nations and states are withholding data" well, yeah, but then the trend graphs would show inverse results.)
    edited July 2020
    cat52
     0Likes 0Dislikes 1Informative
  • Reply 2 of 6
    baconstangbaconstang Posts: 1,197member
    I often put my iPhone in 'Airplane mode' when I drive.  HA!
    watto_cobra
     1Like 0Dislikes 0Informatives
  • Reply 3 of 6
    anantksundaramanantksundaram Posts: 20,421member
    Sure, blame the stupidity of their models on Apple.

    As someone once said (I forget who, and I am paraphrasing), forecasting is difficult, when it's about the future.
    magman1979watto_cobra
     2Likes 0Dislikes 0Informatives
  • Reply 4 of 6
    I often put my iPhone in 'Airplane mode' when I drive.  HA!
    That’s not particularly relevant to what Apple was tracking. The mobility data was based on searches in Maps not actual movement. It tracked if people were looking for driving directions, public transit or walking and how those changed over time. It doesn’t track actual movement. 
     0Likes 0Dislikes 0Informatives
  • Reply 5 of 6
    DAalsethdaalseth Posts: 3,297member
    This is why I have no faith in “the market”. Too many investing analysts, and even investors wave data and algorithms like talismans without any understanding. They treat the business like magic. I’ve seen people dump money into dotcoms that had no business model or plan, or even much more than a concept, let alone hard idea. I’ve seen them put their life savings, or more commonly their clients life savings, into derivatives so complex no one understood what they were saying. In fact they said nothing, but money was bet, and lost, on them. Their understanding of business and economics is on the level of a palaeolithic nomad’s belief in magic. They might as well try to predict the market with chicken entrails. 


    watto_cobra
     1Like 0Dislikes 0Informatives
  • Reply 6 of 6
    entropysentropys Posts: 4,479member
    It’s called GIGO

    very common in modelling.
    edited July 2020
    wonkothesanewatto_cobra
     2Likes 0Dislikes 0Informatives
Sign In or Register to comment.