Hacker News Reader

Comment by timthelion | Hacker News Reader

Is it really fair to use such non-public information? How far can you go till it becomes insider trading. I'm aware that some investors have been looking at parking lot volume via satellite imaging in order to predict the performance of retail companies. Is that fair? Would it be insider trading if Amazon trades stocks of companies which sell through Amazon? Amazon would then ACTUALLY KNOW their sales numbers. Perhaps this whole insider trading thing is a broken idea to begin with. What if I figured out the sales numbers for a cell phone manufacturer by analyzing MAC addresses using a network of wifi sniffers, sales numbers for car manufactures by analyzing traffic camera data using ML. Used security camera footage from stores that I own to figure out what kind of hand bags people caried and then shorted Prada when I knew fewer people were bringing Prada bags into the store...

View on HN

Funny enough that Google is exactly in that position. Google owns a company called double click who does adword bidding optimization in between other things. As I work in the business, one thing that is clear is that the only way to do that efficiently is to keep track of conversion data. This puts Google in a highly unfair position since they manage the auction and own the data which drives the auction. That said, this is not even close to what Google knows about businesses in general...

But the really interesting thing is not what Google can do with the ad exchange. But what they can do with that ad buy data in order to reconstruct the sales numbers for a company and unfairly play the derivitives markets around it's earnings call. When I say unfairly, I mean that ONLY google is privy to such knowlege. If there is any reason to think that insider trading is wrong, then Google would be similarly in the wrong here.

Indeed, when I think about it. Isn't big data in a way "insider trading via induction". Like isn't the goal of big data to gain the type of information that you're not aloud to take advantage of due to insider trading laws?

Yes, this has been used as a defense against insider trading accusations, sometimes unsuccessfully: https://en.m.wikipedia.org/wiki/Mosaic_theory_(investments)

But isn't it equally true, that a mosaic of publicly known information could constitute non-public information? For example, the example I gave of analyzing satellite imagery of parking lots to predict Walmart sales numbers. Is that really fair?

I don't think you can deduce what is legally insider trading from an idea of what is "really fair". There really isn't much of a relationship. Matt Levine frequently writes about this.

You are allowed to trade based on non-public information. It doesn't matter if it's fair or not.

However, I'm not sure why this wouldn't be fair. More accurate valuations seem like a positive thing.

Another example I've been wondering about lately... Could Mercer and Renaissance potentially be using data from Cambridge Analytica (he is a partial owner) to identify market opportunities or even make markets?

Reply on news.ycombinator.com