Tesla is a good example of why alternative data hasn’t helped the hedge fund industry as much as hoped - even if you knew sales or earnings in advance, it’d have been hard or impossible to make $. We live in cash flow from financing world, not cash flow from operations world
-
My response to “why hasn’t alternative data helped the hedge fund industry” is two-fold. First, much of the data is bad (eg low breadth, short history, poor data quality, not updated in a timely manner etc) or useless (already baked into earnings estimates or reflected in price)
Second, it’s not true that alternative data hasn’t helped. Some of it has been and still is very predictive, you just need to know where to look (and have tested a lot of data sets out of sample).
-
-
Would like to see results of the strategy in my first tweet though, as I may be wrong.
-
quarterly releases contain a lot more info that can move the stock in the opposite direction from what you’d expect just from the earnings number. still I think your strategy would work especially on the outliers. isn’t this how people make money insider trading?
End of conversation
New conversation -
-
-
would agree the issue with alternative data is that a lot of it might have issues, but there will be some datasets which are useful. I often hear about which datasets are not very useful, but for obvious reasons if folks find some useful, they'll generally be quiet.
-
Not only keep quiet, but actively try to prevent discovery eg with exclusivity agreements, lawsuits or even buying the firm that produces the data and in-housing it.
End of conversation
New conversation -
-
-
Anecdotally I have heard inhouse NLP extracted sentiment does create uncorrelated and performing signals for US single stocks even in tough year like 2019 (for statarb) or 2020.
-
I have heard similarly but am somewhat skeptical… I think you have to manually curate your inputs extensively and combine with flow signals. Could be wrong, but “nlp + Twitter firehouse” just seems like you’d be chasing pump and dumps
End of conversation
New conversation -
-
-
i worked with alex, he absolutely knows that alternative data can help :) he basically pioneered it at our firm. but was also abundantly clear that you need to back it up with some qualitative info as blowups from crowded data trades hurt really bad (i still wince at
$SKX 2015)Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.