/biz/ - Business & Finance

>>23136185
count myself lucky that I can cover without triggering an infinite squeeze because the other hedge funds shorted GME so much that the real float is too big for an infinite squeeze to happen (for now)

practically all of the short sellers must be underwater by now

quantitatively inclined anon here in response to the fellow asking about estimating FTD shares

The SEC provides FTD for all securities at regular intervals
https://www.sec.gov/data/foiadocsfailsdatahtm
the latest available is through August of this year, so I was hoping to see if I could correlate it to any of the short data metrics from ortex (Estimated SI, short volume) and try and make some predictions about FTD shares into September

Unfortunately even the best predictor (the finite differences derivative of the SI, i.e. daily change in estimated SI) was very crappy, with all of it's predictive power driven by a single FTD outlier of 347,372 shares on the 25th of August. Essentially the data is too noisy and incomplete for a "quick and dirty" empirical approach.

Just to ameliorate concerns about settlement time
(it takes 2 business days for share transactions to settle which is when shares actually FTD, but the SEC wasn't super clear about whether the date in their data was the transaction date or the settlement date)
I tried testing against predictors timeshifted by T-2, but then all the correlations went to essentially zilch (no relationship) which makes me think the SEC date is the transaction date.

I'm gonna try to put together some sort of differential equation model for the shares (imagine modeling a tank filling with water) but it's gonna take some time, because
a) I don't work in quantitative finance
b) I couldn't find anything relevant to work off with a cursory literature search - everyone just reads the SEC numbers

GME data I compiled if anyone wants to look at it:
http://www.mediafire.com/file/v7yxd8xtaki49n4/GMEdata.xlsx/file

>>22839379
>These kind of people have probably been shorting since the 30s
Most people think this, but it's actually not true
I recall Dopierala saying that 70% of shorts have a cost basis under $7 (I'll find the quote next) and that seems to line up with the short volume on ortex.

Most shorts are deep underwater right now if we look at the numbers. Their only way to make profit is if it hits $0 and goes bankrupt