Cash

Hi all,

today I ran a simple screen which uses cash as parameter and is limited to 100 stocks:
https://www.portfolio123.com/app/screen/summary/179630

Can anybody explain me why over large periods the screener cannot even find 100 stocks within the Russell 3000 universe?

If I don’t use any stock limit, then only from mid 2010 onwards the screen finds around 3000 stocks. From this it would appear that the databse has no cash data prior to this date, but such kind of limitation is not mentioned in the function reference:
https://www.portfolio123.com/doc/doc_detail.jsp?factor=Cash&popUpFullDesc=1


Weird

CashEquivTTM and CashA seem to work back in time.

Cashttm seems to break, though.

The CashTTM data is NA for many, many companies before 2010 (see results for buy rule “CashTTM > 0”).

When working with “rare” factors, I always check for NA’s.

It would be great, if P123 could display the data availability of all factors à la “x% of companies have reported this number”.


Cash is kind of strange thing in data collection. The call of whether to put a particular current asset into cash or equivalents or to leave it as its own line is all judgement, based not only on company reporting but on the history of the data collection for that company. I don’t know if anyone still has access to the Reuters database, but last I saw neither The Kroger nor Activision Blizzard had a pure cash line in it. Instead, based on where they put it in the past, Reuters reports it all as equivalents. I don’t have any specific examples from CompuStat, but clearly there’s the same sort of thing going on.

We give the pure cash line for use in analyzing extremely low quality companies: At the end of the day, cash per share is usually the absolute minimum value for a company. For everyday use, you want to use Cash and Equivalents. CashEqivTTM is much more consistently filled in, too.

Thanks Paul!