Compustat / Bloomberg / FactSet Data

[west]

Hey all.  Over the last year or so, I’ve been working on a program to help me screen for undervalued companies in more nuanced ways than just your basic valuation ratios.  Recently I started using it to analyze historic data sets.  My goal in doing this was/is to write a paper about different quantitative value strategies and how well they’ve worked in recent years in Japan.

 

Unfortunately, I’ve hit a bit of a snag.  I scrape most of the financial data I use off of various free websites.  While this works great for getting recent financial data, it’s not so great for historic data, especially price data.  The data on these websites has a lot of flaws in it, and adjusting for all of these issues... well, let’s just say I’ve started to think that it won’t be a great return for all the time I’d need to invest in it to do it right.

 

Before I give up though, I was wondering if I could persuade someone to help me get access to a more professional source of financial data, like Compustat, FactSet or a Bloomberg terminal.

 

In exchange, I can show you what my screener can do (it does waaay more than just your basic comp/value ratios, like I said) and maybe do some custom historic analysis of your choice for you as well.  In addition, I can point you to some pretty sweet Japanese companies I’ve found with it (a net-net trading at less than 4x normalized FCF anyone?), as well as send you photos from the most recent Japan Company Handbook (Winter 2014 Edition) about those specific companies, or any Japanese companies you’d like for that matter.

 

I really appreciate any help I can get on this.  Especially since I don’t want to see all the software I’ve written so far or the back-testing analysis work I’ve done go to waste.  Thanks in advance.

 

west

[ZenaidaMacroura]

Not to derail, but I've been wondering how essential it is to have access to one of these data sources when researching potential investments.  I remember having them available in college but is that level of depth wholly necessary for value investing?  It would certainly be nice but is that level of nuance a pre-requisite for good due diligence?  *sorry end rant/question*

[Kraven]

Not to derail, but I've been wondering how essential it is to have access to one of these data sources when researching potential investments.  I remember having them available in college but is that level of depth wholly necessary for value investing?  It would certainly be nice but is that level of nuance a pre-requisite for good due diligence?  *sorry end rant/question*

 

Schloss made it work with the local paper, Value Line, a pencil and a pack of gum.

[racemize]

Not to derail, but I've been wondering how essential it is to have access to one of these data sources when researching potential investments.  I remember having them available in college but is that level of depth wholly necessary for value investing?  It would certainly be nice but is that level of nuance a pre-requisite for good due diligence?  *sorry end rant/question*

 

Schloss made it work with the local paper, Value Line, a pencil and a pack of gum.

 

Schloss was amazing.

[oddballstocks]

Hey all.  Over the last year or so, I’ve been working on a program to help me screen for undervalued companies in more nuanced ways than just your basic valuation ratios.  Recently I started using it to analyze historic data sets.  My goal in doing this was/is to write a paper about different quantitative value strategies and how well they’ve worked in recent years in Japan.

 

Unfortunately, I’ve hit a bit of a snag.  I scrape most of the financial data I use off of various free websites.  While this works great for getting recent financial data, it’s not so great for historic data, especially price data.  The data on these websites has a lot of flaws in it, and adjusting for all of these issues... well, let’s just say I’ve started to think that it won’t be a great return for all the time I’d need to invest in it to do it right.

 

Before I give up though, I was wondering if I could persuade someone to help me get access to a more professional source of financial data, like Compustat, FactSet or a Bloomberg terminal.

 

In exchange, I can show you what my screener can do (it does waaay more than just your basic comp/value ratios, like I said) and maybe do some custom historic analysis of your choice for you as well.  In addition, I can point you to some pretty sweet Japanese companies I’ve found with it (a net-net trading at less than 4x normalized FCF anyone?), as well as send you photos from the most recent Japan Company Handbook (Winter 2014 Edition) about those specific companies, or any Japanese companies you’d like for that matter.

 

I really appreciate any help I can get on this.  Especially since I don’t want to see all the software I’ve written so far or the back-testing analysis work I’ve done go to waste.  Thanks in advance.

 

west

 

You can get pricing data via TradeKing's API if you sign up for an account.  I don't think you need to even fund it, just sign up.

 

Is it possible to pull historical pricing data from EDINET, or yahoo.co.jp?

 

The Tokyo Exchange seems to have increased the amount of information they have: http://quote.tse.or.jp/tse_n/quote.cgi?F=listing%2FEdetail1&QCODE=1904&MKTN=T&cht=21&mode=M#chart

 

Looks like historical quotes going back three months, plus shares outstanding and book value.

[west]

Not to derail, but I've been wondering how essential it is to have access to one of these data sources when researching potential investments.  I remember having them available in college but is that level of depth wholly necessary for value investing?  It would certainly be nice but is that level of nuance a pre-requisite for good due diligence?  *sorry end rant/question*

 

I have never used a Bloomberg terminal, FactSet, or Compustat.  However, my gut feeling is that, at least for their screening capabilities, you probably don't need one.  My screener, while awesome for finding things in Japan where things are ridiculously cheap and you can actually screen for them, is much less useful in the US.  At least this has been my experience so far.  I still get most of my good ideas from other people or from following events.

 

[west]

You can get pricing data via TradeKing's API if you sign up for an account.  I don't think you need to even fund it, just sign up.

 

Is it possible to pull historical pricing data from EDINET, or yahoo.co.jp?

 

The Tokyo Exchange seems to have increased the amount of information they have: http://quote.tse.or.jp/tse_n/quote.cgi?F=listing%2FEdetail1&QCODE=1904&MKTN=T&cht=21&mode=M#chart

 

Looks like historical quotes going back three months, plus shares outstanding and book value.

 

I actually scraped the entire Japanese market 2013 price data from yahoo.co.jp and finance.google.com :).  I scraped both to compare them and find inconsistencies in the data.  The goal was to deal with the (hopefully minor amount of) inconsistencies by hand.

 

Unfortunately, there are too many to deal with.  And the price data on those websites doesn't always play nicely with the financial data I get from money.msn.com.  So I kind of just want to deal with one data source at this point, versus trying to get different data sources to play nicely with each other.

 

(An aside: I think I may have actually found a bug in Google Finance!)

 

[turar]

A library should have access to Compustat.

[west]

A library should have access to Compustat.

 

I checked at both the San Francisco (where I live) Public Library, which has a lot of awesome resources, and they don't have it.  The Berkeley Business library has it, but limits access to faculty and graduate students that have special permission to access it.

 

I'm guessing it's too expensive for them to grant general access to it.

 

[DoddDisciple]

I've been using Screener.co. I'm not sure if they have historical price data (I see there is some technical data, but I haven't played with it), but it seems to be sufficient for screening for net-nets.

[SpecOps]

You will probably find it difficult to get access to the data. These terminals usually have a limit to 1-2 users in the terms & conditions and anyone with the raw datasets will be a reseller and wont just give it out free.

 

Your only hope is to find a library that has one of them, or to ask one of the companies sales departments for trial access, but then you might have a problem publishing figures from this data.