Finding Japanese Net-nets

[west]

Kind of busy, so I'll quickly respond.

 

I use a proprietary program I wrote to look for high ROIC, low ROIC variance, low EV/Normalized and TTM EBIT companies.  I add to this list companies that have low P/TBV plus medium to high-ish BVPS growth rates (with the growth having low variance from year to year).

 

I then pick through the list by hand, checking accruals, DSO+DIO+DPO+CCC, and that my program did EV and ROIC calculations correctly.

 

My position sizes are not more than 2% except for in the case where I need to because the minimum block size to trade the company requires me to go above that threshold.  And the company is worth the risk.

 

My total allocation to Japan is going to be 30%, maybe 35%, of my portfolio for everything for a grand total of about 15-17 Japanese stocks.

[west]

For those asking, the proprietary program is not going to do most people any good unless you're very confident with large data + data munging.  It does do output to Excel, which would probably be useful and I can post some time in the future, once I've run the program to get new prices again.

 

[west]

writser,

 

Thanks for pointing out Okayama Paper.  My program dropped the data for it for whatever reason.

 

I spent some time looking at it.  It's cheap on an assets basis (negative EV as of today), but I'm a little nervous about it's lack of earnings power and the fact that EBIT and EBITDA margins have been in decline for the last five years.  It could slip into negative earnings/FCF territory very easily if this keeps up.  Did you see this?

 

[randomep]

I'll add to the information overload with yet another data source. To gather your own info which saves money and allows you to find original ideas, you should get a list of all securities in the world and filter what you need.

 

The following is a wonderful source: it lists all the stocks in the world and gives key statistics. I use it as a first pass to get what I want. In this case it is Japanese small cap stocks.

 

http://people.stern.nyu.edu/adamodar/New_Home_Page/data.html

 

Then you can write a program or some such to mine the internet for data on your pruned list of stocks. The internet is like a giant database of information. All this is free.

 

AFAIK there is no free filter for netnet stocks.

 

I use finance.yahoo.com and ft.com for mining my data.

 

 

[west]

I'll add to the information overload with yet another data source. To gather your own info which saves money and allows you to find original ideas, you should get a list of all securities in the world and filter what you need.

 

The following is a wonderful source: it lists all the stocks in the world and gives key statistics. I use it as a first pass to get what I want. In this case it is Japanese small cap stocks.

 

http://people.stern.nyu.edu/adamodar/New_Home_Page/data.html

 

Then you can write a program or some such to mine the internet for data on your pruned list of stocks. The internet is like a giant database of information. All this is free.

 

AFAIK there is no free filter for netnet stocks.

 

I use finance.yahoo.com and ft.com for mining my data.

 

Great thought.  My data is essentially what's in Damodaran's Japan data sets.  I can dig into the data a little bit more than what you can with his data, but not much more.  They have most of what you need to know, and I think you should check the rest by hand anyways.

 

If I had known about his data when I first started I probably wouldn't have written my program.

 

Fwiw, here is a link to the beginning of 2014's data:

 

http://www.stern.nyu.edu/~adamodar/pc/datasets/Japancompfirm.xls

 

He also has data going back further but, unfortunately, you can't really do long-period back testing with it because he does weird Yen->USD conversions before, I think 2007 or 2008.  And they're not simple conversions.  He does the whole by-the-book process where, say, cash is at current exchange rates, COGS is at the average exchange rate over the period analyzed, etc.  It's all quite stupid.  I'm glad he wised up and started keeping things in yen.

[writser]

writser,

 

Thanks for pointing out Okayama Paper.  My program dropped the data for it for whatever reason.

 

I spent some time looking at it.  It's cheap on an assets basis (negative EV as of today), but I'm a little nervous about it's lack of earnings power and the fact that EBIT and EBITDA margins have been in decline for the last five years.  It could slip into negative earnings/FCF territory very easily if this keeps up.  Did you see this?

 

Yes, I agree. It's a bit of a smaller position. Not a good business probably but I really like the significant share buybacks at 0.3x TBV. To be fair, in general I'm not sure if my qualitative screening of these companies adds any alpha (probably it doesn't). I'm not a very proficient analyst and there is also the language / culture barrier. Nevertheless I see it as an interesting experiment for a small part of my portfolio. As long as I stick to the basket approach my personal opinions should not interfere with future returns too much. And the numbers just look unbelievable.

 

Thanks for the links both of you. My solution for picking stocks is a bit lazier easier: just look at all the stuff other bloggers / forum members come up with.

[matjone]

Thanks for your ideas guys.  These are the ones I've bought so far:

 

+7501.T

+7928.T

+9760.T

+2055.T

+2737.T

+3599.T

+3766.T

+5965.T

+6466.T

+6930.T

+7314.T

+7521.T

+7558.T

+7559.T

+7885.T

+8141.T

+8150.T

+9476.T

+9932.T

 

When I first started looking over there I was looking strictly for net nets.  I was using FT and reuters for data and I was lumping the long term investments and long term notes receivable items in when calculating NCAV.  Now I've realized that this was not correct.  The long term investment line can mean different things - sometimes it is marketable securities, and sometimes it is investments in associates.  So I have been discounting them when calculating NCAV.

 

Oh, one more thing - ft and reuters seem to be including treasury shares in the shares out when they calculate their market caps.  So I have been calculating it myself with the shares out on the balance sheet.

[cobafdek]

Thanks for your ideas guys.  These are the ones I've bought so far:

 

+7501.T

+7928.T

+9760.T

+2055.T

+2737.T

+3599.T

+3766.T

+5965.T

+6466.T

+6930.T

+7314.T

+7521.T

+7558.T

+7559.T

+7885.T

+8141.T

+8150.T

+9476.T

+9932.T

 

When I first started looking over there I was looking strictly for net nets.  I was using FT and reuters for data and I was lumping the long term investments and long term notes receivable items in when calculating NCAV.  Now I've realized that this was not correct.  The long term investment line can mean different things - sometimes it is marketable securities, and sometimes it is investments in associates.  So I have been discounting them when calculating NCAV.

 

Oh, one more thing - ft and reuters seem to be including treasury shares in the shares out when they calculate their market caps.  So I have been calculating it myself with the shares out on the balance sheet.

 

Great list.  Thanks for it.

 

I just did a quick glance of the summary stats for these.  Many seem to have poor-to-middling quality of earnings, at least when merely eyeballing 5-yr net income compared to cash flow (I'll need to look more closely).

 

Is your approach to give heaviest weight to discount-to-NCAV or discount-to-BV, and lesser weight to other considerations (such as earnings quality)? 

 

Also, are you avoiding any industries, such as retailers, specialty retailers, shipbuilding, etc.?

[writser]

Yeah, most of these have a < 2% ROE for the past decade, it looks like. Not really my cup of tea - though buying a basket of them is probably not a bad strategy. Couple of exceptions: Tomen Electronics looks decentish at first glance, growing revenue, growing book by 4% p.a., sporting a 2% dividend yield and trading around 0.75x NCAV. Tomen Devices also looks reasonable. Fujimak has been discussed already, I own that too. What I like about these names is that, even if everything in Japan stays the way it is (no activists, terrible capital allocation, market multiples stay the same) at least these names should grow by 5-10% annually.

[matjone]

I guess I have generally preferred lower p/b companies, especially net nets, as long as they are at least holding steady.  I am not sure that is the smartest strategy though, and I may branch out and buy some low pe stuff.  Actually I already did that with Fujimak.  And honestly I didn't look at quality of earnings while picking these.  I was simply buying a basket of net nets.

 

I've usually avoided net nets in the retailing sector because the ones I've seen were usually struggling businesses with a lot of their current assets in inventory of uncertain value.

 

Writser, where are you seeing a decade worth of financials for Japanese stocks?

[writser]

You can get the annuals from the TSE website and translate them using chrome, but I didn't do that. However, in Morningstar you can check the past 6/7 years on the summary page. If you look at, for example http://financials.morningstar.com/ratios/r.html?t=9760&region=jpn&culture=en-US I'm fairly certain that the CAGR for the past 10 years isn't spectacular (if it is even positive to begin with).

 

Obviously this is extremely superficial 'research'.

[bizaro86]

Yeah, most of these have a < 2% ROE for the past decade, it looks like.

 

Is the low ROE partially due to holding significant excess cash? Although Japan is not the US, it's still reasonable to exclude obviously excess cash from the ROE when calculating the quality of the business, imo.

[oddballstocks]

Yeah, most of these have a < 2% ROE for the past decade, it looks like.

 

Is the low ROE partially due to holding significant excess cash? Although Japan is not the US, it's still reasonable to exclude obviously excess cash from the ROE when calculating the quality of the business, imo.

 

Some is due to excess cash yes.  I believe if you adjust the ROE for deflation and productivity as well.  I remember reading an article or book about this a while back, if you do the calculations their adjusted ROE's are similar to the US.

 

This is a great thread, I own Fujimak as well.  Always in the hunt for more of these things.  I avoid any and all retailing net-nets regardless of the country.

[matjone]

Some of these companies are running overly conservative balance sheets, which is affecting their profitability.  9760 is a good example of what I was talking about earlier where the nature of the assets listed as long term investments is important.  If they are marketable securities then it is obviously undervalued even if there isn't a business there at all, and the business wouldn't look nearly as bad if you excluded those securities and the excess cash when measuring profitability.

 

What we need is someone who can read Japanese.  We are really shooting in the dark on this hunt.

[matjone]

Thanks for directing me to morningstar.  I didn't even know they had Japanese stocks on there.

 

I am inept when it comes to technology, so this may be a stupid question.  But how can you translate the reports in chrome?  When I tried finding annuals on TSE they only came up as PDFs.

[randomep]

Some of these companies are running overly conservative balance sheets, which is affecting their profitability.  9760 is a good example of what I was talking about earlier where the nature of the assets listed as long term investments is important.  If they are marketable securities then it is obviously undervalued even if there isn't a business there at all, and the business wouldn't look nearly as bad if you excluded those securities and the excess cash when measuring profitability.

 

What we need is someone who can read Japanese.  We are really shooting in the dark on this hunt.

 

I have Japanese relatives and I work in a Japanese company. On occasion I have prodded them to translate. But they don't help much at all, the small cap financial reports are just management-speak w/o delivering much information.

 

[writser]

Yeah, most of these have a < 2% ROE for the past decade, it looks like.

 

Is the low ROE partially due to holding significant excess cash? Although Japan is not the US, it's still reasonable to exclude obviously excess cash from the ROE when calculating the quality of the business, imo.

 

Sure, the ROIC is much higher than ROE for these companies due to all the excess cash & investments on the balance sheet doing nothing. But what's the point if they keep hoarding all this cash for decades? If nothing changes in Japan the 2% ROE companies will continue growing BV by ~2% p.a. So unless the market decides to rerate the P/B at which these companies are trading you are stuck with a basket of perennial net/nets with stock prices that increases by 2% p.a. on average. Granted, it is probably an ok approach to buy a basket of these stocks. You have a huge margin of safety and occasionally something good will happen either with the stock price or the underlying business.

 

However, given the choice I'd rather buy something like Fujimak, maybe not the cheapest net/net out there but it grows book value by ~10% p.a. I would expect that to be reflected in the stock price over the long term. The way I see it you get paid to wait. A somewhat higher ROE suggests either a better business or better capital allocation and that's what I prefer in the long run. Especially in Japan, given the lack of shareholder activism.

[writser]

Thanks for directing me to morningstar.  I didn't even know they had Japanese stocks on there.

 

I am inept when it comes to technology, so this may be a stupid question.  But how can you translate the reports in chrome?  When I tried finding annuals on TSE they only came up as PDFs.

 

The Japanese equivalent of the SEC filings page is EDINET: http://disclosure.edinet-fsa.go.jp/EKW0EZ1001.html . If you run a search query you can view the resulting documents online or download them either as PDF or XBRL. Download the XBRL version (it's a zipfile), unzip it somewhere and you can open the HTML-files it contains with Chrome.

[cobafdek]

Thanks for directing me to morningstar.  I didn't even know they had Japanese stocks on there.

 

I am inept when it comes to technology, so this may be a stupid question.  But how can you translate the reports in chrome?  When I tried finding annuals on TSE they only came up as PDFs.

 

The Japanese equivalent of the SEC filings page is EDINET: http://disclosure.edinet-fsa.go.jp/EKW0EZ1001.html . If you run a search query you can view the resulting documents online or download them either as PDF or XBRL. Download the XBRL version (it's a zipfile), unzip it somewhere and you can open the HTML-files it contains with Chrome.

 

How are your Google Translates working with Japanese?  My past experience has resulted in some bizarre translations, possibly related largely to different syntax (Japanese and Korean syntax is heavily dependent on declension/conjugations, kind of like Latin, whereas English and modern European languages and even Chinese rely more on word order in sentences). 

[oddballstocks]

Thanks for directing me to morningstar.  I didn't even know they had Japanese stocks on there.

 

I am inept when it comes to technology, so this may be a stupid question.  But how can you translate the reports in chrome?  When I tried finding annuals on TSE they only came up as PDFs.

 

The Japanese equivalent of the SEC filings page is EDINET: http://disclosure.edinet-fsa.go.jp/EKW0EZ1001.html . If you run a search query you can view the resulting documents online or download them either as PDF or XBRL. Download the XBRL version (it's a zipfile), unzip it somewhere and you can open the HTML-files it contains with Chrome.

 

How are your Google Translates working with Japanese?  My past experience has resulted in some bizarre translations, possibly related largely to different syntax (Japanese and Korean syntax is heavily dependent on declension/conjugations, kind of like Latin, whereas English and modern European languages and even Chinese rely more on word order in sentences).

 

My $.02, if you're investing at a big enough discount I'm not sure the minutia in the filings matters all that much.  If you can read the essence of the filing and know that the business isn't going under then you should be fine.  I will invest in these things without reading anything beyond the financials or a few metrics.  Most investors probably think I'm crazy, and I'd agree to that, but to each his own.

[writser]

Agreed with Nate: in most cases I'm fine with an imperfect translation. I uploaded the Fujimak annual a while ago: http://www.writser.nl/fujimak/translated.htm , translated in English. I'd say it is quite understandable. Good enough for me in any case.

[randomep]

Thanks for directing me to morningstar.  I didn't even know they had Japanese stocks on there.

 

I am inept when it comes to technology, so this may be a stupid question.  But how can you translate the reports in chrome?  When I tried finding annuals on TSE they only came up as PDFs.

 

The Japanese equivalent of the SEC filings page is EDINET: http://disclosure.edinet-fsa.go.jp/EKW0EZ1001.html . If you run a search query you can view the resulting documents online or download them either as PDF or XBRL. Download the XBRL version (it's a zipfile), unzip it somewhere and you can open the HTML-files it contains with Chrome.

 

How are your Google Translates working with Japanese?  My past experience has resulted in some bizarre translations, possibly related largely to different syntax (Japanese and Korean syntax is heavily dependent on declension/conjugations, kind of like Latin, whereas English and modern European languages and even Chinese rely more on word order in sentences).

 

My $.02, if you're investing at a big enough discount I'm not sure the minutia in the filings matters all that much.  If you can read the essence of the filing and know that the business isn't going under then you should be fine.  I will invest in these things without reading anything beyond the financials or a few metrics.  Most investors probably think I'm crazy, and I'd agree to that, but to each his own.

 

No I don't think you are crazy at all!

 

Let's invert the problem. When was the last time you lost money because you failed to read some minuta in the management discussion?

 

When I look at financials I just jump to the balance sheet and income statement. Then I look around and get an impression of the company from their website (esp. for japanese companies)

[west]

Yeah, most of these have a < 2% ROE for the past decade, it looks like.

 

Is the low ROE partially due to holding significant excess cash? Although Japan is not the US, it's still reasonable to exclude obviously excess cash from the ROE when calculating the quality of the business, imo.

 

Sure, the ROIC is much higher than ROE for these companies due to all the excess cash & investments on the balance sheet doing nothing. But what's the point if they keep hoarding all this cash for decades? If nothing changes in Japan the 2% ROE companies will continue growing BV by ~2% p.a. So unless the market decides to rerate the P/B at which these companies are trading you are stuck with a basket of perennial net/nets with stock prices that increases by 2% p.a. on average. Granted, it is probably an ok approach to buy a basket of these stocks. You have a huge margin of safety and occasionally something good will happen either with the stock price or the underlying business.

 

However, given the choice I'd rather buy something like Fujimak, maybe not the cheapest net/net out there but it grows book value by ~10% p.a. I would expect that to be reflected in the stock price over the long term. The way I see it you get paid to wait. A somewhat higher ROE suggests either a better business or better capital allocation and that's what I prefer in the long run. Especially in Japan, given the lack of shareholder activism.

 

writser, there are a few others like Fujimak that I haven't posted yet.  I've been slacking on buying them because I hate pulling the trigger on things.  I like to overanalyze things.  Forever.

 

Pester me regularly (I don't mind) and the chances are much better you'll see the ideas sooner :)

[writser]

I assume you don't want to share before buying into a full position? Fair enough :) . I really like your ideas so far, appreciated!

[cobafdek]

My $.02, if you're investing at a big enough discount I'm not sure the minutia in the filings matters all that much.  If you can read the essence of the filing and know that the business isn't going under then you should be fine.  I will invest in these things without reading anything beyond the financials or a few metrics.  Most investors probably think I'm crazy, and I'd agree to that, but to each his own.

 

Thanks, all.  This approach accords with my experience with English language 10Ks as well!

 

I'll revisit the question I brought up in a related thread, which is when to sell.  Assuming I will have only these essential metrics in the future, with no plans of acting like a faux owner/operator, I will sell in 3-5 years if the stock is dead-in-the water, or when it reaches NCAV, whichever comes first.  West had an interesting approach, loosely based on the magicformula yearly-turnover tactic (see the Car Mate thread).  But in general, I'm with writser, and will procrastinate when it comes to the actual time of sell-decision.

 

Any other approaches would be welcome to hear.