- Small Japanese net-net
- Market capitalization: ¥2,743 million (US$22.37 million)
- P / t. B = 0.28
- P / NCAV = 0.38
- P / Liq. value = 0.56
- P / NNWC = 0.61
Katsuragawa Electric manufactures large format printers, copiers and micro motors. This obviously isn’t exactly a growth industry. I expect the printer business, in general, to slowly decline in the future. I think, however, at these prices, the stock is trading significantly below intrinsic value. At ¥179, the stock is trading at a 62% discount to NCAV (including net long-term investments). The company is currently marginally unprofitable. Management guides to a profitable FY 2016 (ending March 2016), however. They expect ¥110 million in ordinary income, from which ¥10 million is attributable to holders of the parent, ¥140 million in operating income and ¥10,500 million in sales for FY 2016. Management identifies the usual suspects to restore profitability: streamlining, reducing fixed costs, better inventory management, cutting executive compensation, firing people and developing new sources of revenue – nothing extraordinary that qualifies as a hard catalyst. On the other hand, I also couldn’t find (which doesn’t mean there aren’t) any signs of major adverse litigation, off-balance-sheet-liabilities etc. that might indicate severely negative surprises in the future. Below is the translated balance sheet, calculation of NCAV, NNWC and a conservative approximation of liquidation value:
(Source: Quarterly Reports for the First and Second Quarter Ending June 30, 2015 and September 30, 2015 Translated via Google Translate)
I want to reemphasize that I can’t read a word in Japanese and might miss something important. This stock represents only a small position in my diversified net-net basket. Decisions in this basket are based almost entirely on quantitative variables.
Disclosure: The author is long 6416:Tokyo.
Deep Value Ideas 
Hey, cool blog. I like how your posts are succinct, but still impactful and insightful. During your idea generation process, have you run across Sorl Auto Parts; SORL? It is statistically cheap – as you like – and management has offered to take the company private. Cash + ST investments (net of all liabilities) is more than double the market cap.
LikeLiked by 1 person
Thanks a lot for your kind words! I remember SORL coming up on my screens some time ago. I didn’t investigate further, though, as I’m very skeptical of net-nets with their main assets in the PRC. I wasn’t aware there was an offer from management. This is indeed very interesting. I’ll take a look. Thanks for bringing this to my attention!
LikeLike
No problem. Your skepticism is well placed as there have been quite a few frauds perpetrated by PRC-operated companies listed in the U.S. (reverse merger frauds). We’ll see if shareholders approve of the deal given how egregious (cheap) the valuation is. Happy that I could bring the idea top your attention, keep up the excellent work.
LikeLiked by 1 person
Hey! Thanks for sharing your ideas. What do you think about the fact that this stock and a lot of the Japanese net-net stocks didn’t trade above NCAV for more than 5 years (at least, sometimes 8 years).
Best regards, Pascal
LikeLiked by 1 person
This is a good question. I don’t have a rule regarding that. It might be that stocks that only recently fell below NCAV do, c.p., better than the long-time net-nets. On the other hand, frustration in stocks that have been trading below NCAV a long time is probably much higher. This might be a non-fundamental factor that depresses the price. Which is good, I think. Also, Greenblatt once mentioned in his lecture at CBS that 5 years of bad performance is a positive momentum factor. I don’t know much about that, however, and I don’t use momentum in any way. People often talk about avoiding long-time net-nets. But I think their experience is strongly biased. They likely remember the frustration of a non-performing net-net much stronger than a net-net that did as expected. Their sample seize is probably much too small as well. The correlation between prior years of trading below NCAV and future performance is actually a thing you could backtest. I might do this when I have the chance. One problem I can see with such a test already is insufficient sample seizes. There may be not enough stocks in either of the groups.
What I’m doing at the time is basically making the NCAV data for different stocks as comparable as possible and investing in the top 20 cheapest I can find while avoiding obvious frauds and high leverage. I rebalance as often as reasonably possible. What that means is I don’t sell a stock at P / NCAV = 0.4 to buy a stock at P / NCAV = 0.39. But if I find a stock at 0.33, I probably would rebalance. The important thing is: (Ignoring trading costs and tax-effects) there are no “holds”. If I wouldn’t buy, I will sell. So, If a stock that I bought at a deep discount now is trading at P / NCAV = 0.95, I will sell that stock. I surely wouldn’t buy it if I could find a similar stock at a 67% discount. This means that I don’t hold stocks until they reach their NCAV anyway. And the problem that they haven’t traded above NCAV for a long time becomes somewhat less severe.
LikeLike
Very good point, thanks. My strategie is to sell near NCAV. I will consider changing that strategy for my Japanese basket.
For this basket I might also change how I compare stocks that are depressed for more than five years and maybe use a modified version of P/NCAV.
LikeLike
Hi there,
I enjoy reading your write-ups. One thing I was wondering about is what tool(s) you use to screen for the stocks (Bloomberg, etc)? I’ve found that many databases are so-so in terms of quality for smaller foreign companies (as you point out, although Bloomberg does tend to get the Treasury share issue right).
LikeLike
Glad to hear that! I use Screener.co. For the relatively low price (in comparison to Bloomberg, at least), this is really a great product. You can build very sophisticated screens with it. It, however, doesn’t get the treasury share issue right without modifications. For some very strange reason, it provides the correct number of outstanding shares as a separate data point, though. So you can just calculate the accurate Mktcap yourself using the custom formulas.
LikeLike