Nam Tai: A net-net with multi-bagger upside

Nam Tai Electronic (NTE) is a contract electronics manufacturer with factories located in China. It has a market cap of $270m and no debt. It's been in business for 40 years and listed on NYSE since 2003. It business has changed and gone through many phases. Currently it assembles LCD modules (LCM), with Apple as the primary customer. It has always been profitable and has never lost a single dime in the last 10 years. However, it has warned investors in the recent years competition is intense and its contracts could end abruptly. On Jan 27, it confirmed it would receive no more orders and all production would be terminated for good in April. It also confirmed it would exit its existing lines of business completely.

Investment thesis

NTE is a net-net with a hidden asset worth multiple of its existing market cap. Its management is competent and aligns with shareholders. Catalyst is already in place. The management has committed to realise its value.

Downside

At the time of writing, NTE is trading at around $6. Current assets less all liabilities is $300m ($6.65 per share) and book value is $360m ($8.10 per share). In other words, NTE is trading at 10% below NCAV and 25% below BV. And most of the current assets is cash.

The number one risk investing in any company with any China connection is fraud. In the case of NTE, this risk is virtually zero. Every evidence indicates it's a legitimate business run by a legitimate businessman: Its primary customer is Apple. It sells real electronic components and receives real cash. In the last 5 years, it distributed $58m dividends in total with an average payout ratio of 70%. There was no capital raising nor debt.

Or we can look at it from the history perspective. Nam Tai was founded by Hong Kong entrepreneur Koo Ming-kown (as in Chinese tradition, the surname goes first; so, "Koo" is the surname). He is currently NTE's Chairman and CFO and holds 11.7% of NTE. Another director Peter R Kellogg holds 14.4%.

NTE's history shows Koo is an opportunist and capital allocator. NTE's business has gone through 3 major transformations in the last 4 decades. NTE started as a distributor of Japan-made calculators. At its peak, about half of the calculators used in Hong Kong were imported by NTE. In 1979, taking advantage of China's economic reform, Koo expanded the business into China and started manufacturing his own brand of calculators in China. Later, around 1985, economic environment became harsh and NTE's business went downhill. With its existing relationship with Japanese manufacturers, Koo seized the opportunity to transform the business again to export electronic components and raw materials from China back to Japan.

Upside

One of NTE's factories locates in Bao'an, Shenzhen, China. NTE now intends to redevelop this land into a commercial complex and become a property developer and manager.

A quick discussion of the economic backdrop of this region is warranted here.

In 2010, Chinese central government extended the Shenzhen Special Economic Zone (SEZ) to include Bao'an and Longgang. Besides, Qianhai, the bay area in Nanshan, is earmarked to be turned into an international financial services centre. It is China's ambition to have its currency RMB playing a more significant role in global finance. 

Expansion of Shenzhen SEZ (Source: China Daily News)

NTE's factory is 7km from Bao'an CBD, 12km from Qianhai CBD and 10km from Shenzhen's international airport. Local government has recently been rezoned the region from industrial use to commercial use.

"A" is NTE's site. Left box is Shenzhen's international airport. Middle box is Bao'an CBD. Right box is Qianhai district.
(Source: NTE new releases, Google Map)

In China, like Singapore and Hong Kong, lands are not bought. Instead, land use rights are leased. The total historical cost of land use rights is stated at $10m on NTE's balance sheet. Bao'an site makes up probably less than half of that. i.e. it makes up less than 1.5% of the its book value. In other words, it's not material.

NTE's site has a land area of of 52,600 sqm. It can accommodate a business complex with a plot ratio of ~6. That means, after redevelopment, it will have a 300,000 sqm gross floor area (GFA). How much is this land parcel worth now after the rezoning?

NTE's Bao'an site (Source: NTE new releases, Google Map)

We can look at comparable land sales (or more correctly, auctions of land use rights). Two parcels of land in Qianhai with plot ratio 6.5 & 7.8 (T201-0077 & T201-0075) were auctioned in July 2013 and sold for ~$2,700 per sqm (GFA). Sale prices were ~50% higher than the opening prices. But that's Qianhai, the red hot region in recent months. How about Bao'an? Two parcels of land in Bao'an CBD with plot ratio 6.0 and 8.1 (A004-0154, A002-0046) are currently listed with opening prices at around $2,100 per sqm. Two years ago, a parcel of land a couple of kms outside CBD (A002-0042) was sold for $1,600 per sqm in Jun 2012.

So, if NTE's land can command $1,000 per sqm, it will be worth $300m. If it can command $2,000 per sqm, we arrive at a more aggressive value of $600m. Adding this back to the book value, we are looking at a value of $14 to $21 per share.

Alternatively, we can look at cash flow. NTE has commenced 3 feasibility studies for the project from 3 different consulting firms. The recommendations centre arround building a mix-use complex consisting of a hotel, offices, R&D offices, retail space and residential apartments, and some combinations of these. With assumed rental charges in the range of $0.8 to $3.0 per sqm per month (if you consider the current rental market in the area, these rental rates are very reasonable), the expected net rental cash flow will be in the vicinity of $50m per annual. If you apply a very conservative 10% capitalisation rate, the complex will be worth $500m (in 3-4 years). Apply a more aggressive cap rate of 7%, you will get $700m. i.e. We are looking at $11 to $16 per share. Here, we haven't added back the book value because the development will consume cash. However, it is almost a certainty NTE will mortgage the land to fund the development. There will be un-utilised asset value not accounted for here.

Why is it cheap?

The main give-away is the 11% drop on Jan 27th when NTE announced it would cease all manufacturing and committed itself to develop its lands. Its existing shareholder base sees it as a electronic manufacturer with good dividends, not a property developer that won't be able to distribute in the next few years and needs to lever up the balance sheet.

Second, value investors who screen for cigarbutts will shun the stock with its China connection.

Third, its small $270m market cap is pretty much not on most fund managers' radar.

Risks

Koo is 70 this year. NTE is his company. Without him, the direction of the company is completely unknown.

Second, my contact in China who is familiar with its business environment and property laws told me there is a real chance the local government will not approve NTE's redevelopment plan. Typically, the local government will have a lot of financial incentive to take back the lands and then auction them off. It will pocket the hefty profits while the original property owners will only be compensated for what the lands were original designated for. e.g. Another industrial land in a nearby region. Besides, the local government favours established larger developers and give them significant advantage.

However, the feasibility studies indicate the local government is "supportive" of NTE's project. One possible explanation is NTE's project fits well in the government's overall agenda. They want to fast track the redevelopment of the entire industrial region. There are rivalries between different SEZs. They are willing to give up a very small value in the overall scheme of things in order to get a pioneer to create a showcase. In addition, Koo's 3 decades of business relationship will the local government should also count. Undoubtedly, we have to trust Koo's experience and judgment here.

But even if the project is rejected by the government, what do we lose? We are still holding onto $1 worth of asset value that we have paid only 75 cents for it.

Third, if the redevelopment goes ahead, we won't see any cash flow for 3-4 years and NTE will lever up its balance sheet. Besides, China's grand policy plan for the Qianhai and Bao'an will take years to materialise. So, you need a 5+ years time horizon to invest in NTE.

Fourth, I've said for years China will slow down in the coming decades. We now start to see this effect on Australia's mining sector and its currency. In addition, China's huge debts in its shadow banking system is destabilising its economy. The chance of a catastrophic collapse of its economy or property market is remote but possible. This can derail the planned economic development of the region for 5 to 10 years. Or more.

Closing thoughts

One way to look at this investment is, it is buying a property development project at a discount. Another way to look at it is, the property development project has tremendous option value. We are effectively getting paid to hold this option.

How often do we see management of companies operating in deteriorated business environments hanging onto their status quo, knowingly to lose money years in years out, hoping to "turn around" the businesses? When the market sees NTE's exit from its existing business a negative, we should see it a strong positive. Koo refuses to continue to operate a business that will no longer earn its cost of capital. It is Koo's brilliance to turn hindrance into opportunity. We have seen it happened before multiple times in NTE's history. Buying NTE is to bet that he will do it again. But strangely, this bet costs us nothing.

Resources

• NTE's New Releases (in English) - including SEC filings and the 3 feasibility studies
• Shezhen Land and Real Estate Exchange Center (in Chinese)
• SoFun (in Chinese) - online rental markets and land transaction data
• Brief bio of Koo and NTE (news piece in 2003, in Chinese)
• Koo commented on its redevelopment plan (July 2013, news piece in Chinese) - He speculated the land was worth RMB 9 billion!  (USD$1.5B).

*I didn't discover NTE myself. The credit goes to Batbeer2 who published an investment case on Gurufocus in July 2013. I also owe Batbeer2 a thank you for flushing out some of the details.

(Disclosure: Long NTE & AAPL)

Value of the century - the Aussie edition

I spent some time last month combing through the bottom end of the ASX market, looking at companies with market cap less than A$300m. I haven't found anything worth investing so far. Then, last night I saw Whopper's post ACGX: value of century? I just couldn't resist and decided to write this up for your amusement.

Details at a glance:

• Company: Richfield International Limited (ASX: RIS)
• Market Cap:A$1.6m
• NACV: A$6.3m, mostly cash!!
• Business: It operates shipping services in Singapore. It used to operate trucking services years ago.
• Profitability: Mildly profitable or break-even most of the years.
• Dividends: Never in the last 8 years.
• Insider ownership: 72% controlled by the directors

So, here you are. A crazy pile of cash. Buying a share of RIS is buying 25 cents dollar. Or, to flip it around, you are getting an instant 4 bagger. While the value isn't as extreme as Whopper's ACGX, RIS is a listed company with audited accounts and regular filings.  

What's the catch?  

I have serious doubt the company exists mainly for the purpose of running a business. It looks more like the directors' tax shelter. It's basically their family's piggy bank. The most likely end-game I can foresee is, when the business dries out, they will shut down the operations and gradually draws down the cash pile as salaries until it reaches zero. Well, there is no certainty. They may declare a surprise special dividend or announce someone tendering for the company. But I won't bet on it.


(Disclosure: No position)

Will PGNT become a mini replica of BRK?

No, Paragon Technologies isn't Berkshire Hathaway. But the current situation shares some interesting aspects of the old BRK when Buffett bought it in 1960s.

PGNT is a $4.3m microcap. It provides conveyor systems for assembly lines and order fulfillment operations. It lost money 7 out of 10 years between 2001-2010. Current share price is ~$2.8 while it has a NCAV of $3.25 which consists mainly of cash.

Normally I would quickly dismiss companies without a track record of making money. However, I noticed Sham Gad, a value investor, was involved. I took a closer look.

Downside protection

Gad was elected to the board in 2010. He subsequently built up his position to 25% throughout 2011. In March this year, he didn't only take over the chairmanship, he also got the 2 directors elected. These are the 2 directors that he originally recommended in the proxy fight back in 2010. So, effectively, Gad has the complete control of the company.

Why is this important? This is important because it puts a very solid floor on our downside. Gad intends to bring the business back to profitability. After he gained a seat at the board, he managed to reduce cost and steer the business to break even in 2011. What will happen if he senses he can't achieve it? As a value investor, he won't have any emotional baggage. He will immediately liquidate the business. The liquidation process may not be smooth. But since the bulk of the assets is in cash, we should get back most of the $2.8 invested. The presence of a value investor collapses the range of possible outcomes to almost a single point if the business fails to deliver. This wouldn't be the case if it were the founder controlling the company.

Valuation

And what is our upside? We should consider this a turnaround and handicap it. I take a stab at this in the spreadsheet below:

(If your rss reader doesn't show the spreadsheet, you need to visit my blog directly.) 





Under the "turned around" scenario, I assume it requires a 1.0x quick ratio to keep the business running. Hence $2.75 cash can be distributed. I assume it can double its revenue, back to the level before GFC. I assume it can earn a generic 5% net margin. And I give it a conservative 8.5 P/E multiple. This gives us a valuation of $7.13 per share.

Next, we need to guess how likely the business can turnaround. Again, this is a wild guess and a pretty aggressive one. But one thing that helps is the recovery of the US economy is on our side. I put down a one-fifth chance. The rest of the calculation in the table is self-explanatory. We end up with an expected return of 42%. You can try to plug in different numbers in different places. But the general risk/reward profile doesn't change much.

(Another thing to be aware of is, the 42% return or the $3.99 value won't exist in the real world. We will either end up with one of the possible scenarios. There is nothing in between. The expected value is only indicative.)

Capital allocation

Why did I make reference to BRK at the start? This has to do with how Gad intends to use the cash in PGNT. I don't believe Gad will actually distribute the cash if the turnaround fails. Gad has lay down his intent in his chairman letter published in March:

Through a disciplined capital allocation process, we will examine ways to utilize the Company's assets to increase the intrinsic value of the Company.

This is how I see it. He will try to keep the business breakeven and plow any operational cashflow back into the business (e.g. in R&D) while waiting for recovery of revenue. At the same time, he will invest the cash pile in any opportunities he can find. If you are familiar with the history of BRK, this is effective what Buffett did to BRK.

Gad is a Buffett disciple. Investing in PGNT will feel like investing in BRK in its old days. You need to be comfortable to be Gad's junior partner to invest in PGNT.

Final thoughts


So, we can look at the investment case this way: At $2.8, we are basically taking a stake in Gad's managed fund and at the same time getting a free option on the PGNT's business.


The asymmetric risk/reward profile here is a classic "tail I win, head I don't lose much" case. This is "high uncertainty, but low risk". I think the market misprices it because everyone focuses on the middle scenario. I imagine many value investors don't dare to dream wildly on the turnaround possibility because this is not usually how one will reason a net-net.

p.s. I have no position because my capital is deployed and locked up in other places.

(Disclosure: No position)

SODI shareholders

If you are a SODI shareholder, please look at this post at Oddball stocks.

(Disclosure: Long SODI)

Research in Motion is absurdly cheap (RIMM)

(Ed: I'm flattered when someone plagiarised this post word for word after a full month I wrote it and published it on gurufocus. --18 May 12)

I don't think I need to introduce Research in Motion (RIMM), the Canadian maker of the Blackberry smartphones. RIMM is currently trading at around $13 with a total market cap of $6.8 billion.

A quick run down of RIMM's business: RIMM makes Blackberry smaretphones. Blackberry phones offer two distinct sets of features: (1) Blackberry Messaging (BBM) - This is basically instant messaging (IM) on  phones. (2) Blackberry Enterprise Server (BES) - This is a corporate solution which provides remote phone administration and secure email connectivity ("push" service). Both BBM and BES relies on RIMM's global network infrastructure. Both BBM and BES are valuable franchises. BBM exhibits strong network effect. BES is sticky in corporate settings. If we dissect RIMM's market along these two technologies, we can see that RIMM serves 2 distinct groups of clients: (1) youngsters who use BBM and (2) large corporations which use BES.

The problems RIMM facing are well publicized. It's been losing market shares to both iPhones and Android phones. New phone models have been delayed because of delay in getting 4G LTE chips from Qualcomm. And its tablet offer Playbook has been a flop. Future prospect looks grim.

I initially looked at RIMM half a year ago. At the time, I didn't know how to establish a valuation with conviction as its cash flow was deteriorating fast. Since then, RIMM has lost 50% of its market value. What got me interested again was the news that Fairfax's Prem Watsa had doubled down on RIMM (now owning ~5%) and acquired a seat on the board, and Greenlight Capital's David Einhorn has bought a small stake in RIMM instead of shorting it.

I'm amazed by what I found. It's possible RIMM is a sinking ship, but it's a sinking ship stacked with gold bars on the deck in plain sight.

Fire Sale Value

The golden rule of investing is "don't lose money". What is the worse case scenario here? What I'm interested in is how much RIMM is worth if it is dissolved today and sells off its assets and businesses.

RIMM currently has a book value of $19 per share and a net tangible asset (NTA) value of $12 per share. i.e. RIMM is trading at 30% discount to its BV and very close to its NTA value. However, both of them may not be good measurements of RIMM's net worth. It is because RIMM is a technology company. Hard assets may not reflect its true value. (What use is a manufacturing plant which makes Palm PDA today?)

Here is how I calculate its liquidation value. Actually I've cheated. You will see why later. (All figures in million $USD except per share figures.)

 

Current assets (consists of mainly cash and account receivables), long term investments and total liabilities are straight off its balance sheet.You may argue that inventories have to be marked down significantly. But on the flipside, I mark PP&E down to zero. We only need an appoximation here.

Patents are valuable these days as strategic assets to technology companies for both offense and defense purposes. RIMM has ~2500 patents filed in United States. How to value them? Reports from analysts value them as low as $1 billion to as high as 10 billion, with a valuation of $2.5 billion widely quoted. There are also a number of recent comparable sales: Microsoft's purchase of Novell patents values the patents at $510k/patent. Google's purchase of Motorola Mobility, assuming patents are the primary assets, values them at $310k/patent (after backing out the $3 billion cash and $1.7 billion net operating loss tax benefits). Microsoft's more recent purchase of AOL patents value them at $1,320k/patent.

There are big price variations. How can we sure they were not overpaid out of panic? Besides, the quality of the patents counts. How can we be sure RIMM's patents are of similar quality? I've randomly sampled some of RIMM's patents and Motorola's patents on USPTO. I couldn't come up with a conclusive answer.

But RIMM was part of the consortium which purchased Nortel's patents for $4.5 billion in mid-2011. RIMM's share is ~$750m. This is a more reliable figure. If RIMM needs to sell this block of patents, it will be easier for RIMM to convince the buyer the other four guys in the room also thought this was a fair price. More importantly, these patents have been (hopefully) evaluated by experts in the consortium. This gives us more confidence in the quality. By entering only $750m in the above spreadsheet, I have effectively mark RIMM's own two thousand odd patents down to zero!

Next, the BBM network. This consists of a user base and the underlying technologies and network infrastructure. The user base is still growing. Number of subscriptions grown from $50 million to $75 million last year. My first thought was this network was comparable to the Skype network. Microsoft paid $8 billion for Skype. Skype had about 170 million active users. BBM has about 75 million subscribers. Let's say BBM is only half as valuable as Skype on a per user basis. Then, BBM will be worth $2b. But this is very imprecise. (Half? Why half, not a quarter? And didn't people say Microsoft overpaid?)

Can we do better? BMM generates cash flow. Users pay a monthly subscription fee to use BBM. BBM provides annuity kind of incomes. I read elsewhere RIMM gets $2-4 per month per user. (I can't verify this as RIMM does not disclose its revenue details down to this level. One thing we know though is RIMM's service revenue grew by $1b last year. With a 25m user base growth, it works out to be about $6 per month per user. But I suspect this includes also BES revenue.) Let's say RIM is making $1 profit per user per month and the user base doesn't grow anymore. RIMM's gross margin on services is 85%. So this is implying a 25-50% net margin which is high but believable. Let's also assume RIMM can milk it for 3 years. With a 15% discount rate, I also get a $2 billion figure. [Ed: The 15% takes care of both time value of money and decline of subscriptions. A sign of laziness on my part...]

Now, if we simply stop here and add up what we've got so far, we have $12.76 per share liquidation value, effectively the current stock price. Not only we have marked down its handset business, which has been very profitable, to merely its inventory value on the book, we have also excluded two very valuable assets in this valuation: (1) RIMM's own patent portfolio and (2) RIMM's BES franchise.  And this is where I've cheated. I don't attempt to value them. And we don't have to if we are interested in finding out our downside. We have built up layers of margin of safety at every step. RIMM's true liquidation value should be higher than this. It's hard to see we won't get our money back in case RIMM fails as a company. (Did I mention RIMM has another 2000 odd pending patents?)

Upside

Stop for a moment and consider RIMM's business in the last 10 years.  It's been extremely profitable. Even the much hated 2011 result which included some material non-cash write downs (goodwill and inventories) still commanded a 12% ROE, without using any debt. ROE was 20-40% in the previous years. BBM user base is still growing. Handset sales is still going strong in emerging markets.

If we ignore the current noises, I think it's absurd RIMM is trading at the current price level. I don't know exactly what the upside is. And we don't have to know. When we watch our downside, the upside will take care of itself. But if you really want to push to speculate the possibilities, I would say, if RIMM can just maintain the current profit level, its stock price can easily double at a P/E of 12x. But we are looking at a trough year. If RIMM can fix its problems and improve its profitability, it will be a multiple-bagger.

Challenges and opportunities

It's understandable the market generally fears RIMM will fade in obscurity very quickly like Palm did a few years back. Besides, production issues have forced RIMM to delay the release of its new generation of phones powered by the brand new QNX OS. This damaged RIMM's its revenue and gave its competitors a window of opportunity to steal its loyal customers.

How likely will RIMM fade in obscuirty?

I think the comparison to Palm is flawed. RIMM has 2 franchises, BES and BBM, with lock-in power that Palm didn't have. They may not stop from RIMM fading into obscurity if the management doesn't do anything. But they can slow down the deterioration and give the management enough breathing space to fix their problems.

Besides, these franchise are immensely profitable. While service sales makes up only 24% of the revenue, it makes up 57% of the profit. Since service revenue commands way higher gross margin (85%) than hardware sales (20%), the significant drop in sales figures is masking the profit growth in the BES and BBM.

   

RIMM should give these 2 franchises 120% of its attention.

BES is a corporate IT solution. It doesn't have to be coupled to Blackberry phones. In actual fact, RIMM has introduced new technologies last year (called "Fusion") to make BES work with other phones. RIMM is a trusted brand and the biggest player in this space. If RIMM is to sell off this business, BES will be very valuable to people like Microsoft, Oracle and IBM. Not only the revenue stream, but also the client relationship.

While BBM is an instant messaging service, it shares the characters of a social network. BBM is popular in pockets of population. This shows its network effort among small social circles. It will be a strategic fit for any existing social networks: Google+, Facebook and LinkedIn. 

We may not be able to decouple it from Blackberry phones (i.e. offer it on other phones) without damaging both the phone brand and the BBM brand. At one point I thought RIMM may simply ditch its handset business and focus on its services, moving the BBM and BSE franchises onto other phone platforms. But it appears RIMM does have pockets of loyal customers. Its phones are actually selling well in Latin America and Asian countries. The falling sales data in United States and worrying trend in Europe are masking its growth in Latin America and Asian countries. 

   

The battle RIMM is facing at the moment is a battle on OS platforms, a battle on the ecosystems. With two ecosystems (iOS and Android) dominating the market, it's tremendously difficult to maintain a third one. We learned in the desktop OS battle that whoever commanded the biggest pool of the applications commanded the market.

I believe RIMM can actually sidestep this battle and the management can then give its full attention to strengthen its franchises. What they have to do is to make Android apps run natively on their phones. Technically, this may require re-implementing QNX to run on the top of Android. This may not be the only solution to its ecosystem problem. But this is the simplest. This is effectively what Amazon has done. Nokia is arguably going down a similar route too, with the exception that it's selected Windows instead of Android. (Whether it's a wise choice will be a topic of another post.)

But in reality RIMM is doing it the other way round. RIMM is making Android apps run on the top of its OS. While this is not my preferred choice, this may work too. And I won't fault their thinking. This will probably be too late to change course without causing damages both internally and externally. How well it will pan out will depend a lot on how trouble-free it is for both app developers and consumers to bring Android apps onto Blackberry devices. (At the moment, sadly, it's not.)

Catalysts

Co-founders Mike Lazaridis and Jim Balsillie resigned from their posts as co-chairmen of the board and co-Chief Executive Officers in January. This removed one obstacle for RIMM to break away from its engineering root and explore different strategic directions. (Don't get me wrong. It was no small task to grow RIMM into the current form. Both Jazaridis and Balsillie have done a very respectable job. It's just that being the founders makes it hard for them to decouple their affection from cool-headed decision-makings when the competition landscape has shifted.)  While the new CEO Thorsten Heins is largely untested, his promise to explore all strategic directions is encouraging.

We need to have the faith that the management won't destroy the value by burning its assets. Its history doesn't stack up well here. But it offers a bit of comfort that RIMM has started cost cutting initiatives.

Time is of essence here. While value in RIMM's patent portfolio and the value in BBM and BSE franchises won't evaporate overnight, they will deteriorate over time quickly. If RIMM can't fix its problems in a year, our investment thesis will quickly reduce to simply the liquidation scenario. It is good to see they now have a sense of urgency. We also have activists like Prem Watsa on the board (and on the strategy committee) to keep things in check. [Ed: I should've emphasized, the presence of Watsa is pivotal in the entire investment case. Without a strong capital allocator on the board, the company can forever lose its path.]

High uncertainty, low risk

This can be a bumpy ride. Profits can fall further in the next few quarters if hardware sales don't pick up. Stock price can fall further in tandem. Visibility is low and no one can predict what will happen to RIMM. Will it be broken up like Palm was? Will it sell off pieces of its businesses or assets? Will the Canadian government veto an acquisition by a foreign firm on national security ground? Will it form strategic partnerships with other big boys? Will it stay in one piece?

No one knows.

But the risk of permanent impairment to our capital is low because we can establish a pretty solid floor on the valuation. On the flip side, the stock price will pop if anything good happens to RIMM.

This is a classic "head I win, tail I don't lose much" risk/reward profile.

The market sentiment is at its maximum pessimism. Short ratio is at its highest. One really needs the nerve to believe one's facts are right and one's reasonings are right.

(Disclosure: Long RIMM, ORCL & MSFT)

GLG Corp, a case study of (not) doing proper due diligence

GLE (ASX)

GLG Corp (GLE) is a Singapore-based company listed on ASX in Australia which provides apparel/knitwear supply chain management services. GLE's major customers are clothes retailers in the United States. It acts as a middleman between the retailers and the clothes manufacturers in China and other Asian countries. GLE is recognized in the industry locally as an established player. It's a micro-cap with a market cap of $18.50m.

GLE's share price has been hovering around $0.25 for quite some time. I originally looked at this company in mid-2011. Its value has improved substantially since then. NACV is now $0.34 per share and with the improvement in US economy its operation risks have subsided significantly. GLE has never lost any money since it was listed in late 2005. Both its margins and ROE before GFC look good. Average earning in the last 6 years comes to $0.085 per share. With a conservative 6.5x multiple, it will be worth $0.55. (GLE reported earnings in USD. But since exchange rate is close to 1, the difference isn't material in this discussion.)

So, we are looking at a 50%-100% upside. What did I do? I quickly bought a stake, of course. But the fun starts now.

Because the value looks so good, I actually wanted to buy more. But before committing more money, I decided to do more due diligence. There were 2 things in its financial statements that I initially grossed over. First is an "Amounts advanced to other parties" appeared in the financing activities section of the cash flow statements in 2012H1, 2011 and 2010. What the heck are they? I couldn't reconcile them to the balance sheets.

The other one is how GLE accounted for its trade receivables. It disclosed in the Notes that it did some kind of "offsetting" which seems to be related to how GLE accounted for its trust receipts. I re-read those few paragraphs a few times but was still not sure how the offsets worked. Besides, an entity called GLIT was mentioned here. GLIT was the spin-off from GLE when it was initially listed. It is a clothes manufacturer. In other words, it's GLE's supplier. It's actually GLE's main supplier. I'm not an expert in trade financing. So if I draw the wrong conclusion, someone please correct me. But why did a supplier have anything to do with receivables? On the top of that, GLE has also provided some $16m loan to GLIT since 2010.

With some google searches and through some Singaporean contacts, I tracked down 2 other public companies operating in the same industry in Singapore with comparable size: Ocean Sky and FJ Benjamin. Comparing their balance sheets to GLE's, I noticed a few things straight away. GLE holds far less cash, inventories and account payables than its competitors. GLE essentially has a very different capital structure than its fellow competitors. How so? Also, I didn't see the kind of trust receipt offsetting that GLE used.

If GLE is not a outright fraud, the only explanation I can think of is GLE doesn't do its own manufacturing while the other two competitors do. GLE outsources its manufacturing to GLIT. But, is it truly outsourcing?

By piecing together all these observations, I come up with a theory: Legally GLIT is an independent company. But it isn't, both commercially and financially. It's still part of GLE. As one can imagine, their operation is more capital intensive and they has probably lost money in the last few years. GLE has been shuffling money down the pipe to keep GLIT alive. Beyond the $16m loan shown up in the balance sheets, I guess GLE swept the transaction details all under those "trade receivables". Beyond that, nothing else about GLIT appears in GLE's book. The operation is basically off balance sheet. How profitability is the combined entity, GLE and GLIT together? What is the overall ROE? No one knows.

These are not facts. It's a guess.

But I was uncomfortable enough that I got rid of my stake at a small lost.

p.s. I was fully aware of Steve Johnson's post about a mistake in their financial statements. That alone didn't deter me from buying GLE. Stupidity? Greed? Maybe. But now with other supporting evidence, it fits the theory. It also fits the theory why GLE used a big name accounting firm in a small town.

(Disclosure: No Position... now)

Lesson learnt from my mistake with Lakeland

LAKE (NASDAQ)

Not buying LAKE was one of my biggest mistakes in 2011. This has led me to rethink how to judge risk/reward balance in net-net investments.

Lakeland (LAKE) is a protective clothing manufacturer. It's a microcap with a current market cap of  $54m. I looked at on and off for nine months with a passing interest. In November last year, it traded below $7.00, that was 20% below its net current asset value (NCAV) and 50% below its book value. That got me very interested and I took closer look. Whopper Investments has a nice writeup of the investment thesis on his blog. I'm not going to repeat the analysis here.

To cut the story short, I was troubled by their Brazil VAT liabilities. I couldn't reconcile the figures disclosed in the cashflow statements with the details else where in the 10-K. Together with a few circumstantial facts*, I started wondering if there was fraud at LAKE. After some email exchanges with Whopper and some more thoughts, I dismissed the fraud idea because there wasn't much incentive for the management to do so. But I did conclude their VAT mess (resulted from an acquisition) was a result of bad management. I concluded their incompetency would destroy value and their good ROE in past years was pure luck. Later in their fourth quarter result, they shuffled the India money losing operation into "discontined operation". That further enforced my thinking. With no catalyst in sight, I was worried that they would bleed money in foreseeable future.

Fast forward to today. LAKE is now trading at $10.50. That's 40% return in less than 3 months. What has happened? In December before Christmas, Ansell, an Australian protective clothing company, took a 9.7% stake in LAKE.

What's gone wrong in my reasoning?

I always wanted an exit strategy or catalyst in place but there wasn't one. I forgot a net-net is a net-net because it has a few warts and no obvious resolution in sight. Otherwise it won't be a net-net. Buying a net-net is basically a calculated bet on some "positive black-swan" event, if you wish, that some positive event you have no way to anticipate or foresee will happen. At the same time, the backing of the assets gives you the staying power and downside protection.

p.s. If you don't know what a black swan is in the context of investing, read Nassim Taleb's Black Swan or Fooled by Randomless.

----

*  Other circumstantial facts I found: (1) the proxy-advisory firm Institutional Shareholder Services (ISS) has advised sharesholders to vote against director John Kreft (sitting on the audit committee) and Lakeland's audit WAKM in its Jun 2011 AGM, objecting the high non-auditing fee paid to WAKM. Kreft nearly lost his seat.  (2) Lakeland only switched to WAKM it the last couple of years. Switching accounting firm always raises concern. (3) WAKM was being sued for negligence in auditing of a bankrupt furniture maker.