Description

Business model

COGO acts as an intermediary between semiconductor-component manufacturers and local Chinese equipment manufacturers. Another way to think of COGO is as an outsourced marketing team for large non-Chinese semiconductor companies (such as Broadcom, Matsushita, Scandisk, etc.) to reach equipment manufacturers in China. 80 blue chip customers compose 70% of COGO revenues, while 1,500 SMEs compose the remaining 30%. End applications for chips distributed by COGO include handsets, set-top boxes, telecom equipment, rail applications, as well as others.

COGO employs hundreds of engineers to design software modules for the imported chips. These modules are customized to the application required by the manufacturing customer. Thus, COGO uses industry relationships to identify customers, leverages sourcing scale to import chips economically, applies design expertise to build modules for the chips, and sells the chips to manufacturing partners. Margins for large, blue-chip customers are quite small, but COGO is able to leverage designs and software of large customers to resell similar applications to smaller customers; thus, blue chip customers are used in part to develop IP which is marketed to smaller manufacturers. 

This is not a generic import business, and COGO's return on tangible invested capital reflects their value added. On the other hand, it also isn't a very high-moat business. Return on tangible invested capital have declined from 24% in 2006 to 12% in 2010. The 12% ROIC is actually quite attractive given that the business trades for 46% of tangible book value.

Fraud risk

The bear-case on COGO is fraud risk, and COGO's stock price has deteriorated with general backlash against Chinese ADRs. Write-ups on SumZero on August 4th and June 30th argued that COGO is fraudulent. The August 4th piece attacked COGO's recent website aimed at marketing products to Chinese SMEs. According to the author, the new website was evidence of fraud; red flags included that the website seemed to be produced with a limited budget, did not allow for automatic purchasing, and quoted prices in U.S. dollars. However, the website was made active just one quarter ago, and is not a part of COGO's core business. The website is not evidence of fraud.

The June 30th piece also failed to present compelling evidence of fraud. The major issue raised was insider transactions, which are quite common in Aisan stocks (and even some U.S. stocks). On the other hand, some legitimate risk factors were presented, including changing auditors in 2007 (note that the new auditors, however are KPMG). Furthermore, COGO has a history of growing through M&A, and often these transactions commanded high multiples. However, this data, alone, does not present a compelling case for fraud. 

On the other hand, there are many factors that argue against fraud in the case of COGO:

1. Jefrey Kang, a respected member of the Chinese business community has been CEO and Chairman since 1999; Jeff Kang continues to own approximately 30% of the company; Kang's total assets are in the 9 figures, and he's not likely to risk jail and continue to operate a fraud
2. Top management has been consistent for years and has not turned over
3. The company is audited by KPMG
4. The company has been listed on the NASD since 2004, with no significant capital raise since 2007; if COGO were a fraud, why wouldn't it have folded by now, with its main purpose (raising capital) having been complete since 2007? Cogo could have folded during the 2009 financial crisis
5. The company has been buying back shares over the past 3 years ($33M in total). Why would Cogo buy back shares if it were a fraud?
6. The company has ~$400M in yearly revenue, 2/3s of which comes from large blue chip companies like ZTE. I spoke to someone through GLG who represented that he knew first hand of departments within ZTE that purchased from COGO. Moreover, falsifying $400M in revenue from large companies is not feasible given extensive analyst coverage
7. Realistic financials -- many companies that were demonstrated to be frauds had extremely high margins and ROICs; COGO, on the other hand, has had falling ROIC since 2006; during the downturn in '09, margins fell, and the stock price cratered; also, accounts receivable have been increasing; why permit mediocre economic performance if you are cooking the books anyway?

I think the most worrying thing about COGO is their increasing accounts receivable (and factoring of A/R at the same time) -- it is possible that assets and earnings are overstated due to low quality sales and understated allowance for doubtful accounts. However bad accounts receivable would not be responsible for eroding >50% of tangible book value, as implied by current market prices. 

Valuation

For an upside case, I think COGO can trade for tangible book, or 119% upside from $2.80. COGO could be worth even more, as it has proven that it can grow and out-earn its cost of capital since inception as a public company in 2005. As a very conservative base-case, I take a scenario where COGO's receivables and inventory are worth 50% of stated value, getting me to $3.53 per share, or 26% upside. The downside is that COGO goes to zero, which is certainly possible given the space and possibility of fraud. However I think upside is significantly more likely given the factors listed above.

Catalyst