Chinese Reverse Mergers: An Objective Examination

by: Ryuto A. Kawai

Part 1: Introduction:

About a month ago, Rames El Desouki posted a rebuttal to a controversial Barron's article, "Beware This Chinese Export", written by Bill Alpert and Leslie P. Norton, an article that both warned of the risks of investing in Chinese revere merger stocks, but also indicated significant underperformance for a select 10 such stocks.

Mr. Rames El Desouki rightly pointed out that analyzing stock performance with the starting point of merger completion may not be relevant for many investors, given that trading in the stock immediately following a reverse merger is typically amongst insiders. While looking at the data post-merger is most appropriate for those willing to buy in on those first weeks of trading, most investors understand the inherent risks involved in buying a stock with little price history.

Thus Mr. El Desouki took the first 10,000 trade volume day as the starting point, and selected the last pre-Barron-article day of 8/27/2010 as the ending point. The rationale for the end date was that the Barron’s article had a large negative impact. Using these new parameters, he pointed out that the 10 stocks in the original Barron's article, implied to have management or ethical issues, actually OUTPERFORMED rather than UNDERPERFORMED Chinese stocks as a whole, using the Shanghai Composite or Halter Index as the benchmark.

His piece was certainly a contribution to the discussion, and a balancing article with contrasting conclusions from the original Barron's article. Here is a relevant table from that article:

(Content from:

As can be seen, if one agrees with the methodology, the conclusion seems to be that these 10 stocks have been stellar investments. As I have followed many Chinese reverse mergers and ADRs, I have come to know that Chinese ADRs, and especially Chinese reverse merger stocks, require a different type of due diligence beyond just financial statement analysis and market analysis. Namely, I am talking about the sort of due diligence that tries to identify risk factors correlated to accounting fraud and earning restatements, such as examining the history of the company, its auditor(s), its product line and marketing, and the affiliation and histories of major investors, executives, banks, and promoters involved with the Chinese company in question. As pointed out on numerous occasions here on SA, comparing Chinese SAIC figures to US SEC figures can also provide clues.

Though I applaud Mr. El Desouki in pointing out the misleading nature of taking the merger date as the starting point, I also think the Barron's original article did have the right tone of warning individual investors and professionals alike who may only use more traditional analysis, and may not be aware of how common earnings restatements, auditor changes, or other outlier-loss events are in this space. In my current article, I would like to present an objective analysis that neither completely confirms the Barron article implication that these 10 stocks, and more generally the Chinese reverse merger space stocks, are not worth investing in, nor completely confirms Mr. El Desouki's implication that these 10 stocks examined by Barron's have been, on average, decent if not great investments. Both Barron's and Mr. El Desouki are correct -- Barron's in pointing out significant risk, and Mr. El Desouki in pointing out significant gains. My conclusion, as will be seen, is that investing in this space should not be seen as too risky and unworthy of consideration nor broadly lucrative. The reverse merger Chinese ADR space is both significantly risky and significantly lucrative. The characteristics of return distribution also suggest a few strategies, which I will touch upon in the final remarks section.

Part 2: Summary of Analysis, Parameters, and Certain Assumptions:

The framework for my analysis rests on the following parameters and assumptions:

1.) The list of stocks analyzed are those cited by the Barron's article and rebuttal: OTCPK:CAGC, OTCPK:CBEH, CGA, OTC:CHNG, DEER, OTC:FEED, GFRE, ONP, OTC:RINO, and SPU. Nothing has changed here.

2.) In selecting the time period for performance analysis, rather than selecting the passing of a volume threshold, I have taken the 5-day moving average of volume, and an arbitrary but more reasonable 100K volume MA threshold as the starting point. Most investors need some consistent level of liquidity in order to avoid high bid-ask spreads. Larger institutional investors also need a relatively higher level of trade volume to avoid running the stock up in accumulation. Thus, taking the first 10K volume day as the starting point for stock performance analysis, may not be appropriate.

3.) Rather than selecting the last date for analyzing stock performance as the last trading day before the Barron's article, I have decided to present the analyses of the 10 stocks for two time periods, side-by-side: a.) Through to a more current 10/1/2010, the date of data download, and b.) The same analysis through 8/27/2010, before the Barron's article publication. Mr. El Desouki took the Barron's article date as the cutoff, which I believe skewed results favorably.

I completely agree that the Barron's article did in fact have an impact on share prices in the Chinese reverse merger space. However, quantifying such an effect precisely is difficult, and we cannot assume any level of impact, high or low. In the case of SPU, for example, that company announced $25.9M in additional shares, counter to what they announced before in their second quarter filing. This development and other significant developments make measuring the negative impact of the Barron’s article difficult, if not impossible.

I believe in light of these difficulties, the only objective way to present the data is to present both, through to current, and through until right before the Barron's publication, which is how I presented my data.

4.) I have introduced some risk-metrics into my analysis. While some of the tables in the rebuttal showed spectacular returns, we don't know if this was a hot period, and what the risk cost was.

The first, and most simple addition was the Sharpe's ratio. Due to space constraints, I will refer unfamiliar readers to a google search or investopedia for the definition of the Sharpe's ratio.

The two key concepts in the ratio are, a.) That performance is measured against the risk-free rate of return, which I took from 1yr Treasury yields on 9/17/2010, and b.) That this 'excess return' is divided by the standard deviation of returns, in order to scale the returns for risk.

In my analysis, I tried binning returns into 1-week, and 1-month periods but nonetheless could not eliminate additional statistical modes (local maxima) and high kurtosis or the so-called 'fat-tailed' distribution, and thus I stuck to the simplest methodology of utilizing daily returns. Note here that this ratio can understate risk when there is high kurtosis and/or skew of returns, which was observed in this sample of 10 stocks.

5.) To compensate for risk understatement in the Sharpe ratio, I have also calculated the maximum drawdown, which informally, is the largest loss in a 'trend'. I simply took the largest loss that is unbroken by a consecutive 4 days of 0%+ returns. Where a downtrend is broken by 4-consecutive 0%+ return days, that was taken as the end of that downtrend. The maximum drawdown was simply taken as the largest loss of all these individual downtrends.

6.) To benchmark performance against a readily accessible Chinese equity fund, I have also calculated information ratio ratios to two separate Chinese funds.

a.) PGJ as the benchmark. The PowerShares Golden Dragon Halter USX China ETF is an exchange-traded fund based on the NYSE Arca Halter USX China Index® which has the stock-selection criteria of a $50M capitalization minimum over a minimum of 40-days, NYSE or Nasdaq listing status, and approval by a commission.

b.) FXI as the benchmark. The iShares FTSE Xinhua China 25 Index Fund is based on The FTSE Xinhua China 25 Index®, an underlying index of 25 of the largest and most liquid Chinese companies. Compared to the Halter USX China Index®, this index tends to select larger capitalization companies.

7.) Because both Sharpe ratios and Information Ratios are inadequate in incorporating the data contained in the maximum drawdown figure, I have calculated a custom figure that scales annualized return (and not excess returns over the risk-free rate or some benchmark) by the maximum drawdown. I utilized the following formula:

MDD-adjusted return = { Annulized Gain % X (1+Maximum drawdown), if [Annualized Gain % X (1+Maximum drawdown)] > 0

MDD-adjusted return = { 0. if [Annulized Gain % X (1+Maximum drawdown)] =< 0

Only positive values are scaled down. All negative values are equated to 0 as a minimum value for the custom risk-adjusted return figure.

8.) Finally, I have constructed a virtual portfolio of $100,000 allocating $10,000 each into each of CAGC, CBEH, CGA, CHNG, DEER, FEED, GFRE, ONP, RINO, SPU, to gauge its performance, including absolute gain/loss, maximum drawdown, Sharpe ratio, and information ratios. I have taken the latest starting point of adequate liquidity as determined by the 100K 5-day moving average volume threshold for the entire virtual portfolio.

This analysis suffers from the short time period due to the latest starting point coming from SPU stock which achieved the criteria-level of liquidity on 12/15/2010. However, the exercise is nonetheless instructive.

Part 3: Data:

I will let the data speak for itself. The first table shows the absolute gain %, annualized gain %, the Sharpe ratio, and the Information ratios against the 2 benchmarks. The maximum/attractive value is highlighted in green, whereas the minimum/unattractive value is highlighted in red. The first table is data through 8/27/2010.

A. Risk Adjusted Return and Performance Figures, Individual Stocks

click to enlarge

Performance Table 3.1: Performance through 8/27/2010

The second table is the same analysis through 10/1/2010:

Table 3.2: Performance through 9/17/2010

Next, the drawdown figures, and the custom MDD-adjusted return table:

Table 3.3: Maximum drawdown and MDD-adjusted return, custom ratio.

B. Virtual $100,000 Portfolio Construction and Performance:

In the construction of the virtual $100,000 portfolio of equal weight in these 10 stocks, I have rounded down share size to the next whole number down. Keep in mind, that in real investing there are expected additional costs of commission, slippage, and the bid-ask spread.

Table 3.4: Initial portfolio, starting date 12/15/2009

The performance and risk-adjusted metrics of this virtual portfolio for two different ending periods are tabulated below:

Virtual $100K, end 8/27/2010

Virtual $100K, end 10/1/2010

Starting Date



Starting Value



Final Value



Gain/Loss %



Annualized %



Sharpe Ratio



IR vs. PGJ



IR vs. FXI



Maximum Drawdown



Table 3.5: Virtual portfolio performance.

I want to make clear in the presentation of this data that the time periods were not selected with any bias. However, the time period is too short to make any generalization over a broader time-frame. The information only serves a purpose of presenting one possible scenario for a specific time period, if one were to invest in these 10 stocks equally starting from a point when all the stocks have passed the liquidity/trade-volume thresholds.

Part 4: Conclusion:

From the data presented in Part 3, I believe the case is clear that there is significant opportunity for gains as well as a significant potential for loss. While the historical annualized gains of GFRE or CAGC may seem attractive on the surface, the maximum drawdown for CAGC was calculated as a -66.5% loss, and for GFRE a whopping -93.5% loss, with a maximum total loss without leverage obviously being -100%. In a very broad overview, both the idea presented by the Barron's article that this investing space can be extremely risky, and the idea presented by Mr. Rames El Desouki that these stocks have massive upside potential are indeed correct.

Reflecting back on historical performance, those performances (since the point of adequate liquidity) of CAGC, CGA, and SPU seem to be 3 of the most attractive. CGA and CAGC both had equivalent maximum-drawdown-adjusted returns of 0.60. CAGC produced the larger annualized return through 10/1/2010 of 264.67% with a massive -66.5% maximum drawdown. CGA was the relatively lower risk stock compared to CAGC, with an annualized return for the same period of 86.95% with a much more reasonable maximum drawdown of -31.2%.

Incidentally, CGA had the highest Sharpe ratio for that period, though, the Sharpe ratio tends to understate risk where the return distribution graph is not normal/Gaussian. SPU had the lowest maximum drawdown of -21.9%, and also had the most attractive information ratio by a good margin (around 1) for all periods and against both benchmarks. The annualized return through 10/1/2010 for SPU was 50.27%. It goes without saying that past performance may or may not indicate some level of future performance.

Other notable stocks include CHNG, ONP, and FEED which were all historically unattractive investments producing negative returns. ONP, the Muddy Water's short-sale target was the worst performer on an annualized basis, with an annualized loss of -55.26% through 8/27 and -52.1% through 10/1. Interestingly, the stock was relatively flat post-Barron's-report. GFRE also had the largest maximum drawdown of -93.5%. Referring back to the original Barron's article cited in the introduction, it may be noteworthy that both ONP and GFRE were associated with Mr. Kit Tsui, which may say something about the quality or ethicality of the companies he is involved with, or alternately, the short-selling attention his name may be bringing to these stocks or both.

The largest absolute losses for both periods go to FEED (-80.51% thru 8/27 and -78.52% thru 10/1). Contrary to the assumption that the Barron's article had a negative effect on all stocks after its publication, FEED has also very slightly trimmed its losses since the Barron piece publication. Notably, AgFeed also has the 2nd largest maximum drawdown at a loss of -77.6%. Referring back to the original Barron's article, FEED is a stock affiliated with Mr. Benjamin Wei/Wey who many investors will remember as having regulatory issues with NASD and the State of Oklahoma. Many associate Mr. Wei/Wey with the stock Bodisen Biotech, which produced significant losses for investors when AMEX de-listed the company with actions starting late 2006, and ending early 2007.

Finally, the construction of the virtual $100K portfolio clearly indicates that at least for the last 8-9 months, these 10 stocks collectively were a poor investment. An initial investment of $99,935.57 would have been reduced to $79,387.15 if the portfolio were liquidated before the Barron's article for a -29.27% annualized loss, or reduced to $67,690.51 if the portfolio were liquidated on the close of 10/1/2010 for a -33.96% annualized loss. The maximum drawdown for the portfolio was -35.49% for both periods.

In conclusion, the reverse merger space is fraught with risks for both longs and shorts, as evidenced by large losses as well as massive gains. In a following piece, I will detail how the more risk-tolerant investor can exploit such stocks reasonably to add gains to their portfolio.

Disclosure: No positions

See How to Invest in Chinese Reverse Mergers- Part II