I invested in Hollysys Automation ("HOLI") about 3 years ago because I believed that their positioning, business model and product portfolio would benefit greatly from China's need to increase its efficiency and productivity in the future particularly as labour becomes scarce and, consequently, more expensive in the future.
But, the more I follow this stock and analyze the company's results, the less I am convinced about my investment. Indeed, I have become concerned about the company's dependence on several elements to keep publishing EPS growth. These include the focus on reporting "non-GAAP" earnings, the growing importance of "VAT refunds and government subsidies" within the P&L, acquisition led growth, and the appreciating RMB/USD exchange rate.
I therefore suggest looking at these elements in more detail below. I would value any feedback enabling me to decide whether or not I should be holding on to the stock, or selling it at these levels (note that I am now, finally, making a small profit!).
1/ "non-GAAP" earnings
Like most US listed companies Hollysys tends to focus on "non-GAAP net income" when communicating on its financial performance. While I have no objections with company managers trying to isolate recurring income from the rest, I am always interested to know what the good-old-fashioned "Net income" based on the general accounting standards was for any given year.
Looking at the year ending June 2013 ("FY13") for example, I believe "non-GAAP net income" gives us a very flattering view of Hollysys' performance last year. Indeed, according to the company's filings, "non-GAAP income attributable to Hollysys" came in at US$57.6 million in FY13, presenting growth of 0.4% year-on-year. However, when looking at the P&L's "Net income attributable to Shareholders" (based on GAAP), the company only produced US$51.994 million in FY13, a 7.5% decline year-on-year. Going from GAAP to non-GAAP can therefore make it look like there was some growth while there actually wasn't.
So what did Hollysys exclude from its costs when it went from GAAP to non-GAAP in FY13? To begin there was the share-based compensation plan, which cost US$1.599 million in FY13. In my book (and in Hollysys' accounts) this is a recurring expense (it happens every year) and, therefore, should not be excluded. But, in addition, Hollysys also excluded a total of US$4.011 million of acquisition-related costs. For a company that has made it its strategy to grow in South East Asia through acquisitions, and that has completed two such acquisition in a little over 2 years, this also does not strike me as being absolutely "non-recurrent" in nature. So the "non-GAAP" adjustments of US$5.611 million made in FY13 do look to me like a way to flatter net income a bit.
2/ VAT refunds and government subsidies
Hollysys seems to be increasingly relying on government support to maintain margins and earnings. Although I prefer to invest in companies enjoying government support rather than in companies not enjoying government support, I still find this trend worrying because 1/ we don't know how long this will last and/or how to forecast it, and 2/ this distorts our understanding of a company's true commercial performance.
According to Hollysys' FY13 annual report (http://www.hollysys.com.sg/home/pubdown/Hollysys_Automation_Technologies_Ltd_20-F_FY2013.pdf), the total value of "VAT refunds and government subsidies" amounted to US$10.812 million in FY11, US$18.306 million in FY12 and US$22.982 million in FY13. While the company's revenues grew 32% over FY11 to FY13, the value of "VAT refunds and government subsidies" grew 112%!
Moreover, with Hollysys reporting "Income from operations" of US$44.691 million in FY11, US$65.438 million in FY12 and US$57.702 million in FY13, this implies that the share of "VAT refunds and government subsidies" has been growing tremendously as a percentage of profits. Indeed, based on the amounts above, it means that they represented 24% of "Income from operations" in FY11, 28% in FY12 and 39.8% in FY13!
In other words, it is quite clear that Hollysys' profits have become increasingly reliant on government policies and hand-outs in China. While I was not overly concerned about it in the past, I am now starting to question whether or not this will likely be a serious problem in the future because it represented nearly 40% of earnings last year. In my view this also sheds new light on the company's performance last year.
3/ acquisition led growth
Hollysys has announced 2 significant acquisitions in recent years, both of which were outside China. Concord, a Singapore-based business, was acquired in July 2011 (thus impacting the FY12 accounts) for a total consideration of US$42.910 million. And in April 2013, Hollysys acquired the Bond Group, a Malaysian-based business (which impacted accounts from the fourth quarter of FY13) for a total consideration of US$73.805 million.
The reasons why I want to focus on this are: 1/ by changing its growth model from an organic one to an acquisition based one, Hollysys is already seeing a deterioration in its profitability ratios - this is normal when you pay for goodwill over the price of net assets; 2/ it may imply that growth in its domestic market - China - has slowed a lot; and 3/ the impact this has on accounts, and how one views the company's growth.
Information relative to both acquisitions and their impacts on Hollysys' accounts can be found in Hollysys' FY13 annual report, available on the company's website and the SEC's database. Look up "NOTE 3 - Business Combinations", pages F-21 to F-24.
I want to focus on the Bond Group in particular because its consolidation in Hollysys' accounts will have a significant impact on the current year's (FY14) reporting. Indeed, according to page F-23 on Hollysys FY13 annual report, had the Bond Group been fully consolidated into Hollysys' accounts last year (what is usually called 'pro-forma accounts'), the company would have reported total revenues of US$392.481 million and a Net income of US$57.813 million. But, because the Bond Group was only consolidated in the fourth quarter last year, it only contributed an additional US$22.795 million in Revenues, and US$1.791 million in Net income (this also means that, without the Bond Group, Hollysys' Net income in FY13 actually dropped 10.7% versus FY12).
The pro-forma accounts for last year are of particular interest to me as they enable us to better understand the FY14 guidance management is giving today. Indeed, in its "Outlook for FY14" (,http://www.hollysys.com.sg/home/investor-relations/quarterly-financial-report/712-fy2014-q1-ended-on-sep-30-2013-nov-18-2013-hollysys-automation-technologies-reports-unaudited-financial-results-for-fiscal-year-2014-first-quarter-ended-on-september-30-2013-) management states that it expects "…revenue in the range of $460 million to $490 million and non-GAAP net income in the range of $65 million to $69 million". Considering Hollysys reported US$349 million in revenues and a Net income of US$52 million in FY13, this all looks like pretty strong growth.
But based on the pro-forma accounts, this implies Revenue growth in the range of 17.2% to 24.8% - this is strong, no question. But it also implies Net income growth in the range of 12.4% to 19.4%, i.e. roughly two-thirds of revenue growth. More importantly though, this guidance is based on "non-GAAP net income", which could mean Net income growth might come in lower.
4/ the appreciating RMB/USD exchange rate
I have analyzed several overseas-listed companies in my career and it is my experience that one must always convert their accounts in the currency in which they do business to get a good understanding of their true operating performance. Too often people focus on accounts published in a "reporting currency", which distorts the real performance of a company, either positively or negatively.
In the case of Hollysys, the company reports all of its accounts in US Dollars, but actually conducts the large majority of its business in China, in RMB. Therefore, with the RMB appreciating regularly by 2-4% versus the US Dollar since mid-2010, Hollysys' accounts have benefited each year from a favourable FX translation. While the gain may be small each year, possibly around +3% on average, this adds up after 3 or 4 years to well above 10%. So, it's best to always keep this in mind when looking at results…
In conclusion, although the analyst consensus seems to be pointing to a valuation of 14.5x FY14 EPS and 11.8x FY15 EPS at present, which I don't find overly demanding (but not cheap either), I am questioning the quality of the analysts' models, and the assumptions they have made when forecasting Hollysys' future earnings. How have they accounted for currencies? What did they assume would be the future level of "VAT refunds and government subsidies"? And, are they basing their EPS numbers on GAAP earnings or non-GAAP ones?