Description

OVTI ($17.57) OmniVision Technologies, Inc. (common stock)

Have you ever wanted to buy a company that meets Berkshire Hathaway’s acquisition criteria?  Here is a chance to do so at a basement-bargain price.  (Note: I posted this idea about a year ago.  As I intend to focus on recent developments, my previous write-up could be helpful in understanding the business in more detail).

Here are the criteria (from BRK’s 2005 annual report):

1) Large (>$75 million in pre-tax earnings)

2) Consistent earnings power

3) Good ROE with little or no debt

4) Management in place

5) Simple business

6) An offering price

Yes, OVTI meets all of them:

#6) It is offered at a basement-bargain price.  As of April 30, 2006 (the end of its last fiscal year), OVTI has $6.2 of net cash per diluted share.  In FY2006 it generated $1.9 of free cash flow per diluted share.  Thus, the current offering price adds to 6.0x EV/FCF, or a 16.7% FCF yield!

#5) While it involves technology, it is a simple business.  OVTI makes digital (CMOS) sensors that can be used in a variety of applications including camera-cell phones and digital cameras.  Sensors are sold to manufacturers (either directly or through distributors).  Furthermore, OVTI does not own its factories as it out-sources production (i.e., very low capital intensity).

#4) OVTI has management in place.  In fact one of its founders is the CEO (and a major shareholder).

#3) OVTI generates low-twenties ROEs with no debt (<$1 million) and lots of (low-yield) cash.

#2 and #1) As can be seen in the table below, it has consistently increased its earnings:

2003  2004  2005  2006

Amounts in millions of US$

Revenue           109   318   388   492

Ebitda            21    89   105   117

Net income        15    59    76    89

Capex              7    10     1    23

Ebitda-capex      13    79    104  94

ROE               18%  30%  22%  22%

($/share)

FCF/diluted shr   (0.0) 0.5  1.2  1.9

FCF = Cash from operations – Cash from investing (excluding sales and purchases of marketable securities but including acquisitions and other uses of cash)

When I wrote about OVTI a year-ago, the investment thesis was centered on OVTI being able to generate $1.2 of FCF/shr.  As can be seen, OVTI exceeded this amount by almost 60%!  As investors realized the cash flow generating capacity of OVTI, its price increased to the low 30s.  Only at that point, its valuation became more reasonable (as opposed to way undervalued).  That is, assuming no growth in FCF/shr, an investor would obtain a reasonable return on her capital.  Now a year later, it is clear that there is growth in FCF/shr.

Why has the stock pulled back so dramatically over the last few weeks?  My view is two-fold.  First, in the market there seems to be a consensus that we are headed to an economic melt-down.  Second, OVTI, for non-fundamental reasons, is disliked by a number of public commentators and others.  Those able to do a news-run could confirm that for a long time many have declared it, essentially, out of business.  The bear arguments have been multiple.  For example, a year ago there were major concerns with inventory levels.  As I argued, this argument proved to be a non-issue as the inventory was adequate (once you looked at finished goods inventory vs total inventory).  Then, there was concern with competition, declining ASPs, commoditization, and production yields and gross margins.  While these are all key indicators of the company’s business, it seems to me that they have fluctuated within a reasonable range.

The market is ignoring the company’s leadership position in a growing market.  It is also ignoring the company’s execution capabilities.  The fact is that despite a fiercely competitive market, OVTI continues to deliver increasing amounts of FCF/share in addition to larger revenues and earnings.  The market is today as competitive as it was a year ago (and previously).  In this market, the company successfully competes against some of the world’s semiconductor powerhouses.  Declining ASPs are part of the nature of this market (as occurs in other similar markets, like NAND flash memory).  As there is an absence of comparable pure-play companies in the CMOS space, it might be worthwhile using NAND flash as a reference.

I believe the NAND flash market is a good reference point, as both markets exhibit similar dynamics.  Both are based on relatively old semiconductor technology that has revived to serve very large and growing markets.  Both have a fair amount of competitors serving a fair amount of major customers.  In both you see significant declines of ASPs.  Importantly, both are increasingly deployed to new mass-market applications.  CMOS sensors were initially used in digital cameras, just as NAND flash was initially a USB drive product/memory cards for digital cameras.  As CMOS found a huge market in cell phone cameras, NAND flash found one in portable music players (e.g. iPod nano).  Both technologies are replacing older and less suitable ones for the majority of the applications into which they are deployed.  In the case of CMOS sensors, they are replacing CCD sensors which have higher power requirements, are bulkier, and more expensive.  Similarly, NAND flash is replacing hard drives in the music player market for similar reasons.  Both businesses are seasonal, and can be described as commoditized or not (as there are arguments both ways).  More importantly, each of these markets has huge future potential, as there are several large mass-market applications that are just being explored.  For example, OVTI has recently announced design wins into automotive applications, deployment into toys and medical applications, and also sells into the security and surveillance market.

Despite these, and several other similarities, there is a huge valuation difference between OVTI and NAND flash companies like SNDK or FLSH.  SNDK owns some of its manufacturing facilities but also out-sources some of its production, while FLSH out-sources all of its production.  However, in terms of market share (i.e., size relative to that of the total market), OVTI looks a lot more like SNDK than FLSH.  Let’s look at a few valuation metrics (based on numbers for the last twelve months):

                  OVTI  SNDK  FLSH

EV/Ebit           6.2   11.8  18.3

EV/Ebitda         5.6   9.6   16.5

EV/Net income     7.3   17.7  16.8

EV/Ebitda-capex   7.0   12.9  18.7

Even if you disagree that these are good comparables, the absolute valuation of OVTI is inconsistent with the market’s cost of capital.  At a 17% FCF yield, it is ridiculously cheap.  Especially in light of its growth trajectory, its leadership in a growing market and its staying power among other things.  Furthermore, last year when the stock price was also at low levels the company repurchased $80 million worth of stock.  Needless to say, this is very accretive.

So why is it so cheap?  To me, it just looks like the result of Mr. Market changing his mood.  Clearly, the market for CMOS image sensors will continue growing (at a multiple of GDP) for the foreseeable future.  As new applications mature (e.g., automotive) and current applications continue to grow (e.g., camera cell phones), the unit grow continues to exceed the decrease in average price per unit by a large amount.  Secondly, OVTI is not going out of business anytime soon.  On the one hand, it has demonstrated a remarkable ability to maintain its technology leadership.  On the other, it has a ton of cash and minimal capex requirements.  Lastly, based on its track record, it is a lot more likely than not that it will continue to increase its revenues, earnings, and free cash flow per share.

Catalyst