The GSI Group is an under the radar opportunity with all the ingredients in place for a roughly 50-100% gain.  In addition to an undervalued business, there's:

-- a new chairman of the board who owns nearly 11% of the shares.  
-- high quality management (namely, the chairman) with a long track record of making shareholders wealthy in the opto-electronics industry.  
-- a tailwind of increasing demand ... good chance this is the beginning of a multi-year up-cycle for their most cyclical business.
-- some potential for low-hanging fruit (operational improvements) in a company that has historically been under-managed.
-- near term catalysts: filing of last seven quarters of financials by year end, followed by a reverse split and re-listing on an exchange.

(Note: valuation stats exclude final bankruptcy costs for 2010.)

Overview:
GSI's main business competes in the $2.5-3.5 billion "non-diode laser" market - not exactly the world's best known industry, so here's a quick overview.  These are lasers sold to other businesses that integrate them into systems (machine tools, robots, etc.), which are then sold to manufacturers that use them to make all sorts of stuff.  For the most part, I'm talking about cutting and welding metal, but lasers also do plenty of engraving and marking (to identify parts).  Lasers can be much more precise than traditional methods, so this industry grows 6-7% in a typical year as it takes market share. 

The bad news - if you will - is that laser technology improves all the time, and it's quite common for startups to win market share selling the latest and greatest.  So the industry is fragmented, albeit consolidating. The good news is that it's hard to lose market share: once your lasers are designed into some manufacturing process it's difficult for customers to replace you.  There's generally a recurring revenue stream from replacing parts that get worn out, and I'd imagine the margins on these are good.  End users don't care about having the fanciest laser technology - they just want what works in terms of cost, throughput and downtime (reliability).

GSI's smaller business makes three types of semiconductor equipment.  Here, GSI combines its own lasers with other suppliers' parts and sells the complete system directly to end-users (the semi fab's). The bad news here is cyclicality and a concentrated customer base.  The good news is a very limited number of competitors.  These little markets ($200-300 million in total) are basically duopolies, with GSI usually facing off against Electro Scientific Industries (ESIO).  The most important of these - WaferRepair - was absent for GSI around 2001-2004 owing to the loss of a patent infringement suit to ESIO.

Brief history:
GSI had historically been one of the consolidators in the non-diode laser industry, and so it wound up with some nice businesses and good brands.  Unfortunately the whole never seemed to equal the sum of the parts for shareholders owing to chronically weak management: the acquisitions were never really integrated, never enough cost cutting, aforementioned patent issue, etc.  

But the real kick-in the pants came in 2008.  GSI levered-up to do a big, overpriced acquisition of Excel Technologies (XLTC) at the top of the market.  Excel is/was a very nice asset, mind you, but the price/timing/capital structure was all wrong.   Shortly thereafter, GSI 'fessed up to a number of revenue recognition errors (a technical default under the new debt) and along with the economic downturn the shares did a swan-dive from $9 to $1.

At this point, Stephen Bershad began buying in, eventually accumulating over 10% of the equity.  Bershad is the former CEO of Axsys Technologies (AXYS), which he ran from 1986 until he sold it to General Dynamics in June of 2009.  Axsys makes infrared cameras - not exactly the same little niche as GSI but darn close.  He knew the company well as the names GSI and Excel had come up frequently over the years when looking at potential M&A, new hires, etc.

Some good corporate drama ensued.  Here's my interpretation: former management saw the writing on the wall: Bershad wanted them gone, and rightly so.  They put a golden parachute in place, then later threw the company into bankruptcy in a last-ditch effort to save their jobs (and tried to grant themselves 8% of the new equity in the process!).  In my view, bankruptcy was completely unnecessary: GSI was and is fully capable of paying the bills. 

Bershad organized a shareholders' committee which eventually managed to retain about 80% of the equity by the time GSI emerged.  A recent rights offering helped to reduce the company's debt from $210 to $107 million.  Post-restructuring costs, they'll have about $35 million in cash.  The new debt includes a PIK option, but frankly I expect little trouble servicing the debt given EBITDA of roughly $60 million in 2010 (pre-bankruptcy costs) and $80 million in 2011.  I think this debt will be paid down fairly quickly, and note there are no penalties for doing so.

By the way:

While I don't think this is a situation that requires some special talent to work, Bershad's track record deserves some mention.  Here's what I know: he took over Axsys in 1986.  My data goes back to about 1991, when AXYS had tangible book value per share of $1.29.  By 2008, TBVPS was up to $8.17 and he sold it in mid-2009 (not exactly a frothy market) for $54/share.  So my guess is that shareholders made at least a high-teens % return for ~25 years under his tenure.  If the new GSI did a road-show in the future I imagine it would be well attended.

Speaking of track records, I can't resist talking about Excel.  Again, my data goes back to about 1991, when its TBVPS was $0.71.  Coherent (COHR), a competitor, offered about $30/share in 2006 but ran into anti-trust issues.  By 2007 TBVPS was up to $11.45, and GSI paid $32/share the following year.  I'm guessing shareholders made around 20% a year for two decades.  Point is, we own this baby now. 

The numbers:

We're not in the dark here despite the lack of filings.  We have unaudited 2009 financials and a series of 8-Ks detailing progress on bookings, sales, cost cutting and margins through 2Q10.  And we can see a strong rise in bookings that dovetails with what competitors have noted on their conference calls.  Particularly encouraging is the rebound in the semi equipment sector, which tends to move in multi-year cycles, suggesting this increase has legs.

I think the recent numbers suggest some good things:  first, that sales should rebound strongly in 2010, and second, that the cost cutting begun in 2H08 has had a lasting impact.  Former management put out an 8-K in 2009 claiming $26 million in run-rate savings from post-merger layoffs.  The latest sales and adjusted EBITDA numbers suggest to me that these savings have indeed remained in place.  Also, I frankly wouldn't be surprised if the new management can improve on these synergies. (Here's my model: http://rghost.net/2758087   See line 79 and below for projections.)

Valuation:

There are no perfect comps here, but a basket composed of COHR, CYMI, ESIO, NEWP and RSTI probably gets you within the realm of reason.  Note that while ESIO has been mentioned as the closest comp given that GSI and ESIO were duking it out for those semi equipment sales, the merger with Excel significantly lowers GSI's exposure to this area.  IPGP has also been mentioned, but I don't use it.  IPGP is the leader in "fiber" lasers, which are the new, new thing in laser tech, and IPGP's multiples are far higher than where any comp ought to trade.

(By the way, good luck finding any mention of GSI in recent sell-side research.  The bankruptcy really pushed this company off the radar, which is why I think the new financials and re-listing could be a great catalyst.)

So, if GSI traded at the average multiple of ___, the shares would be at ___, for a gain of ___:

  Current year sales, $4.46, 78%

  Next year sales, $3.86, 54%

  This year EBITDA, $3.35, 34%

  Next year EBITDA, $3.32, 33%

  This year EPS, $4.21, 68%

  Next year EPS, $5.41, 117%

Note that I've excluded 2010 restructuring costs in the above calculations. 

While GSI's EBITDA margins are similar to the average of these comps, it ought to turn more of EBITDA into FCF owing to a low consolidated tax rate of about 25%, thus I prefer P/E and FCF based valuations.  At an EV of 10X 2011 (mid-cycle?) free cash flow, the shares would trade at $4.30 for a nice ~70% gain.

Risks (both good and bad):

In the bankruptcy proceedings it was revealed that a number of parties had offered to buy GSI in whole or in part.  One offer equated to a current share price of about $2.40, suggesting to me that today's price is at/near rock-bottom.  Why?  Because this was the very first offer made for a company with delinquent filings going through bankruptcy amidst a huge global downturn.  If that ain't a recipe for low-ball offers ...  The other reason I bring this up is that it wouldn't surprise me to see some individual businesses sold and the proceeds used to retire debt.  Every large laser company is an amalgamation of small businesses and there's a lot of horse-trading that goes on.

We don't know who the new CEO will be yet.  But I think Bershad's involvement as Chairman and major shareholder mitigates this risk.  The shareholder base also includes some former industry exec's and funds who were involved in the bankruptcy process and continue to watch closely.

On their last call, ESIO claimed to have taken some market share in the semi business during the bankruptcy.  I doubt it was a lot, but don't know for sure.

GSI could face some headwinds from new laser technologies.  I'm not very concerned, but thought it's worth mentioning anyway.  Fiber lasers are clearly superior in some respects and are taking share from older laser types (which in turn are taking share from non-laser technologies).  In other ways, however, fiber falls short.  IPGP is the clear leader in fiber, with all the other big guys playing catch-up (GSI sold only $1 million or so of fiber last year).  I think GSI has a good deal of time before fiber captures much of its end-users' demand.  Here's a good article: http://www.electrooptics.com/features/feature.php?feature_id=84

Another possible headwind comes from lamp-pumped lasers - a segment in which GSI is strong.  Sales of these have been flat for some time, which means lost share in an industry which grows 6-7% over time.  In some applications, e.g. marking and engraving,   Another good article: http://tinyurl.com/33tb7ft