Article Info

Vitesse Sells Storage

image After dropping hints in an earlier conference call (see Vitesse Q207 Conference Call Notes) Vitesse announced the sale of a portion of their storage products to Maxim (MXIM). The $63M transaction has a provision for an additional $12M if certain milestones are hit. The company will use the proceeds from Maxim and $30M in new convertible debt to repay the junk debt to Tennenbaum Capital. (see Tennenbaum and Vitesse).

Three months ago I expected the sale of the entire storage unit to fetch 3.5x revenue. Subsequent research and discussions led me to believe this estimate was inaccurate, something reflected in comments I made (see here).

In the end, the outcome was far less than what was originally anticipated. Instead of 3.5x overall storage revenue ($200M+), the company received 2.5x-3.0x (depending if the additional $12M is met) on $20M of it’s $73M in TTM storage revenue. But this $20M represented the vanguard of the product line growth. The 2G Fiber Channel products are an annuity, and Vitesse’s position in ROC without a full complement of storage products is tenable.

A Concentrated Focus

With this divestiture, Vitesse is essentially out of the storage business. Not that that’s a bad thing, but something that needs to be taken into account when valuing the company. One gets a clearer picture of the situation when you look at storage revenue ex-SAS/SATA.

  Q306 Q406 Q107 Q207 Q307
Consumption  $    53  $    59  $    61  $     58  $     61
Networking 28.1 28.9 29.9 29.9 32.7
Storage – Legacy 16.4 15.1 16.4 12.3 9.5
Storage – Divested 5.0 5.0 5.0 5.0
Ethernet 8.5 10.0 9.8 10.4 13.3

Flat SAS/SATA revenue is assumed. If you assume growth then the falloff in legacy storage is steeper.

This transaction puts Vitesse in a position in which it can concentrate R&D and SG&A in it’s core networking businesses. This is where the operational DNA for the company exists and makes sense. The future of Vitesse now depends on it’s ability to grow replacement revenue in this area.

The company is almost certainly generating significant earnings at this point but my concerns about revenue inflation and Opex deflation temper my enthusiasm for this metric.


In the short term the company is generating cash flow on investments made in the past. Adjusted opex is now at $20M on $56M (36%) in revenue, far below it’s competitors (PMC opex is 63% of revenue, AMCC is around 55%). It is unlikely that Vitesse has suddenly reached a pinnacle in fixed cost efficiency. This leaves investors with a decision – can the company make more efficient investments than it (and it’s competitors) have made in the past, or will operational expenditures rise in order to grow the top line?

Another effect to consider – the number one priority at Vitesse from dawn till dusk was to generate cash in the short term, particularly by monetizing existing inventory. This has the side effect of driving revenue forward. This is not a criticism of management as this is exactly what should have been done to ensure liquidity. The problem may be when investors assume top line growth is organic, rather than aggressive cash management. Some metrics from the company in this area would be helpful.

Go Forward

Unlike PMC-Sierra (PMCS), Mindspeed (MSPD), or AMCC (AMCC) which have broad product lines, Vitesse is now a pure-play optical networking component company. They must reach a level of R&D efficiency that can drive replacement revenue at existing Opex levels. No one else in the industry has managed to achieve this and I would argue some consolidation is needed before anyone, including Vitesse, can make it happen. The singular focus of Vitesse in optical networking makes it a virtual certainty they will play some sort of role.

Update 30 Aug 2007: The purchase agreement is now on Edgar. The earn out looks very achievable, it is simply a 1:1 payment by Maxim to Vitesse for every dollar of revenue they ship. I.e. if Maxim ships $5M in ex-Vitesse product in CQ407, they pay Vitesse $5M towards the earn-out. This should result in Vitesse realizing the full earn-out by July of ’08. As long as Maxim ships $12M in product by September ’08, Vitesse should recognize 100% of the earn out. It isn’t clear to me why this was made so complex.

Author is long Vitesse


Comments are disallowed for this post.

  1. Hi Andrew – couple things:

    1) the new debt is not a convert, it’s a straight loan, presumably @ LIBOR xx. whitebox has the option on an incremental $45M convert @ $2 for 3yrs

    2) i suspect the earn out is more a function of a lack of audited financials than performance related, as neither MXIM nor VTSS actually will control sell through for the next 12 months (the duration of the earn out period). if i am a MXIM lawyer, i just want to protect my downside if the revenue #s VTSS is quoting don’t turn out to be kosher. VTSS would be more likely to “gamble” on their #s (they know if right or wrong) vs. the shipment levels of their OEMs a year out. if you include the earn out, the price is quite healthy, particularly if you look at it on ROIC ( $75M in proceeds, $-10M in GP, 10M in OpEx)….

    3) growth is over-rated, ask Eddie Lampert (see any of his investments). running the biz for long term cash optimization is would you or i would do if we owned the whole thing– i tend to think Gardner has transitioned from slash/save to maintain/build. That’s not to say that a cyclical upturn in the networking biz won’t help the cause….what these guys have done in the last 9 months is really remarkable.

    4) glad to hear you are on board. i suspect that if it wasn’t so damn hard to figure out what the #s are here and we has some liquidity, this thing would already be > $2.

    Posted by roger | August 24, 2007, 5:52 PM
  2. I don’t think it is incremental as the release says the proceeds from the convert purchase would pay off the debt, if issued. I am _assuming_ the $45M matches the 30M 15M.

    You are correct I am speculating on the earn out. I don’t know why anyone bothered with this given it is only 20% more.

    Sure it looks good for Vitesse in the short term but realize that this was supposedly a big growth area.

    If people believe profitability is sustained I agree you are looking at one cheap stock.

    Posted by Andrew Schmitt | August 24, 2007, 7:04 PM
  3. Andrew,

    how much is the RAID ON CHIP (ROC) worth? Do u think there is a potential buyer just for the ROC? thx

    Posted by alphnasx | August 24, 2007, 9:04 PM
  4. Any idea what VTSS paid for these lines and when?

    Posted by osxman | August 25, 2007, 6:11 AM
  5. VTSS grew them organically from a small startup they bought 9 years ago for a small amount of $$$. It was like $15m but they did very well.

    Posted by Andrew Schmitt | August 25, 2007, 7:29 AM
  6. Valuations are difficult since financial numbers are a bit sparse. Nevertheless VTSS’s huge R&D investments have generated plenty of IP over the years. The market may not have needed it all though.

    The recent sale has given VTSS the chance to profitably survive. Mere existence was a big question until now and something that the slow trading days in late August have failed to fully account for.

    Sure valuations in the $3.50 range are easy to compile but many investors with a 4 to 8 year horizon are likely looking at a $15 valuation. It all depends on the fiber market.

    Posted by David Gillooly | August 25, 2007, 1:20 PM
  7. How is VTSS a pure optical play?? Networking is flat for the year with a slight pop in the last quarter. There is a reasonable chance is that this is due to one time events (end of life). Meanwhile it looks like Enet is growing nicely. I’m assuming the the majority of this revenue is from switch / cu phy biz. I’m not seeing the light!

    Posted by m@ | August 25, 2007, 8:16 PM
  8. My comment is based on the fact that all future R&D is concentrated in this area.

    Posted by Andrew Schmitt | August 25, 2007, 8:49 PM
  9. Andrew,

    The ROC came from Adaptec, not home grown. They bought the rights to resell the PCI-E to SAS Controller (but never owned the IP!). They bought this from Adaptec with the hope that server vendors would embrace ROMB (RAID on Motherboard). IBM went down that path with this device (as IBM has been Adapatec’s largest customer for a long time). However, other server vendors did not embrace the ROMB solution and instead went for a more flexible and more expensive HBA solution where the controller is mounted on the HBA. This allows a server vendor to offer different levels of RAID (or no RAID) by plugging in different HBAs. As Vitesse was never able to capture any other large server vendors, the ex-Adaptec team has been reduced to a skeleton crew to support their large customer. At this point this business is legacy revenue stream that probably already has seen its pinnacle as MXIM did not pick up this piece.

    Posted by Mr Storage | August 27, 2007, 12:55 PM
  10. Further clarification…

    As pointed out by MrStorage, the ROC products came from a partnership/acquisition with Adaptec, and MXIM did not pick this up.

    The ~ $15M acquisition Andrew referred to was most likely Serano Systems, acquired by Vitesse in early 1999. The Serano product architecture… married with the supreme Vitesse high-speed serializer/deserializer technology… ultimately yielded the SAS expander product family, which was “grown organically” by the Colorado Springs operation. This product family, plus the BMC products that also owe their roots to Serano, appear to be the meat of the $63M/$75M acquisition by MXIM.

    Posted by The other Mr. Storage | September 8, 2007, 9:24 PM
  11. Correct.

    Posted by Andrew Schmitt | September 9, 2007, 7:53 AM
  12. Trackbacks / Pingbacks

  13. Vitesse Q208 | Nyquist Capital | May 19, 2008, 11:03 AM