Over the past three days, Quicklogic (NASDAQ:QUIK) shares have declined 32%, the loss triggered by news that the company completed a public offering of 4.5 million shares of stock at $2.50, which was significantly below the then market price of $2.98. The stock offering included a five-year warrant to purchase 0.45 shares of common stock for each common share purchased exercisable at $2.98.
Given the terms of this offering, a decline in the stock from $2.98 to the offering price of $2.50 would not be at all unexpected as buyers of these news shares would be motivated to immediately flip the stock and lock in a healthy gain. However, with well in excess of 5.5 million shares having traded over the past three days and the stock trading all the way down to $2.03 per share, well-below the offering price, it would appear that the offering has caused quite a bit of panicky selling from the existing shareholder base as well. We think this selling is ill-timed.
To be clear, we were initially surprised to see the company undertake this financing. After all,
- the company had $17.8 million of cash on the balance sheet at the end of the first quarter;
- its new product families (the ArcticLink III-VX and ArcticLink II-CX) are out of development and moving into production in the coming months, which, according to management, will cause the cash burn rate to decline;
- the potentially high-volume Pop Video Pico Projector (an Apple-approved accessory for the iPod and iPhone), which Quicklogic is designed into is now in production; and
- according to management the new VX family of chips is seeing good design momentum with major OEMs across both handsets and tablets. This suggests good visibility into a long-awaited revenue ramp that should be very noticeable in the second half of 2012 and accelerate in 2013.
It appears that investors are puzzled as to why, given these considerations, the company would need to raise any funds and have simply drawn the conclusion that the expected long-awaited near-term revenue ramp may not materialize. Based on our recent conversations with management, we are convinced that investors are drawing the wrong conclusions from this financing and rather than signal a tepid fundamental outlook this financing confirms the company has actually made a clean break with its disappointing history and bolsters the long-term bullish outlook for the company.
Given Quicklogic's track record it is understandable that investors have developed a "flight not fight" instinct around the stock. There is no argument that the company's recent past is characterized by missed opportunities and a failure to predict and adapt to changing industry dynamics. This has left Quicklogic long on promise but very short on results, and investors with a lot of volatility but not much to show by way of investment returns. Recent stock price action not withstanding, we think this is about to change.
A brief review of Quicklogic's background is necessary to put this recent financing into the appropriate context. Quicklogic pioneered a new class of semiconductor it calls Customer Specific Standard Products (CSSPs), which are very low-power programmable parts ideally suited to mobile devices. In a mobile design (handsets and tablets) they sit next to an apps processor and in addition to connectivity fabric, they are comprised of various Proven System Blocks, which are comprised of intellectual property both developed by Quicklogic as well as licensed from third parties. By including programmable CSSPs in a design, OEMs can utilize this intellectual property to customize their devices and add new features to a design on the fly without waiting for a next generation apps processor to be developed. Additionally, because CSSPs are extremely low-power, devices can be designed so that the high-power apps processor offloads tasks to the low-power CSSP in order to improve overall device performance.
One of the most compelling use cases for the company's chips is the company's Visual Enhancement Engine (VEE), which is based on technology (the iridix algorithm) that Quicklogic has licensed exclusively for mobile applications from Apical Ltd., as well as Quicklogic's own IP. The iridix algorithm, already widely used consumer electronic markets (digital camera and televisions), performs a pixel-by-pixel optimization of device displays in order to improve picture and video quality in all lighting environments while simultaneously reducing the power draw of the display.
Quicklogic launched its VEE solution in 2008 and despite the compelling experience VEE provided in terms of display quality and battery life savings (according to OEMs who evaluated the technology) adoption of the technology was anemic. The reason for this was that the initial implementation of VEE was in a standalone chip and therefore required additional board real estate and added to design complexity and cost. After getting a thumbs up from OEMs on the concept of VEE but a thumbs down on its implementation, Quicklogic realized that to overcome the design issues presented by VEE as a function of a standalone chip, VEE would need to be implemented as a feature in a display bridge chip.
A bridge chip sits between an apps processor and the display and in addition to doing screen refreshes and buffering, it also provides an interface between the application processor and the display itself. For example, an application processor with a MIPI interface used in a tablet with an LVDS display requires a bridge chip (a MIPI-LVDS bridge) as the apps processor cannot interface directly to the display. Because display bridges were already common to many handset designs, implementing VEE in a display bridge would not require additional board real estate, add to design complexity, or necessarily add to the cost of the design.
The first display bridge chip Quicklogic introduced was the VX2, which supported an RGB processor interface to RGB display. Unfortunately, this solution found little market traction in handsets as applications processor companies transitioned far more quickly than expected from RGB interfaces to what has become the industry standard known as MIPI -- a standard that Quicklogic now supports in its ArcticLink III family. However, it should be noted that while Quicklogic's VX2 RGB-to-RGB display interface found very limited initial traction in handsets it has recently found a new life in Pico Projectors. In fact, the dramatic improvement that Quicklogic's VX2 chip provides Pico Projector images is so significant that it has been called a changing technology with the potential to jump-start sales of this heretofore moribund product category. Recent estimates suggest that Pico Projector volumes could reach 50 million per year by mid-decade.
On the heels of the VX2, Quicklogic developed its VX4 display bridge to support Qualcomm's (NASDAQ:QCOM) proprietary MDDI interface. However, despite the seeming wisdom of attaching the company's fortunes to Qualcomm, the VX4 also had only limited success. Shortly after this chip came to market -- in a move seemingly coincident with its comprehensive settlement agreement with Nokia -- Qualcomm decided to drop its proprietary MDDI interface in favor of the MIPI standard. As a result, despite having partnered with Qualcomm on the VX4 development and being included in a Qualcomm engineering reference design for Snapdragon processors, Quicklogic's VX4 also found only limited success.
While it is tempting to chalk Quicklogic's lackluster history up to a series of unfortunate exogenous events, a more accurate and critical reading of the history is that Quicklogic had up to this point failed to develop a silicon roadmap that anticipated the changing landscape of the mobile market place and requirements of its customers. Instead, the company tried to market VEE through a series of point products whose time had come and gone before the products even hit the market. As noted, the VX2 and VX4 chips could only interface with RGB displays, which OEMs were already moving away from, and with the VX4 the company tied itself to a Qualcomm proprietary interface standard that was abruptly abandoned. Furthermore, while tablets are an ideal use-case for a technology like VEE, Quicklogic could not address this market as tablets almost exclusively use LVDS displays not RGB.
This changed when Andy Pease became CEO at the beginning of 2011. As his first order of business he had the company develop a long-term silicon road map and accelerated the development of the company's next family of processors, the Arctic Link III-VX. The ArcticLink III VX family of processors -- which sampled in the first quarter of 2012 and will be in production in the third quarter of 2012 -- includes 13 different processors able to support all OEM design preferences and requirements across all end markets: handsets, tablets, and Pico Projectors (stand-alone and embedded).
In addition, the company has already begun work on the next generation ArcticLink IV family of processors. As part of the recent financing, Quicklogic disclosed that it "may also use a portion of the net proceeds for licensing or acquiring intellectual property or technologies to incorporate in our products." In our opinion, this is an important disclosure that investors dismayed by the financing and accustomed to being disappointed by Quicklogic may have overlooked. To us it indicates that, under Andy Pease, there has been an unequivocal clean break with the company's history of failing to anticipate the future requirements of its customers.
So, while the assumption that the company's balance sheet should have been sufficient to manage an impending production ramp of its current ArcticLink III-VX and ArcticLink II-CX processors without additional funding is correct, the balance sheet really needs to be considered in the context of the development of the silicon road map as well. Depending on the size and terms of a technology acquisition or licensing deal the company may be contemplating, the current balance sheet may have been limiting.
There are two important conclusions for investors. First, this financing does not indicate a delay in the design momentum or call into question the outlook for a near-term revenue ramp based on the current family processors. Second, the company is making the necessary investments in ongoing product development, which is essential for it to become a trusted supply-chain partner to OEMs and partner to the large Tier 1 silicon companies such as Qualcomm, Nvidia (NASDAQ:NVDA), Texas Instruments (NYSE:TXN), etc.
Is it preferable to fund product development from company cash flow? All things being equal, of course it is. But given that the revenue ramp of the current products is still a couple quarters away while the company needs to make investments in next generation products today in order to meet future customer requirements, that option was unfortunately not available. While the stock price action since the offering has been clearly disappointing, it represents an excellent opportunity to build or add to positions in front of the fundamental momentum we expect to become apparent over the second half of this year.
Disclosure: I am long QUIK.