IKANOS Study (For Earnings Model Click Here) March 10, 2008

Strategy for Analyzing IKANOS

Commentary in this analysis arises from two sources:

(1) Studying the balance sheet, income statement and cash flows of the company to determine its financial standing, ablility to generate earnings, and need for future capital

(2) Discussing the company and its prospects with various industry contacts to determine:

a. Competitive position

b. Market opportunity (demand)

c. Historical profitability of this industry

d. Strength of management and morale

Balance Sheet and Cash Flow Analysis

The balance sheet does not show any significant signs of stress. DSOs remain tame at 52 days though up 10 days from an unusually low DSO of 42 in the Sept quarter. Inventories remain well contained with turns around 5x for all four quarters of 2007.

Cash flow from operations was close to zero in 4Q07 though the company did improve its overall cash position in 2007 due to some sale of investments. Free cash flow (as defined by Net Income + FA123 adjustments + D&A – CapEx) was also close to zero in 4Q07. Our revenue sensitivity analysis suggests the company has to get to 20% revenue growth to achieve cash flow breakeven in 2008. With consensus growth now estimated at 30%, IKAN would be cash flow positive in 2008 should they meet consensus numbers. With cap ex running about $1M per quarter and D&A running about $2.7M per quarter, free cash flow will be significantly higher than pro forma earnings. In fact, our model suggests IKAN could generate $0.80 in FCF in 2009 based on consensus estimates. However, for reasons we will discuss in this analysis, we believe 30% top line growth with 45% GM in 2008 will be difficult to achieve.

Income statement is highly leveraged to top line & Gross Margin percentage

Even if we assume the SG&A dollars can be modulated with rising and falling revenue levels, the nature of the DSL business requires high levels of R&D which remain relatively fixed regardless of whatever machinations the top line may undergo. In other words, even during tough times, the company has to keep spending on R&D. Thus, the company’s earnings are highly sensitive to the Gross Margin percentage and revenue levels. We now discuss the sensitity of earnings to variations in Gross Margins and Revenues.

Gross Margin Sensitivity

Currently, the Street is assuming that the GM% between 45% and 46% can be maintained indefinitely. Street models thus call for GM% to be 45.3% in both 2008 and 2009. However, historically, there has been significant margin pressure in the DSL chip business and GM% usually heads downwards as the technology cycle matures. In 4Q07, the company reported a huge sequential increase in GM% from 37% in Sept to 45.8% in December. In part, this increase was due to a mix shift towards more Central Office products. We expect this shift to vary in future quarters and thus believe the company may have trouble maintaining GM% at current levels.

For each 100 bp increase/ (decrease) in gross margins, 2009 pro forma estimates increase/ (decrease) by $0.05 while fully-taxed estimates change by $0.03.

On the flip side of the gross margin argument, the company’s recent move to consolidate (buy) Centillium’s DSL business will likely ease GM pressure in Japan in 2008.

Revenue Sensitivity

The conservative and aggressive income statement models show the sensitivity of the earnings to the top line. The Street has modeled a 30% revenue growth estimate for 2008 followed by 20% in 2009. In our conservative model, we assume 20% growth in 2008 and 10% in 2009. In our aggressive model, we assume 30% for both years. We have kept GM assumptions the same in all three models as GM sensitivity has been discussed separately (above).

Street Estimates currently remain untaxed

Pro forma (and GAAP) earnings estimates for IKAN remain untaxed (actually, at a 3% effective rate). While this may be appropriate given that the company will not pay taxes for a long (long) time, for valuation purposes we have also included fully-taxed estimates on the earnings models attached. On a fully-taxed basis, IKAN is currently trading at 20x the fully-taxed 2009 estimate of $0.23.

Industry contacts say Street revenue estimates may be too aggressive for 08

Street estimates currently call for IKAN to grow top line 30% in 2008 and 20% in 2009. Given Cisco’s recent comments about Europe slowing, and IKAN need to diversify outside of Japan into Europe, we wonder if the company will be able to achieve this level of top line growth under this difficult macro environment.

Customer Concentration

IKAN continues to derive a significant portion of its revenues (40% in 2007) from two customers (NEC and Sumitomo) in Japan. We expect this concentration to abate slightly in 2008 as ALU expands in Europe.

Competitive Positioning

Worldwide DSL Landscape
Vendor	Supplier	Carrier Customers	Technology
ALU	BRCM	France Telecom	ADSL2+
ECI	INFY	France Telecom	ADSL2+
Huawei	INFY	France Telecom	ADSL2+
Sagem	IKAN	DT	VDSL2
Siemens	INFY	DT	VDSL2
ERIC	INFY	DT	VDSL2
ERIC	INFY	Denmark	VDSL2
ALU	CNXT	AT&T	VDSL2
ERIC	CNXT	AT&T	VDSL2
Various	IKAN	Japan	VDSL2
ALU	IKAN	Swisscom	VDSL2
ALU	IKAN	Netherlands	VDSL2

By buying CTLM’s DSL business, IKAN has effectively netralized the competition in Japan. (Our contacts at Aware as well as at Centillilum suggest that IKAN bought this business to shut it down). These industry contacts further suggest that despite having good technology, that Infineon is poorly positioned politically in Japan. Meanwhile, Conexant appears to have “bet the ranch” on AT&T’s U-verse build-out in the US and is not dedicating significant resources in Japan. As such, we believe that IKAN has a near lock on the Japan opportunity for 2008.

However, our industry contacts also suggest that the Japanese carriers are not happy about being left with a monopoloy supplier. (They feel that in a previous product cycle, when CTLM had a monopoly position, that the pricing they got from the supplier was unfairly high). Thus, we believe that during 2009, IKAN Japanese customers may attempt to add at least one other supplier. Our industry contacts could not tell us if this competitor is likely to be some existing threat like INFY or CNXT or some new threat such as BRCM.

In addition to the three vendors (IKAN, INFY, BRCM) in Europe, CNXT remains committed to AT&T's U-verse deployment in the US. Historically, silicon competition tends to increase in mature technologies that have well established industry standards, thereby pressuring margins longer term.

Our proprietary sources also suggest that ECI Telecom is attempting to sell its CO-DSLAM business. (So far, there have been no takers). What effect, if any, this may have on IKAN is unclear.

Long term lack of profitability in semi subsector for standard products

Looking at the universe of silicon companies, long term profitability (evidenced by Retained Earnings) is typically only achieved by four categories of semiconductor suppiers:

(1) Analog (LLTC, MXIM, etc.)

(2) FPGA (ALTR, MXIM, etc.)

(3) Monopolies (INTC, TXN, etc.)

(4) IP providers (QCOM, MIPS, ARMHY, etc.)

The only exception we know if to the above rule is MCHP. IKAN does not fall into one of these four categories. Thus, we remain suspect about its longer term profitability outlook.

Conclusion

While shares of IKAN no longer appear expensive by the metrics we have discussed in this analysis, it does not appear exceedingly cheap either. We believe that the risk/reward ratio for IKAN given the difficult economic environment continues to lean towards the negative and thus have formed a NEUTRAL opinion on the stock. The basis of our opinion rests on the idea that they will not be able to achieve 30% revenue growth in 2008 followed by another 20% revenue growth in 2009. If we are wrong about this conclusion and IKAN is able to achieve this top line growth (while maintaining margins), its strong implied cash flow would likely cause the stock price to double from its current level.