Barclays' CJ Muse thinks it's time semiconductor companies rethink how they return cash to shareholders -- and he's got a way that makes everyone happy.
Too much cash? Try a variable dividend
Barclays' CJ Muse thinks it's time semiconductor companies rethink how they return cash to shareholders, given their cash positions. Net cash is currently ~15% of equipment companies' market cap -- but they have been hesitant to drive shareholder value through dividends or stock buybacks that could turn out ill-timed with market fluctuations. When times are good, firms are flush with cash; at other times they're afraid of being locked into fixed obligations during a downcycle. That hesitancy is reflected in investor's lack of support in valuations, too: "Investors do not believe managers, in general, are good shepherds of cash," he writes. "Too many times in the past we have seen dilutive acquisitions, too aggressive compensation, and poor business models that burn cash during the downturn."
But that corporate reluctance needs to change, Muse argues, since free cash flows are likely to continue through future cycles, thanks to "quiet share consolidation" (i.e., non-M&A) that will fuel better pricing and business models. Plus, semiconductor capital intensity is only going up, capex probably won't sink back to abysmal 2009 levels, and maneuvers including outsourcing and cost cutting will make companies more variable-cost in nature.
Muse's suggestion: implement a "variable-rate" dividend, where a payout occurs only if the company reports a quarterly GAAP net income in excess of the proposed dividend. That, presumably, would lead to beefier dividends during good times, and less risk for when markets soften. The alternative -- issuing rather small dividends today (1%-3% for some of the big names) -- "are just not enough to attract value investors," he says.
So what would the sector look like if such a plan were enacted 10 years ago? A number of top semi equipment stocks would have paid out dividends ~50% of the time (average dividend 2.5%), but without depleting cash reserves during bad times. And extrapolating into the future, he thinks average payouts would increase to 75%+ of the time -- "meaning the return of excess cash to investors will be a critical differentiator for equipment companies in the future." Those who might benefit most from such a plan? Logical names include LRCX, TER, KLAC, AMAT, and NVLS -- companies who already pay dividends, or have clean balance sheets, continue to cut costs/narrow focus, and aren't investing heavily in adjacent areas (e.g. solar, other tech hardware).
Takeaways from DB Tech conference
Among several sector presentations at last week's Deutsche Bank Technology Conference:
AMKR: Normal seasonal outlook for 2H10, said CFO Joanne Solomon, with weaker PC shipments somewhat offset by strength in consumer electronics, notably smart phones. The company sees an unchanged ~$500M in 2010 capex, though some peers are reporting customers slowing and pushing out tool shipments. In fact Amkor probably has lower capital intensity than large peers due to end-market mix and a less aggressive ramp in copper wire bonders -- and this should keep revenue growth on pace with others. Look for capital intensity to return to "a more normal 10%-14% level," and stronger cash flow, in 2011, O'Rourke writes.
AMAT: No new guidance, no changes to outlooks; "DRAM, NAND, foundry, and logic capacity ramp profiles have neither changed nor been a surprise." GM Randhir Thakur did acknowledge macroeconomic concerns ahead. Deutsche Bank analyst Stephen O'Rourke notes that a projected 4% gain in etch does not include any position at Intel (and LRCX isn't there either, meaning it's a TEL win). It's also gained share in mask inspection, though mainly from smaller firms (not KLAC).
ECD: No change to guidance from CEI Mark Morelli. Sept. quarter revenues will be down as expected, "but the overall trend is clearly up." Module pricing remains intact (current ASP is ~$2.05/Wp), and manufacturing costs should decline to $1.60/Wp at current production rates. Conversion efficiency goal is still 10% in 2011 and eventually 12%.
LRCX: CFO Ernie Maddock confirmed weaker semiconductor industry data points, but said that no customers had changed delivery schedules or overall spending plans. The company's latest efforts at wet clean "clearly gaining traction," and paired with its dry bevel clean technology the company "is poised to surpass the 50% market share point in overall etch" writes O'Rourke. "Lam has gotten past the label of being a single product company."
RTEC: CFO Steven Roth noted, but did not reiterate, current 3Q guidance and expectations for sequential growth in 4Q10, and echoed other industry sentiment that a robust spending cycle for backend will lead to "a digestion period" and slowdown. In fact, this appears to have already started, with some backend packaging/test outsource suppliers now pushing out tools, notes O'Rourke. With ~40% exposure to backend customers, Rudolph's near-term outlook could be "tempered." It's worth noting how much Rudolph has changed in the past few years, he adds: while markets for its frontend metrology business have underperformed (though notably a new win at a large DRAM/NAND chipmaker), most of its sales are now from new tools, and it's been active with M&A in backend inspection/analysis, segments heavy with use of leading-edge technologies including 3D packaging, flip chip/bump, through-silicon vias (TSV), and probe cards.
VSEA: No change to guidance from CFO Robert Halliday; fiscal 4Q10 (Sept. 2010) is tracking in-line and F1Q11 could be "incrementally positive." DRAM business is weakening as expected, but foundry is growing and NAND is expected to grow. Overall 2011 spending could be flat Y/Y, he suggested.
Notebook outlook: Cloudy, chance of rain?
FBR Research's Craig Berger says checks indicate PC inventory depletions in Asia are nearly over, with HP and Acer both at about four weeks (down from 6-7 in mid-July), though Dell is a bit behind. While sluggish consumer demand is still reported in the US and Europe and iPad seems to be stealing share from notebook PCs, chip firms should see "a slight business uptick as they transition from customer inventory de-stocking to shipping more in line with end consumption," he writes.
Speaking to notebooks, Ashok Kumar from Rodman & Renshaw sees a flat September for notebook shipments from top ODMs, following a weak back-to-school season, "lackluster demand from Europe," inventories plugging the channel, and the aforementioned iPad cannibalization. The December year-end quarter is still unnervingly murky, though early indications point to perhaps a -10% decline or more in shipments, as Acer, Dell, and HP cut back orders to ODMs.
Taking Toshiba to task
With demand for memory going nowhere but up, and robust demand for embedded NAND flash a good area to leverage technology advantages, why be down on Toshiba? Because, says Deutsche Bank's Takeo Miyamoto, the company's memory business has just 14% profit margins (based on guidance) while some competitors post 20% or more. And product mix issues, a strong yen, and startup R&D costs have hampered efforts to improve profitability -- which is vital, he says, for the company to accelerate its ROI in semiconductors where product cycles are so short.