半导体新思维：研发苦海无边，回头讨好股东是岸？

一位美国华尔街分析师指出，有越来越多半导体公司偏向于发放短期性投资报酬给股东，而不再投资长期性的研发项目；如果此一情况属实，对半导体产业界来说会是个发展过程中重大且不幸的里程碑。 “在过去，不投资研发是一个成长停滞以及无法在未来取得竞争力的象征；”该位要求匿名、以便畅所欲言的分析师表示，现在投资人的观点却是：“既然芯片产业不再显著成长，为何还要继续把钱丢进水里？” 据了解，德州仪器(TI) 就是具备上述这种新思维的典范，该公司最近买回了近30亿美元的公司股票，使其股价上涨了40%。在去年5月13日的分析师财报会议上，TI首席财务官发表了 一个已经妥善布置就绪的新模型，让该公司能释放现金流并发放更多股息给股东；TI在过去五年已经回馈给股东的金额，为先前投资目标的113%，累计利润达 140亿美元。 该匿名分析师表示，十年前那些现金可能会被拿来新添1、2座晶圆厂，但今日则不然：“既然成长率一年预期只有5%，为何要拿来投资研发？当该产业的投资报酬率(ROI)下滑，人们也需要调整他们的投资策略。” “每一家公司都在斟酌这个问题，”他补充指出：“成本随着摩尔定律(Moore's Law)不断攀升，但潜在回收却越来越难取得；这些日子以来，你不得不更偏向以一个经济学家的观点来思考，而非技术专家。” 英特尔(Intel) 的策略与TI似乎背道而驰，手上仍拥有许多昂贵的大型晶圆厂，并持续对那些晶圆厂投资；英特尔回馈给股东的短期利润一直以来不那么多，但当该公司某天终于 让那些投资得以回收，也就是获利大潮来临的时候。但投资人的“短视”与企业的“远见”终究是无法契合；该分析师的结论是：“芯片厂商必须要为了走向未来而 投资。” 依笔者所见，半导体技术进展速度总是比成本增加速度缓慢得多，而自从2008年全球金融风暴以来，半导体产业两位数字成长率的时代已经结束；当你开始也加入投资人偏好发放短期股息、而非进行研发投资的转变时，我更伤脑筋的是这个产业要怎么找新成长动力！ 本文授权编译自EE Times，版权所有，谢绝转载 编译：Judith Cheng 参考英文原文：Semis Feel R&D Squeeze，by Rick Merritt

相关阅读：
• 全球CEO峰会：技术引领未来
• 未来十年晶体管成本降速趋缓，半导体价格或提高
• 德州仪器TI希望小学在四川南部县落成并投入使用QNfesmc

{pagination} Semis Feel R&D Squeeze Rick Merritt, SiliconValley Bureau Chief SAN JOSE, Calif. -- Investors are increasingly rewarding semiconductor companies for short-term returns to shareholders, rather than long term R&D investments, one Wall Street analyst said. If that's true, it marks a significant and unfortunate industry milestone. "In the past not investing in R&D was a sign of no growth and poor ability to compete in the future," said the analyst who asked to remain anonymous so he could speak freely. The new view is the chip sector is not growing significantly, so "why keep throwing good money after bad?" he said. Texas Instruments is the poster child of the new way of thinking. It bought back nearly $3 billion of its stock recently, pumping its share price up 40%. In a May 13 call with financial analysts TI's CFO laid out a new model that's already well in place of freeing up cash so it can pay greater dividends to shareholders. In the last five years, TI returned to shareholders 113% of its previous target, a cumulative $14 billion in profits. A decade ago that cash might have bought a new fab or two, but not today. "Why invest in R&D if growth prospects are about 5% a year, ROI has fallen, and people need to adjust their investments," said the analyst. "Every company is wrestling with this," he said. "The costs are rising on Moore's Law, and the potential returns are more difficult to generate -- you need to be an economist as much as a technologist these days," he added. Intel stands on the opposite end of the spectrum from TI, still holding lots of big expensive fabs and still investing in them. Its short-term returns to shareholders have not been so great, but the tide may turn if Intel can make those investments pay off some day. In the end there's a mismatch in the short-term time horizon of investors and the long-term horizon of chip companies. "Chip companies have to invest for way into the future," the analyst said. I know advances in silicon technology are coming more slowly at greater cost. I can see how since the financial collapse of 2008, the sun seems to have set on the days of double-digit growth in semiconductors. When you blend in this shift in investors rewarding short-term dividends over R&D investments, the new dynamics are making me reach for the Maalox bottle.