Letter #15: Intel (1986)

1986 Shareholder Letter from Andy Grove and Gordon Moore

Hi there! I go by KG, and I love studying the history of business and investing. I’ll be sharing some notes from one Investor/Shareholder letter per weekday (mostly from my compilations) here.

Today’s notes are on Andy Grove’s and Gordon Moore’s 1986 Shareholder Letter, Andy’s first year as CEO (See Postscript From The Chairman). For those that don’t know, Andy Grove is a management legend, responsible for inspiring legendary VC John Doerr’s book Measure What Matters and authoring the famous Only the Paranoid Survive and High Output Management. Gordon Moore was a member of the Traitorous Eight and the namesake of Moore’s Law.

You can find this letter in my Andy Grove and Gordon Moore ft. Bob Noynce and Arthur Rock Compilation, which is under “Enterprise Operators.” This particular letter starts on page 35.

If you have any thoughts on what you’d like to see, let me know!


We’re pleased to report 1986 is over. It was, without question, the toughest year in Intel’s history, filled with plant closings, layoffs and deep losses.

As difficult as it was, 1986 also contained within it the seeds of our recovery, seeds planted by the hard work and perseverance of Intel’s people. We’re starting to see the payoff for those efforts.

  • Wow. Andy doesn’t mince words. Right out the gate he addresses the dismal year they’ve had — most letters dance around it (especially HF letters — I’ve actually seem multiple managers leave out their returns lol). “We’re glad it’s over” — what a start.

  • I wonder if Andy taking over from Gordon had something to do with the down year… Was Gordon pushed out?

  • Seems like things started turning around — was that a result of Andy taking over? Not sure if they actually saw anything paying off, or if its just polite talk in an attempt to assure investors/employees

To understand why 1986 was so difficult for the semi-conductor industry, one must go back to 1983-84, a period that can be said to have caused the next two years.

  • To understand the present, look to the past.

BY 1983, demand for semiconductors exploded, fueled in large part by the rapid expansion of the personal computer business. No one could get enough semiconductors, especially Intel microprocessors, which had emerged as the standard for personal computers.

  • The wave died?

  • Sometimes being the de-facto market leader has some drawbacks — if you can’t meet demand, problems arise

Given this atmosphere of scarcity, it was natural that our customers demanded assurance of adequate future supply when deciding what microprocessor architecture to use as the centerpiece of their systems.

  • Makes sense

First, we started a major manufacturing expansion. Capital spending in 1984 was $388 million, up 168% from 1983. Second, we licensed other semiconductor manufacturers to produce Intel microprocessors, peripherals and microcontrollers. We met our customers’ needs and helped expand the total market for our products, but we also lost control over a generation of our products and created our own competition.

  • Lol. Seems a bit like the Apple-Microsoft kerfuffle, thought obviously not exactly

  • Always be careful when licensing

  • Meeting immediate demand may be dangerous in the long run — always have a long-term perspective

The boom collapsed in the fourth quarter of 1984. The computer industry, our major customers, went into a slump just as significant amounts of new semiconductor capacity came on stream throughout the world. The demand projections of 1984 became instant relics in 1985, creating a capacity glut and a severe downward price spiral.

Intel was left with an overhead structure appropriate to the $2-3 billion company we wanted to be rather than the $1.0-1.5 billion company we were becoming. As a result, we have spent the past two years resizing Intel, redeploying assets from low to high growth areas and maintaining strategic programs to get the company back into a position to grow and make money again.

  • Anyone who says capital structure doesn’t matter can go take a hike

  • Take the time to restructure and refocus if you need it

In the past two years we have closed eight of our older, smaller components and systems manufacturing plants. We have been forced to reduce the Intel workforce by 7200 people through attrition and layoffs. At the same time, we have invested $210 million in capital equipment during 1985-86, much of it for manufacturing. The result is a doubling of manufacturing productivity in the past two years due to automation, better methods and reduced staffing levels.

  • I sometimes wonder why people today (2020) are still surprised when automation takes over traditional jobs.

  • I wonder how many people were laid off due to automation vs financial difficulties? If there was no automation, what would have happened?

We concentrated on maintaining our momentum in research and development program. In 1986, we spent $228 million for R&D, equal to 18% of revenues, and an increase of $33 million over 1985. We also introduced 74 new products. Spared 1985s steep price erosion, these new products are in a position to deliver higher margins than older products.

  • When the going gets tough, go back to your core competency — Netflix tried 4 different strategies to increase revenue streams (selling old DVDs, ads, etc) — CEO Reed Hastings later commented that had they just focused on the recommendation system, they would have been fine

  • More products at higher margins = good

We phased out of the dynamic random access memory (DRAM) business in 1985, and decided to sell our bubble memory business in 1986. We had become a minor participant in the crowded, loss-plagued DRAM market; bubble memories had become a niche market that didn’t fit our long-term strategy. We moved into more promising areas such as microcommunications, application-specific integrated circuits (ASICS), personal computer add-in boards and parallel computers.

  • If something’s not doing well, stop it. As the saying goes, “Sometimes it’s necessary to cut off an arm to save your life.” If you’re in trouble, cut off the worst performers and least-aligned, and invest in things that do have a promising future.

While we decided to get out of two memory component businesses, we remained in a third—erasable, programmable, read-only memories (EPROMs). Intel invented the EPROM and has always been the technology and market leader. We decided to defend our industry-leading market share in the face of aggressive targeting by Japanese competitors.

  • If one of your bleeding arms is a market leader — consider staying. However, defending marketshare against a foreign Asian industry… why does this sound familiar?

The International Trade Commission recently upheld our complaint that the Japanese “dumped” (sold below cost) EPROMs. The U.S.-Japan Semiconductor Trade Agreement reached in 1986 should end this practice if enforced properly.

  • Hard to say if today’s competitors would stand by and abide by such laws… I have a feeling they’d find ways around

According to Dataquest, a market research firm, Intel gained EPROM market share in both 1985 and 1986. This was an expensive but necessary victory. EPROMs are our highest volume chips; we need them for manufacturing process technology development. Also, in anything resembling normal business conditions, we can be profitable in EPROMs.

  • Focus on the fundamentals — if the fundamentals work, power through it. Go back to unit economics. But remember — sometimes markets can stay inefficient longer than you can stay solvent.

We concentrated on strengthening relationships with our customers. We significantly improved delivery performance and started programs that enable customers to forego incoming inspection of Intel products, the latter made possible by our high product quality. As a result of these and other programs, customer satisfaction indicators have improved markedly.

  • Focus on your customers — make things easy for them, and they’ll be happy

Innovation has always been our strength; protecting those innovations is one of our highest priorities.

  • Defend your moat — see above about defending EPROMs

  • This should be the US’s motto… shame it’s not

As we move into 1987, we are especially pleased by the strength of Intel’s component and system-level product line.

  • Product-focused — not marketing/sales

We are very pleased to report that the military versions of the 286 and 386 were the first off-the-shelf chips admitted to the qualification process of the U.S. Government’s VHSIC (Very High Speed Integrated Circuit) program. In the past, admittance was limited to chips developed specifically for VHSIC.

  • Off-the-shelf is powerful… this is why DJI was able to succeed over all other drone companies and capture >70% of market share

  • I often forget how many of the OG tech companies were developed for/sold to the military… can’t wait for the Palantir S-1

During the past two years, we have also established a good position in the thriving market for personal computer add-in boards that permit users to add memory, functions or processing power to their systems

  • Personalization + Choice is always a plus for customers

Intel is well-positioned to grow at this point. Our product portfolio, manufacturing infrastructure and customer relationships are all better than they were two years ago.

  • Measure what matters — Intel’s core competencies are products, manufacturing their products, and selling those products

The year came to a close with fourth quarter orders and revenues substantially higher than the first quarter of 1986. If these trends can be maintained, Intel should return to levels of performance that will once again provide satisfaction to our shareholders and employees.

  • End on a positive note — cater to your shareholders and employees

  • Wording reads as “no promises, but be optimistic”

  • No excuses for horrible performance, discusses steps taken to improve, tone gets better over the course of the letter (deliver the bad news first — everything will seem better later lol)


If you’ve got any thoughts, questions, or feedback, please drop me a line - I would love to chat! You can find me on twitter at @kgao1412 or my email at kevin@12mv2.com.

Please DM or email me any time — to share non-obvious intel, views and correct or solicit mine. I appreciate your continued support and partnership :D.

All compilations here.