Here is the pdf file of the investor letter for the third quarter of 2020.

CIO's Market Outlook

Despite the stalemate over a new fiscal stimulus and an increasingly possible second wave of the pandemic, the U.S. economic recovery has been taking hold in the third quarter. The Atlanta Fed's GDPNow forecast, a close to "real-time" tracking of economic growth by incorporating a wide range of macroeconomic data, has shown the economy's strength. We present another high-frequency growth measure out of the New York Fed to avoid cherry-picking and believe it is not a stretch to draw the same conclusion.

Of course, what matters more is if the recovery would lose steam or, even worse, the growth would fall into a double-dip as we enter the final quarter with many additional uncertainties (the election, geopolitics, etc.). We cannot foresee tail events. Instead, we take comfort in the latest forward-looking economic indicators, such as ISM and the regional Feds' PMI new orders and the U.S. capital goods new orders, all showing reasonable optimism and resilience. A continued, decent economic recovery is our base case for the rest of the year.

Close

Our proprietary Esoterica Market Cycle Indicator (EMCI) depicts a relatively optimistic view of risk-taking, too. Incorporating the U.S. treasury yield curve, corporate credit spread, and the U.S. 10Y real yield, EMCI gauges market risk premium's cyclicality. When the yield curve steepens, the credit spread tightens, and the real yield stays low and stable, it indicates an excellent environment to take the risk, and vice versa. As we wrote in the second quarter's letter, the Federal Reserve's emergency liquidity facilities proved effective and created such a risk-on environment for risky assets.

The Federal Reserve's landmark policy shift towards "flexible average inflation target" in August could extend this investor-friendly cycle even further. Under the new approach, the Fed will respond only to "shortfalls" of employment from its maximum level and tolerate inflation running hot (above 2%) for some time. In other words, the Fed is determined to push up inflation by lowering rates for longer to stay more accommodative if needed. As such, the curve is likely to steepen as the inflation expectation rises. The real yield remains low while the credit spread tightens due to abundant liquidity. To be clear, whether the Fed could eventually achieve its inflation target is another matter. The odds are not on their side if history is any guide (think of Japan). However, markets will price that way until proven otherwise.

Close

Then here is the hard question: has the equity market gotten way ahead of itself? The conventional valuations of the U.S. equity market are near the record highs, if not already there (e.g., the U.S. equity market cap/ nominal GDP). In our opinion, first, valuation is a "relative" concept. SPX's current forward P/E of 26x with the real yield of -100bps does not carry the same weight as its 26x P/E with the real yield of 400bps in 2000. Second, since we entered the QE era, the Fed has been flooding the system with massive money, most of which flows into financial assets rather than the real economy. Equities might look expensive against the real economy (fundamentals), but not so much against money in the system (e.g., the U.S. equity market cap/ M2). That is the regime where we are living in, and it is our job to play along.

Close

Observations of Digital Economy

2020 is nothing but typical. It is surprising how much changes can happen in such a short time: nice-to-have things are turning into can't-live-without; developments that usually take years are now getting compressed into weeks; and for us following the semiconductor industry, it increasingly feels like a double major in export law and international relation is a requirement to do our job properly.

We constructed our portfolio around the transition to the digital economy, identifying the companies benefiting the most during the process. Over the past three months, many of these companies reported their first full quarter of operating results since the outbreak of COVID-19, which gave us a clearer picture of the impact on their businesses. We attempt to share with you our observation, grouped by the four pillars in our investment framework.

Pillar One: Semiconductors

Since investing in the semiconductor sector, we never imagined that we would have spent so much time figuring out export laws and geopolitics, etc. The day finally arrived. The tension between the U.S. and China shows no signs of easing, and the U.S. is increasingly weaponizing the semiconductor industry to attain more leverage. The new restriction against Huawei that went effective on September 15th limited semiconductor companies worldwide to work with Huawei if its product uses U.S. technology in any way or format, including equipment, software, or IP.

Given the highly globalized semiconductor supply chain, the new rule essentially applies to every semiconductor company on the planet. We have been explaining that the Huawei ban is not the end goal, but just the beginning. Over the past few weeks, we started to see more evidence: SMIC, the largest and most advanced semiconductor manufacturer in China, also became a target. The list could go on: any Chinese semiconductor company showing any promise of catching up to the current global leader would subject itself to the scrutiny. No wonder the Chinese government made the localization of technology, especially semiconductors, a strategic priority. The U.S. will, of course, try to stop them from the outset. This not-so-delicate conflict is likely to last for years and shape the industry landscape for decades to come.

However, this quarter's most significant development is Nvidia's attempt to acquire ARM for up to $40 billion. ARM licenses its CPU design to hundreds of other semiconductor companies, powering virtually every smartphone globally and billions of other connected devices. It has also made some early progress into the datacenter market. The grand vision of an Nvidia + ARM combination is to

1. proliferate Nvidia's AI product and IP into ARM's vast ecosystem of smartphone and connected devices,

2. leveraging Nvidia's strong position in datacenter GPU to accelerate ARM CPU's adoption in the datacenter, and

3. total control of AI and machine learning workload from every edge device to the cloud.

It is a brilliant plan on paper, but with two significant roadblocks. First, how will regulators worldwide react to such a combination, especially China? Second, how will other ARM licensees respond to the potentially uneven playing field given that many ARM licensees directly compete with Nvidia? If Nvidia pulled off this acquisition and executed it on the vision, this will go down the history as one of the most impactful technology acquisitions. That is a big if, though.

Close

While Nvidia is busy making its way into the datacenter, Intel keeps struggling. The Intel management finally admitted another delay in its 7nm manufacturing technology, giving competitors a broader window to attack its core datacenter CPU market. Intel has a contingency plan to rely on TSMC if it cannot get its manufacturing back on track, which further increased our confidence in TSMC. It is always good when your biggest competitor showed more confidence in your product than its own.

Pillar Two: Cloud

The migration to the cloud continued to accelerate, and this is particularly visible in companies that were slower in cloud transition in the past. If moving to the cloud used to be good-to-have for efficiency improvement, it turned into a cant-live-without utility after COVID-19. Case in point, during the second quarter, the edge computing provider Cloudflare saw "particular strength in Europe, industrial companies and small businesses. Those are not the first three segments when you think about cloud adoption, and yet, COVID has caused even those segments that traditionally are slow to change to adapt in order to survive."

Close

The hyperscale cloud providers kept investing in both infrastructure and personnel for the cloud aggressively. Alphabet CFO Ruth Porat stated clearly on the call, "while we expect the pace of headcount growth to decelerate somewhat in 2020, we're continuing to hire aggressively in priority areas like the cloud."

The convergence of 5G and cloud is moving fast from concept and trials to commercial deployment. Partnering with Verizon, AWS started to offer its mobile edge computing solution, AWS wavelength, to customers in Boston and Bay area in early August and expanded the available region to Atlanta, NYC, and Washington DC shortly after. Microsoft is also actively competing in the same market with partners like AT&T. Ironically, one of the early customers of AWS wavelength is the gaming studio Bethesda, which Microsoft acquired for $7.5 billion recently. It will be interesting to see how AWS, Microsoft, and Bethesda's relationship might take shape going forward.

Pillar Three: SaaS 

The publicly traded SaaS companies had a monster run in summer, with valuations reaching an unprecedented level. Suffice to say, expectations were high going into the earnings season. Most of the companies we followed delivered strong results and offered positive enough outlooks to sustain the valuation. However, we did start to see more evidence of the negative impact of the macro uncertainty.

Slack, for example, had been adding large customers at a steady pace before COVID-19. The number of customers with annual recurring revenue of $100K or more increased by at least 70 per quarter for the prior seven quarters, with revenue contribution rising from 39% to 49%. However, in the most recent quarter, the number of $100K+ customers only increased by 22, the lowest in years. The company saw 50 customers moving from the $100K-$150K bucket down to the $50K-100K bucket as companies reducing headcounts or stopping hiring.

Close

On the other hand, new SaaS IPOs like Snowflake attracted a lot of attention. Snowflake was so popular that even Berkshire Hathaway uncharacteristically took a stake before IPO. The company operates in the relatively obscure data warehouse market. While the value proposition of its product is clear, and the operating metrics point to a company with strong growth momentum, it also comes with a nose-bleeding price tag. The company generated $400 million in revenue in the twelve months ending July 2020, but trading at a market cap of over $70 billion. The set of assumptions to justify the price is likely both aggressive and narrow, leaving little room for error.

Pillar Four: Enabling Technology

While people have been embracing social distancing for safety concerns, they continue to have the same social needs but just fulfill them in a quite different way. Online content platforms like Netflix, Disney Plus, and Instagram added co-watching features to help users better interact with each other. Scener lets you start a co-watching party with friends while having a video conference at the same time. These new features are less about watching the movie but more about hanging out with your friends. What we learned is that social needs are human nature. External forces such as a pandemic cannot stop it. To take this one step further, people like to socialize in a specific context, most likely something they enjoy doing together, such as shopping, music festival, or an online game. We have many generic video communication solutions available today, Zoom, Google Hangout, and Microsoft Team, to name a few. But for many vertical applications, real-time video communication is still an under-penetrated feature. We expect more applications to add that feature and more social activities to happen online in the future.

As more purchasing activities move online, online payment and digital wallets become the utility that no one can afford living without. PayPal added 41 million new active users in the first six months of 2020 compared to 37 million in full-year 2019. The company sees the strong momentum to continue, anticipating a total 75 million increase in 2020. Square also witnessed its digital wallet product taking off, not only for P2P payment but also for spending, investing in the stock market, or even cryptocurrency. The Cash App's revenue increased by 360% in the most recent quarter and 140% if excluding bitcoin-related revenue.

Close

In a time of uncertainty, pessimists and cynics might sound smart, but it takes a calm mind and real courage to stay positive. If history is any guide, optimists are always right in the long run. Sticking to facts and data, not subscribing to narratives, has served us well. In this vein, we collected things that we learned from companies that excite us during the quarter and put them into a series of "What Caught Our Eye." We also regularly share our thoughts and observations on LinkedIn, Instagram, and newsletters. We hope you find them useful and look forward to hearing back from you.