Sister Wood: Why did I abandon NVIDIA?

Author: Wang Yichuan/Boss Dai, source: Yuanchuan Research Institute

What is the difference between Cathie Wood and Mother Dragon? They all have a lot of titles, and then they all like to play with fire.

The titles on Sister Wood include: Female Brother Buffett, Musk’s Chief Iron Fan, Bull Market Queen, Mamma Cathie, Wall Street Leek Crusher, Money Tree Sister (돈나무 언니), The person who founded the company according to the Holy Spirit… The inn where the dragon mother was refused accommodation because she had too many titles, it was also difficult for her to stay.

Dragon Mother likes to ride on the black dragon to breathe fire in “Game of Thrones”, and Wood Sister also likes to sit on tech stocks and light fires on Wall Street. After founding Ark Investment in 2014, Sister Mu took “disruptive innovation” as a spear, forever bursting into tears in technology stocks, and the net worth curve of the fund was as exciting as an electrocardiogram.

But since this year, she seems to be afraid to play with fire, especially in the treatment of the hottest Nvidia.

After ChatGPT detonated the AI revolution, NVIDIA, which was struggling to digest inventory, turned around and became a popular fried chicken, up more than 220% so far in 2023. According to convention, NVIDIA should be the heavy stock and heart meat of the wooden sister. But on the contrary, Ark Invest has been reducing its holdings of Nvidia this year, and is desperately shouting to the outside world: Don’t just stare at Nvidia, there are many better than it. **

This is a “yesterday’s Britney, today’s Cow Lady” style plot. The relationship between Sister Wood and Nvidia can be traced back to the early days of Ark. In September 2013, Ark Invest’s first flagship product, ARK-K, was successfully fundraised, and Wood Sister bought NVIDIA at the first time. In 2015, based on the optimism of the two downstream fields of autonomous driving and deep learning, Sister Mu bought NVIDIA into a heavy position[2] 。

Therefore, if the performance of several ETFs under ARK is attributed (as of May 31, 2023), the contribution made by NVIDIA is very significant. NVIDIA is the fourth largest contributor to the performance of its flagship ARK-K, and the second and third largest contributors to ARK-W and ARK-Q. It is precisely because the past story is so sweet, and it seems that the operation of the wooden sister Chop Cang NVIDIA is so weird.

According to information revealed by Cathie Wood in various interviews, this pre-publicized breakup is related to another company: Cisco. **

Cisco Apocalypse

In the eyes of Sister Wood, Nvidia is copying Cisco’s bubble story that year.

After the birth of the Internet in the early 90s, U.S. stocks began to favor Internet-related companies. At this time, Cisco, a company that provides network hardware equipment, is considered to be the “shovel buyer” in the Internet era - no matter which Internet giant wins, Cisco is the one who benefits the most. This is very similar to NVIDIA’s current positioning.

Cisco has indeed fully benefited from the explosion of the Internet. From 1994 to 2000, Cisco’s revenue grew 10-fold, but due to investors’ preference for “certainty,” Cisco’s stock price rose 55-fold over the same period, far higher than the growth in revenue and profit over the same period. By the early 2000s, Cisco’s price-to-sales (P/S) ratio was 61 times.

The story that followed was familiar: the dot-com bubble burst, and Cisco’s stock fell more than 90 percent at one point. In the 23 years since, Cisco’s stock price has hovered repeatedly, never exceeding the bubble-era high. At the same time, Microsoft, Amazon, Oracle and other software Internet companies that have also “cut their knees” have hit new highs.

In Sister Wood’s view, in the early stages of the industrial revolution, companies that “sell shovels” were indeed more sought after, and Cisco was one of the best large companies to rise throughout the 90s. However, with the increase of Internet penetration, the software ecology built on hardware is more likely to give birth to super bull stocks.

Historically, Cisco’s revenue growth has slowed significantly since the Internet infrastructure in major countries around the world has been built (Intel is facing the same situation), and software stocks such as Microsoft have returned investors almost 20 times more than Cisco and Intel.

Comparison of cumulative returns of Cisco, Microsoft, and Intel this century (as of June 30, 2023)

To put it simply: pure hardware infrastructure companies benefit in the early stage, but space is limited, and investors should adjust the direction of fire in time.

ARK believes that in the era of artificial intelligence, every dollar of hardware spent can be exchanged for $8-21 in software revenue. Therefore, compared to the hardware-oriented NVIDIA, Sister Mu prefers software companies such as streaming platform Roku, video conferencing software Zoom, cloud communication company Twilio, and UiPath that can automate business processes.

But onlookers may have to ask: Even though the ceiling of software companies is higher, Cisco rose 55-fold between 1994 and 2000, and now that the era of artificial intelligence has just begun, will it be too early to sell Nvidia?

In this regard, Sister Mu also has a set of self-justifying reasons: in ARK’s research framework, once the price-to-sales ratio (P/S) of a hardware stock exceeds 25 times, it is a red flag. At the beginning of the year, NVIDIA’s price-to-sales ratio (TTM) was about 30 times.

Sister Wood even made an analogy between Cisco in 1994 and NVIDIA in 2023: Cisco in 1994 accounted for 2.5% of the S&P IT industry index and 0.2% of the S&P 500 index; In 2023, Nvidia’s share of the two indices is 4.7% and 2.8%, respectively - she believes that Nvidia’s transactions are currently too crowded.

Therefore, to summarize the view of Sister Mu, the reason for abandoning NVIDIA is twofold: one is that it is too expensive (> 25X P/S), and the other is that there are more worthy targets for investment (Tesla, software stocks). She used Cisco’s case to hint to everyone: ** Right now, just like that moment. **

There is actually another reason why Sister Mu is so nervous about “NVIDIA is very similar to Cisco”: the fund company that was heavily invested in Cisco did not end well. The company is called Janus.

Janus’ lessons

Goldman Sachs once described the dot-com bubble in 2000: "The burning of money by Internet companies is not a long-term risk for the industry. Morgan Stanley’s Mary Meeker reassured investors at the top of the mountain with a mantra: “Now is the time to be rational and reckless.” ”

A few months later, the bubble came to an abrupt end, hundreds of Internet companies went bankrupt, and the funds that bet on them came to the final judgment, including the most shining mutual fund of the 90s, Janus (two-faced).

Tom Bailey founded Janus in Denver in 1969, and after surviving the difficult period of the 70s, Janus earned more than 18% annualized throughout the 80s and began to be widely reported in the media.

Thomas H. Bailey

In 1984, founder Bailey sold 80 percent of the company to Kansas City Southern Industries (KSCI) for $14 million. After the sale, Bailey spent most of his time sightseeing, while the fund was handed over to the Wolong Phoenix: Jim Craig, 28, and Tom Marsico, 29.

Tom Marsico’s style is “high concentration + growth stocks”, which just met the market style match, and became a star in the industry in a very short time. He also received the Fund Manager of the Year Award and the World Fund Manager Award from Morningstar in 1989 and 1994.

Marsico and Craig replicated this style throughout the company, mass-producing star products. In addition, Janus has also put a lot of effort into marketing, with millions of dollars in product marketing dollars, and even public relations executives will call business reporters one by one to promote products.

Janus’ two brilliant architects of the 90s (Credit: The Wall Street Journal)

Throughout the ‘90s, Janus’ management grew from $3 billion to $318 billion. **At its peak in 2000, Janus’ new fundraising was even 30 percent of industry-wide (all mutual funds in the United States), and almost $1 billion poured into Janus’ account every day.

Although not as large as the established giants Fidelity and Vanguard, Janus is obviously more famous, and even Clinton handed over part of his IRA retirement account to Janus to manage.

From September 2000 to September 2002, five of Janus’ funds retreated more than 60% in two years, and it was only after a year of dot-com bubble bursting that Janus reluctantly began to rebalance[8] 。 In the end, customers with huge losses were redeemed and Janus plummeted.

In the words of the moment: Janus is also a group, and defeat is also a group. The five most popular funds of Janus have highly similar holdings and three achieved exactly the same results in 1998[10] 。 And one of the largest stocks in these funds is: Cisco. **

Not just overlapping positions, Janus’ holdings were too concentrated for a fund company with a size of more than $300 billion, and the company’s analysts at one point only had 26 people, covering only more than 500 companies. The concentration of positions makes Janus like a fish on a chopping block during a crash, unable to move.

Sister Wood has completely experienced the whole process of the Internet bubble, and is familiar with Janus’ “rising tall buildings, banquet guests, and building collapse”. So when she started Ark Invest and invested in the same growth stock strategy, her biggest fear was to draw an analogy between Ark and Janus.

The image of Sister Wood has always been that she dares to say and do, ignoring the eyes of others, but judging from her actions, she is afraid that history will repeat itself on her.

Long and short board of wood

Using Janus’ story to map ARK, there are indeed many things in common.

For example, it is also at its peak and has a reputation comparable to that of a rock star; For example, they also like high concentration and are keen to bet on disruptive technology companies; But ARK’s biggest highlight is that Janus or any other Wall Street firm has never had — proactive transparency.

One day in August 2012, while Sister Wood was alone in her 6,000-square-foot house, she suddenly felt the call of the Holy Spirit. After her epiphany, she decided to create a new type of asset management company in the age of social media and to appear in front of the world with transparency that Wall Street had never seen before.

For Wall Street, transparency and venture capital have always been two different concepts. Most fund managers in charge of large funds, in order to prevent sniping from others, avoid communicating current positions and views with the market as much as possible, except for quarterly statutory quarterly reports.

But Sister Wood did the opposite, producing podcasts, white papers, YouTube videos, newsletters and other content with the company’s support. Through this content, she promotes her image and her perspective on investing in the moment to millions of fans. ARK also discloses the day’s actions and the reasons for doing so in daily emails to investors.

By combining radical transparency with risky tech stock investing, Sister Wood sometimes comes more of a Reddit desperate Vat than a traditional institutional investor, despite having worked in the traditional asset management industry for more than 30 years.

Different from traditional institutions, Sister Mu has gained a large number of fans, but it also makes ARK’s investment research system focus too much on the grand narrative and ignore the granularity of research.

Taking NVIDIA as an example, Ark was successful in 2015 when it was qualitatively judged that GPUs would become digital infrastructure. But at the beginning of 2023, looking at NVIDIA too much from the perspective of “carving boats and seeking swords”, but misestimated NVIDIA’s profit elasticity and sold trillions of chip leaders.

In fact, many top subjective funds (including some Chinese investment institutions) can roughly predict the performance explosion in Q2 2023 by tracking the orders of NVIDIA customers in a timely manner and giving TSMC demand guidance. When they reduced their positions in Sister Mu, they were increasing their positions.

Although there are many problems in the granularity of investment research, it is undeniable that Wood Sister is changing the industry.

With the success of Wood Sister in 2020, BlackRock also released three “transparent funds” at the end of 20 that disclosed positions on a daily basis. In 21 years, Goldman Sachs and JPMorgan Chase also announced that they would issue “transparent funds”. The entire Wall Street, under the influence of Sister Wood, is trying to improve its own transparency as much as possible.

Wall Street has had impressive performance over the years, but only Sister Wood has become the top, partly because there are fewer and fewer fund managers with clear views and open-book operations like Sister Wood.

In the asset management industry with serious homogenization, how to make differentiation must not be enough based on temporary performance. Sister Mu provides another way of thinking: export ideas, create characters, and strengthen beliefs.

For individual investors, Sister Mu’s homework is worth referencing; For people in the asset management industry, it is more than a reference, many people from ** questioning the wooden sister, to understanding the wooden sister, and finally to imitating the wooden sister. **

In hindsight, Craig and others, the core of Janus, despite causing huge losses to customers, were still able to retreat from the bubble and make a lot of money with the management fees and performance commissions they had accrued.

So whether Sister Wood will copy Janus’ ups and downs, she may or may not care.