Tomorrow’s Titans 2023: Hai Wei

Finding extraordinary and world-shaking equities

The Hedge Fund Journal
Originally published on 01 August 2023
  • Hai Wei, Founder and Chief Investment Officer, Jungle Gene, Tokyo

Full Report

Tomorrow’s Titans 2023

Jungle Gene’s concentrated and high conviction long biased equity strategy annualized at 17% over five years to December 2022, meeting its target of beating the S&P 500 index, and making profits in all calendar years – bar a small loss of 1% in 2018. The strategy, which is designed for longer term investors, has accomplished this with high confidence in the margin of safety of long-term valuations over the whole life cycle of companies it owned. Average holding periods are two to three years and portfolio turnover is low because very few stocks satisfy the manager’s demanding criteria for long term valuation metrics, which quantitatively include being able to estimate ten year and perpetual cashflows. “We cannot calculate a stable long-term valuation for most companies because we do not have confidence in their future. We need to know the company and the industry well. They need wonderful, extraordinary, amazing and surprising products,” says Jungle Gene founder and CIO, Hai Wei, whose investment career started in 2005, initially in economic research and currency trading at China’s fifth largest bank, Bank of Communications, where he worked for nine years including three in New York. He then traded his personal account for seven years from 2011 before setting up the fund in 2018.

In the era of information explosion, ordinary or indistinguishable products cannot be remembered, and extraordinary and world-shaking products are needed.

Hai Wei, Founder and Chief Investment Officer, Jungle Gene, Tokyo

Qualitative criteria

Qualitatively, Jungle Gene’s criteria for stock picks also demand a personal love of the company’s products and services. “We use the products and services of the companies we invest in. We can see the differences and feel they pay more attention to design and user experience with more advanced technology,” explains Wei.

Beyond valuations, the five key criteria sought from companies are: a sustainable and leading technological advantage; elegant and refined design; a wonderful user experience; a corporate culture fostering the emergence of creativity, and accurate judgment instincts from management. These factors partly reflect the zeitgeist of the 21st century: “In the era of information explosion, ordinary or indistinguishable products cannot be remembered, and extraordinary and world-shaking products are needed,” says Wei.

Banks and insurance are generally avoided due to opacity and other companies are eschewed because they lack competitive edge. “Commodity producers such as oil, metals and mining are avoided because they just make raw materials with no added value and their products will be replaced by green technology. ESG can be a secondary criterion. We seek human value added in products such as design, movie making, chipmaking and defence, which cannot be replaced,” says Wei.

This also provides some comfort over valuation target ranges, which can expand or contract over time. “When we meet unknowns, we need to expand the margin of safety and seek deeper discounts,” points out Wei. But some uncertainties are too great for any valuation discount: “We avoid Chinese technology due to regulatory risk which could lead to excessive uncertainties. The fund does not own any Chinese companies,” confirms Wei.

AI-enabled investment process

Artificial intelligence, and artificial generative intelligence, which have been driving forward the US equity market in 2023, are examples of technological advantages which feed into portfolio ideas and expedite Jungle Gene’s own analytical process. A technology specialist has developed ways to make the workflows more efficient in terms of data gathering and productivity. The manager also uses a large language model (LLM) for analysing and translating news articles, which increases coverage of macro, industry and company news. “AI is a real game changer for the world, and we are using AI and our own tools toextract data and establish databases for analysis,” says Wei. The team of seven people contains six in investment research and one in administration, mostly in Shanghai, but Wei, who oversees most steps and stages of the process, moved to Tokyo in 2023.

AI and tech stocks

Jungle Gene compares artificial generative intelligence to nuclear fusion in terms of its major technological impact. Wei reckons that AI could be a big threat to established technology firms including Apple and Google and even Microsoft, notwithstanding its OpenAI unit. “When OpenAI makes enough earnings, it will become independent, and all of its shares then revert to the executing managers. OpenAI owns the technology and Microsoft cannot get the core technology. The OpenAI CEO interview in The New York Times shows that it expects to capture most of the market when AGI comes to fruition. This will be an earthquake in the technology industry”.

Anticipating AI implications is already leading to some portfolio sales. Wei bought Apple for its “wonderful design, user experience, ecosystem synergies, high software and hardware quality standard,” but he sold a large part of the Apple holdings due to the emergence of ChatGPT. Apple’s TV series Extrapolations expressed Apple’s vision of the future technology ecosystem. Apple understands AI is the holy grail of that ecosystem. But its AI technology is behind OpenAI. Our benchmark assumption is that Apple may not be able to gain the advantages of its previous comprehensive ecosystem in the new AI era, but it can become a very good supplier in terms of hardware and user experience. The release of Vision Pro headset proved its outstanding ability in this field. It could be a potential collaborator in the tech era possibly dominated by OpenAI, akin to Mercedes and BMW. But this is a far cry from its previous overall advantage in both hardware and software.”

AI is only one of many threats to established tech behemoths. Jungle Gene bought Google because, “We like YouTube, which is growing fast. We sold it because other Google products are replaceable, and the company is large and bureaucratic and cannot work out which part of its business is most important until competitors make that clear. And Google’s investment success rate has been low”.

Free offerings from new entrants can also lead to exits: “We bought Dolby because it is a leading technology in audio and video. We sold it because Samsung has started a video standard which is free, even if it is not as good as Dolby”.

Yet some technology firms are well placed to thrive in this climate. Dutch lithography machine maker, ASML, is a good example of a firm with a near monopoly that keeps investing in R&D to advance technology and keep their leading position in the industry. “Japanese companies, Nikon and Canon, also make lithography machines, but they are not as good. ASML is the key supplier to leading chipmakers Intel, Samsung and TSMC and it has no real competitors,” says Wei.


Wei expects that creative, fashion and media companies with multi-dimensional products will not be drastically impacted by AI. In the fashion sector – the most profitable one so far for the strategy – Jungle Gene has a huge database monitoring leading fashion companies and collecting data on new product releases, including images, and reviews to predict future fashion trends. This mainly focuses on luxury and high-end brands such as LVMH, Kering, Adidas and Nike, which are more innovative, rather than fast fashion. 

But some active management is still needed to stay on top of the winners. “We bought Hermes because Chinese customers obsessively like the brand. We sold it because we cannot identify design advantages versus other luxury companies, based on our fashion products rating system. We bought Adidas and Muji due to their distinctive design and mass market appeal. We sold them due to lower brand loyalty.”

E-commerce is not a new story, but it could threaten some established retailers: “We bought Costco as it is the best wholesale grocery store but sold it because its slow reaction to e-commerce means it may not survive 50 years”.


Similarly, timely data is collected on media companies in film, TV and gaming to find the best ones, based on big data ratings. The fund has owned Disney and Time Warner, and monitors firms such as Netflix. AI is also a threat to some media companies. “We bought News Corp because we like The Wall Street Journal. It grows at a stable rate. We sold it because the WSJ is a limited percentage of revenues. Algorithm-based news platforms like Apple News threaten the editorial edge at leading news agencies.” 

But media outlets with a truly differentiated offering may survive AI: “We bought The New York Times because it grows as the population becomes more educated. We sold it for the same reasons as News Corp, but bought it back due to its screening offering, which together with in-depth reporting develops customer loyalty. This may defeat algorithm-recommended news platforms and help NYT become a direct source news and opinion boutique bypassing the aggregators”.


Competitive dynamics can be a moving target: “We bought TripAdvisor due to a great experience on attraction rating. We sold it because it is too easy for competitors like Google and to copy. We bought Booking due to a great experience in booking hotels and reading reviews. We ultimately sold it for the same reason as TripAdvisor: because competitors are catching up and Google entered the market”.

Selling discipline

In summary, Jungle Gene is disciplined on selling companies facing threats. “No company keeps its edge forever, and it is also vital to calculate when their competitive advantage vanishes. Every company will go down at some stage after its peak. We look at how long a company can grow and maintain its competitive advantage within its own sector and across sectors. “Generally speaking, we sell stocks out for the reason of unclear or negative long-term outlook for the companies. We ask ourselves “Will these companies’ growth last 30 years, 50 years or 80 years?” says Wei. “We are not sure if Airbus will survive the next industry revolution,” he adds.

Indeed, the true motivation to make investment his vocation reaches back much further into Chinese history. “Commercial battles are like the era of Warring States and Three Kingdoms, where feudal Lords were fighting each other for a very long era. Companies in the long term are just like kingdoms, and their pattern of prosperity and decline matches history, because human nature is constant. The Tang Dynasty historian, Wu Jing, said “Take history as a mirror, and one can understand the rise and fall, in his book Political Affairs overview in Zhengguan Era,” explains Wei.

Hedging in 2022

A recent video update revealed that News Corp and Google were sold in 2018; TripAdvisor in 2019; Hermes and Dolby in 2020; Goldman Sachs,, Airbus and Costco in 2021; Twitter, Adidas, and Muji in 2022; and The New York Times and a large part of Apple in 2023.

Sometimes selling out of individual firms is not enough, and the macro backdrop requires more active hedging to address a market-wide shift in valuations. 

There are typically 10-15 stocks in the long book while index hedging is carried out on an opportunistic basis, including in 2022. Based on accurately predicting interest rate rises, the fund had a gross short of between 130% and 150% in 2022, peaking at net short exposure between 50% and 70%, until around July and August. Cash could also be held as a means of defence. “Historically in 2018 the Japanese Yen was used as a hedge, but it did not work as well as equity futures. There were less hedges in 2018 because we judged that the trade war was difficult to predict,” recalls Wei. The strategy has the flexibility to trade currencies, bonds and commodities, but in practice invests mainly in equities. 

The manager generally does not attempt to predict short term equity market movements or those in macroeconomic variables such as inflation, interest rates and GDP growth rates, and only adjusts macro positioning when markets are judged to be at extreme levels or apparent policy mistakes are made by central banks or governments like in 2021. A version of the Shiller Model of long-term risk premia, modified with some innovative proprietary cross paradigm adjustments, is used to determine when market emotions are at extremes. They were extended for much of 2022 but are currently classified as stable. Hence as of April 2023, the fund was running at a net long of 90% and has no shorts.

The investor base for the Cayman fund, which is listed on Bloomberg, includes friends and family and Chinese offshore assets. The firm is audited by MHA and Apex is the administrator.