Jeff Wang on Public and Private Investing, AI and Sequoia Global Equities

In October 2024, Sequoia Capital Global Equities managing partner Jeff Wang speaks with Harry Stebbings about crossover investing, AI and building an investment firm. His company views and forecasts reflect his fund’s positions and interests at the time, not certain outcomes.

Wang joined the public-equities business in 2010, before its external launch, taking a substantial pay cut to build alongside Sequoia. He says the firm began with $50 million of internal capital and then managed around $9 billion, roughly two-thirds public and one-third private investments.

Its shorts are intended to express the other side of a long-term disruption thesis, rather than primarily to find fraud or valuation arbitrage. For example, enthusiasm for satellite connectivity might lead the team to consider implications for incumbent satellite or rural telecom businesses.

Wang says the portfolio uses relatively low leverage because technology shares are already volatile. Its thematic long and short positions can still suffer together when markets rotate away from growth, so he emphasizes a long-term investor base able to weather volatility.

He says the fund had one down year in approximately 15 years as of the interview. That is his historical account, not a verified performance record or assurance about future returns.

Knowing whether a decision is wrong or needs more time is difficult. Wang says the team integrates data scientists into investment discussions to test public-market theses with outside data and assess whether competition or demand has changed.

He cites Shopify as both a major winner and a regret: after COVID, his team assumed an elevated e-commerce trend would persist, then held that expectation too long as data weakened. He says the company continued growing, but the industry trajectory was lower than his model.

To reduce attachment, the firm re-underwrites positions quarterly. For controversial holdings, a different partner may build an independent devil’s-advocate case rather than relying only on the original investment team.

In 2016 Sequoia considered closing its global-equities business after parting with its original portfolio manager. Wang and the remaining team prepared a plan to concentrate on growth technology and co-investments, where they believed the wider Sequoia ecosystem gave them an advantage.

The team examined its prior results, removed unrelated sectors from the analysis and presented a new strategy to the partnership. Wang credits Michael Moritz with framing the decision as a smaller leap than the original launch because the team, capital base and infrastructure already existed.

Wang distinguishes active public-market investing from a continuation fund holding existing venture positions. Some holdings are Sequoia companies, but the fund also buys businesses such as Shopify that Sequoia had not invested in privately.

He prefers holding a company the team knows well over trading around small valuation differences. For a three-to-five-year thesis, he says a price 10% ahead of expectations may matter little, whereas a much larger advance can materially affect returns and justify trimming.

He acknowledges the team should have trimmed more during the 2021–22 period when some growth stocks ran ahead of fundamentals. Shopify is one example he discusses; the lesson is an investment judgment made with hindsight.

The crossover strategy also invests in late-stage private companies to build relationships and knowledge before they go public. Wang cites Nubank, where SCGE invested privately and at and after the IPO; he says it was a top-five position at the time.

He says knowledge of private competitors can inform public investments and vice versa. In the 2024 conversation he saw more opportunities in public markets, while suggesting that the exit of some temporary private-market investors could improve crossover opportunities.

Wang does not know whether private markets will overtake public markets. He points to public-market liquidity, access to capital and the visibility of a listing as reasons companies may still choose to go public.

He views continuation funds partly as a response to limited liquidity and is unsure how permanent they will be. He predicted the IPO market might reopen from late 2024 into 2025 as financial profiles and revenue predictability improved; that was a forecast at the recording date.

About one-third of the portfolio was private, more than his preferred steady-state range of 20–25%, partly because expected IPOs had taken longer. His preference reflects a balance between access to private opportunities and portfolio liquidity.

The public portfolio is concentrated: he describes roughly 15–20 long positions, with the top five representing around 35–40% of the portfolio. Shorts express additional conviction about disruption rather than mechanically targeting a particular net exposure.

Wang compares thematic research to finding the right island on a treasure map. Several companies can benefit from the same trend; the more serious mistake is choosing a theme that never develops. He estimates 70% of research effort goes to the theme and 30% to the company.

He recalls holding Twilio too long as competition increased. Margin pressure, acquisitions and executive departures could each be explained individually, but together warranted closer scrutiny. A fresh internal review helped the team assess the pattern more dispassionately.

He agrees in principle with being greedy when others are fearful, but says it is hard to practice. The firm tries not to react impulsively to sharp market moves and instead focuses on long-term evidence.

Fundraising was difficult initially because venture firms’ attempts to launch public funds had mixed histories. Existing Sequoia relationships helped, and Wang says a track record made it easier to attract limited partners.

SCGE’s structure is designed around long horizons: he describes investments held for three or more years, a three-year minimum capital base and incentives crystallized over three years. He believes monthly redemptions and annual bonuses can encourage shorter-term behavior.

He credits Sequoia leaders with giving each business investment discretion while sharing mentorship, brand and support. In his view, public equities, venture and regional markets require different expertise, and excessive central control can undermine both decisions and recruiting.

The trade-off of being embedded in Silicon Valley is exposure to its own echo chamber. Wang says technological shifts can occur quickly but still take longer than peak enthusiasm suggests; the firm invites outside investors to challenge its assumptions.

He expects call-center disruption from AI but rejects the simple claim that software companies will disappear immediately. Large enterprises change slowly, he argues, and early examples do not establish that every company can replace complex applications.

Wang expects some large technology platforms to benefit initially because they control distribution and data. In his view, Meta can apply AI to recommendations and ad matching, potentially improving the performance of its auction-based advertising business.

He contrasts that with Google search, where a user often states an intention directly. The fund was neither long nor short Google at the time, he says, but he saw more threats to its search business and predicted Meta could become a significant competitor through its distribution.

Wang estimates AI was already contributing about $15 billion of incremental EBIT to Meta through content and advertising effects. That is his estimate, not a disclosed or independently verified company result; the interview does not establish the calculation.

He agrees that substantial AI infrastructure spending ultimately requires returns from applications, but is more optimistic than some investors about the pace of monetization. He points to potential pricing and productivity gains while acknowledging results may arrive later than expected.

On model training, he says very large spending creates real overinvestment risk. He speculates that a small number of major foundation models may remain, but does not claim to know the eventual number or cost.

SCGE owns some AI hardware and semiconductor investments but also seeks application businesses with strong existing operations and potential AI upside. He cites ServiceNow’s possible use of AI for support-ticket resolution as a clear thesis, not a guaranteed return.

Wang calls NVIDIA’s valuation reasonable if demand and earnings keep growing, but stresses uncertainty about future hyperscaler spending and market share. He says the fund was long NVIDIA and short some server-hardware businesses it considered less defensible; those positions can change.

He has not invested in data-center deployment businesses merely because demand is strong. His concern is that new competition may weaken long-term economics even if near-term growth or share prices are attractive.

For European software and fintech companies, he views the US as a major expansion opportunity. On deglobalization, he says investing in China has become harder partly because he feels less able to understand some companies’ internal decisions; this is his assessment.

Wang remains optimistic about Chinese entrepreneurship and expects leading companies to expand globally. He is also optimistic about India, citing consumer and financial companies and improvements he sees at Zomato; these are historical investment views, not predictions of individual stock performance.

He praises Doug Leone’s commercial instincts and support for colleagues. Of Michael Moritz, he values talent judgment and communication, and says he learned to borrow complementary strengths from teammates rather than develop every skill alone.

Wang recalls Moritz challenging whether private equity had made him too focused on financial numbers. His eventual takeaway was to combine analysis with imagination about what founders and companies might become over time.

He credits Roelof Botha with focusing on team development, culture and patient improvement. Wang describes a leadership style attentive to placing people where they can succeed and comfortable with the process of getting better.

In a quick-fire round, he argues that preserving every possible option can become costly: at some point a person must choose a direction. He admires Altimeter’s ability to connect insights across public and private markets.

The fund’s difficult year was 2022, he says. Anonymous employee surveys showed morale falling with performance; his lesson is to stay even-keeled and improve the investment process rather than let market results dictate every decision.

Wang regrets not taking Google seriously enough when he encountered it early. He distinguishes being first from being early: in public markets, a few months can still matter, but enthusiasm must be matched by evidence.

He sees the scale of AI investment as unusual because large technology companies now generate enough cash to fund enormous capital spending. The amount ultimately justified by returns remains uncertain.

In a personal discussion, Wang says marriage and family are important to him and cites a book by Christian pastor Tim Keller. He acknowledges that people’s circumstances differ and frames the lesson around commitment and sacrifice.

If he were not investing, he says he could imagine building products used by many people. Stebbings says he might have become a filmmaker, drawn by the chance to tell people’s stories.