It is an honor to provide testimony to the US-China Economic and Security Review Commission on the topic of outbound investment from the United States to China, its potential to create national security concerns, and ways in which to address these concerns in a balanced and effective manner. In this testimony, I provide:

• Descriptive data that show the United States is the primary overseas investor in China, mostly through venture capital, though the volume of these flows has declined substantially in recent years.
• An overview of the four key components of the executive order on regulating certain types of US investment to China that I believe are most important to maintain, primarily at the strategic design level.
• A review of three key questions/challenges in implementation that remain after reading the advanced notice of proposed rulemaking (ANPRM) that was released in August 2023, along with recommendations for how to address these challenges.

My core recommendations are as follows:

1. That the Commission should affirm the importance of maintaining any outbound regulation as a notification/prohibition regime rather than a screening apparatus.
2. That the Commission should endorse a sector-based approach to outbound investment regulation. List-based approaches, notably the NS-CMIC list, can be judiciously used to complement sector restrictions, but the bulk of the outbound regime should rest on narrow sectoral restrictions.
3. That the US Congress should consider providing a statutory basis for the NS-CMIC list and extending its reach to include non-public subsidiaries of NS-CMIC listed companies as well as to non-public companies that are determined to be part of China’s military-industrial complex.
4. That Congress should refrain from adding non-national security-related tests, such as supply chain diversity or local employment considerations, to any legislation related to outbound investment regulation.
5. That Congress recommend that Treasury’s final rule for implementing E.O. 14105 include an intangible benefits test to scope covered investments as described above and that any legislation regarding outbound regulation include the same provision.
6. That Congress recommend that Treasury’s final rule for implementing E.O. 14105 further clarify “routine intracompany actions,” and ensure that the rule does not allow for material expansion or operational pivots into covered activities. Any legislation regarding outbound regulation should include similar clarity.
7. That Congress, in addition to adopting recommendation 3, further modify the CMIC program to authorize the designation of Chinese entities beyond the current scope to include any Chinese entity operating in sectors important to US national security, as defined through a regulatory process. These sectors may be broader than the three sectors identified for the purpose of the current implementing rules for E.O. 14105 but should be relatively narrow and stable. A subset of the Critical and Emerging Technologies List (CETL) is a good place to start.

Level-setting the scale of US outbound investment to China

US outbound non-passive investment flows to China have declined substantially in recent years, likely due to policy changes in the United States as well as the Chinese Communist Party (CCP)’s crack down on Chinese tech companies.

Greenfield investment

According to available data, in recent years greenfield investment in China has declined dramatically – both from US investors and the rest of the world. Figure 1 comes from fdiMarkets, the pre-eminent data source for greenfield investment. This chart illustrates that greenfield investment from any foreign source – not just the United States – has declined from a peak in 2018 of roughly $120 billion to under $20 billion in 2022. Note that fdiMarkets uses announcement data rather than realized investment, so many FDI experts believe their numbers are likely to be a bit inflated. Clearly, global investors are avoiding greenfield investment in China, likely due to a mix of push and pull factors. US sources of greenfield investment totaled $8.69 billion in 2020.

Mergers and acquisitions (M&A), private equity (PE), and venture investment (VC)

Pitchbook data can provide more insight into non-greenfield US investment to China. As Figure 2 illustrates, Pitchbook data reports a high watermark of US investment in companies headquartered in mainland China, Hong Kong, or Macau in 2018. Investment volumes have declined every year since 2021; in 2023, US outbound investment to China was 30 percent of its 2021 value. To compare volumes across greenfield and these other forms of investment, US investment through M&A, PE, and VC was about three times as large as global greenfield FDI to China in 2022. A key feature of US investment in China is that a large portion of these flows happen through VC. However, an important caveat is that Pitchbook’s data relies on systematic web crawling and is unable to capture investments that have not been reported in regulatory filings, news articles, or press releases. Because US investors are not currently required to notify outbound investment – at least until E.O. 14105 is implemented – we simply do not know the full universe of US investments into China. Indeed, the reporting component of the E.O. will be very important to better understanding the full scale of US investment to China, and better allow policy makers to scope any regulation appropriately given the true volume of such investments. A costly and burdensome regulatory process to address a tiny concern is not in the long-term interest of the United States.

A deeper dive into the sectors that E.O. 14105 currently contemplates regulating suggests that US participation in these areas is quite small and almost exclusively concentrated in VC, as Figure 3 reports. Furthermore, this investment is almost entirely in the semiconductor industry; in 2020, investments in all other sectors amounted to only about $700 million.

Furthermore, the US is the most important global source of investment to China. Figures 4 and 5 present capital raised in China from investors headquartered in places other than China and the US in all industries (Figure 4) and in key national security technology industries (Figure 5). The United States supplies greater than half of all inward FDI to China and is even more dominant in the relatively small volumes of FDI into national security technology. Moreover, we do not see the United States’ relative position as major supplier of FDI to China diminishing, even as the US government has indicated it will place more restrictions on these kinds of flows.

Taken as a whole, these figures suggest that the size of U.S. investments in Chinese companies of concern is relatively small, but also that the United States is the primary global investor in these sectors. Even small deal values can generate national security concerns if US investors provide capital and expertise to a small set of key entities. However, the available data suggest that the scale of the concern – and particularly outside of the semiconductor industry – is modest. The data also suggest that an effective approach to this potential national security problem needs to address VC, since that is the dominate mode of US investor participation in these core sectors of concern.

Assessing E.O. 14105 – Addressing US investments in certain international security technologies and products in countries of concern – and its proposed rules

The executive order, for which we expect draft rules to be released within the next several weeks, is directionally an appropriate step forward in addressing national security concerns that arise from US investment in sensitive, national security-relevant technology in China. Four of the likely design features outlined in the related ANPRM that are important to maintain are:

1. A notification and prohibition regime rather than a case-by-case review. Initial policy conversations around an outbound regulation envisioned a screening process typically referred to as “reverse CFIUS.” However, the administrability of outbound case-by-case review would be much more complicated than is inbound. This is because The US government has better visibility into the capabilities and national security vulnerabilities of US businesses – which are the targets of inbound investments – than of such capabilities and vulnerabilities of businesses based in China. Additionally, the US government has more leverage over companies that wish to invest in its market – and therefore need its ongoing regulatory approval – than it has over companies that operate in foreign markets over which the US government does not enjoy regulatory authority. In the absence of such investigatory capability or compellence power, a screening mechanism would likely be very challenging to implement effectively. A notification and prohibition regime has the added benefit of providing industry and investors with bright lines about what investments are allowed and which are prohibited, which makes compliance and developing forward-looking business strategies more possible.
   
   Recommendation 1:  The Commission should affirm the importance of maintaining any outbound regulation as a notification/prohibition regime rather than a screening apparatus.
   
   
2. A (narrow) sector-based approach rather than an entity/list-based approach. Some in Congress have suggested that a sector-based approach is inadvisable because, while a sector-based prohibition regime would prevent US persons from investing in Chinese sectors of concern, it would not prevent investors from other countries from doing so. To make restrictions more biting, and to make them apply to investors beyond the United States, some have suggested a list-based approach in which the US government would regularly update a list of Chinese entities that are connected to the Chinese defense and/or surveillance industrial base and impose asset blocks on these entities through the Specially Designated Nationals (SDN) List.
   
   It is my view that this approach creates many problems. First, overuse of the SDN list generates substantial incentives for economic actors to further shift their activities out of the US dollar. While dollar dominance enjoys substantial persistence due to network effects, there is mounting evidence that country governments and related economic actors are increasingly finding ways to avoid US dollars – and thereby the reach of US financial sanctions – through cross border payments systems that do not use the dollar as an intermediary, and by shifting economic activity into other currencies. Dollar avoidance not only erodes the power of financial sanctions more generally, but it also makes it harder for the United States to track patterns in investment flows globally. This, in turn, makes enforcement of existing sanctions and disruption of money laundering activities more challenging. Thus, the unintended negative consequences of a list-based approach are high. Furthermore, the designation process is investigatively burdensome and exposes the US government to litigation. As a civil action, SDN packages need to provide substantial evidence that a designated entity is a national security threat, and designated persons can sue the US government to be removed from listing. Because of this legal structure, an SDN approach would be unable to address risks associated with US investments in Chinese entities working on more speculative but high consequence technologies. This is the exact type of national security concern – that is, early-stage investments and assistance through knowhow in pre-commercialization stages – that the US government identified as a gap in authorities because US export controls are not able to capture these kinds of emerging technologies well.
   
   Recommendation 2: The Commission should endorse a sector-based approach to outbound investment regulation. List-based approaches, notably the NS-CMIC list (more below), can be judiciously used to complement sector restrictions, but the bulk of the outbound regime should rest on narrow sectoral restrictions.
   
   
3. A focus on non-passive investments. There has been a flurry of policy entrepreneurship and innovation around addressing national security concerns related to US investments in Chinese military/surveillance technology. The current E.O. develops a regulatory structure around non-passive investment (colloquially, often referred to as “money plus,” meaning money that comes with control, knowhow, or other forms of more active engagement with the Chinese entity obtaining the investment. Others have argued that such an approach does not go far enough, and instead desire to completely remove all US money from the China market, including passive investment through securities. Indeed, proponents of a list-based approach argue that designations would stop flows of all kinds of US investments to listed entities, not just foreign direct investment (FDI) orVC. While preventing any US money from entering the China market may be symbolically satisfying, this kind of divestment is least likely to have an appreciable effect on decreasing China’s capacity for indigenous development and deployment of advanced technology for military and surveillance purposes. This is because money is fungible and the global equity market capitalization outside of the United States is roughly $62.8 trillion. Moreover, US share of global capital markets is projected to decline from about 42.5 percent today to about 27 percent in 2050.
   
   Thus, the bar for preventing such passive investments much be higher than restrictions on non-passive investments since the benefit-cost ratio of such actions is lower. Already, the non-SDN Chinese Military-Industrial Complex Companies (NS-CMIC) List allows the US government to prevent passive investment in designated entities that are identified as part of China’s military industrial complex, even if they are not state-owned. These authorities exist through E.O. 13959 and amendments. Currently, these restrictions only apply to relevant Chinese companies that are publicly traded.
   
   Recommendation 3: The US Congress should consider providing a statutory basis for the NS-CMIC list and extending its reach to include non-public subsidiaries of NS-CMIC listed companies as well as to non-public companies that are determined to be part of China’s military-industrial complex.
   
   
4. A focus on national security objectives rather than broader “economic security” or supply chain resilience concerns. Discussions on outbound investment regulations began in earnest after Senators Bob Casey and John Cornyn circulated a preliminary version of their draft legislation – the National Critical Capabilities Defense Act – in 2021. As the name implies, this early version of an outbound regulation concept was rooted not only in national security but also broader objectives around supply chain resilience. Over time, and through substantial and rigorous policy discussions, supply chain resilience components were eliminated from draft legislation on this matter and from the E.O. that was ultimately released. I view this as sound policy.
   
   As discussed in greater detail in a 2022 policy report I co-authored with Emily Kilcrease, supply chain concerns are largely due to features of the domestic and global economy that make supplier diversity and localized production commercially unviable. In that report, we recommended that the US government address supply chain resilience concerns through industrial policy and other actions that could incentivize re-shoring and friend-shoring to trusted suppliers. The Congress’ actions on supporting the semiconductor, EV, infrastructure, and other related industries and supply chains through legislative action such as the CHIPS and Science Act and the Inflation Reduction Act are far better able to address the underlying market challenges that have created supply chain fragilities in the first place.
   
   Moreover, by focusing squarely on national security and related technology, the United States is better able to act in a coordinated fashion with partners and allies. The Summer 2023 G-7 communique on economic resilience and economic security is indicative of the positive returns to such multilateral engagement, as leaders affirmed the legitimacy and importance of targeted outbound investment measures to protect “sensitive technologies from being used in ways that threaten international peace and security.” The European Commission’s January 2024 package on economic security, which includes monitoring and evaluation process for considering outbound controls, further illustrates the benefits of multilateral engagement around narrow, technology-related regulations on outbound investment.
   
   Recommendation 4: Congress should refrain from adding non-national security-related tests, such as supply chain diversity or local employment considerations, to any legislation related to outbound investment regulation.
   
   Outstanding design issues for an outbound investment regime
   At time of writing, the draft rules for the outbound E.O. have not yet been released. However, the advanced notice of proposed rulemaking surfaced several issues that a final rule will need to address.
   
   
5. Aligning covered investments more closely to the concept of intangible benefits. As discussed above, it is my assessment that it is correct to focus on non-passive investments. To do so well, the final rules will need to differentiate between purely passive investment and capital that confers some form of intangible benefit. Currently, there is no such test. Instead, the draft rules scope jurisdiction to investment that either rise above a control threshold or confer some form of special rights. But this rights-based approach may not be appropriate for a country with weak rule of law and shareholder protection such as China where control and influence are often exercised in more informal and extralegal ways. Such an approach may also lead to rule circumvention as investors interested in maintaining or expanding China presence simply shift their activities in China away from traditional FDI and VC structures and into uncovered forms of participation such as venture debt, business consulting, and/or university-to-university research collaborations. A final rule may better ensure that all relevant forms of intangible benefits are covered by constructing an intangible benefits test, in which a transaction would be covered if any one of the following conditions is met:
   
   • The US investor has a role in “substantive decision making” regarding the invested entity, leveraging this concept as it exists in the CFIUS context (see 31 CFR 800.245);
   • The US investor conducts one of a range of specified activities with respect to the invested entity, including the provision of management expertise;
   • The US investment conveys control of the invested entity to the US investor, with “control” set as a clearly defined percentage threshold; or
   • The US investment conveys a defined set of management or governance rights short of “control.”
     
     Recommendation 5: That Congress recommend that Treasury’s final rule for implementing E.O. 14105 include an intangible benefits test to scope covered investments as described above and that any legislation regarding outbound regulation include the same provision.
     
     
6. Coverage of growth transactions and operational pivots. Under the current text, it is unclear how the new outbound authorities will apply to follow-on transactions that are made after an initial investment, both in scenarios where the initial investment was made prior to the effective date of the new authorities and those made after the effective date. The ANRPM envisions exempting “routine intracompany actions,” providing an explicit exemption for the “intracompany transfer of funds from a US parent to a subsidiary located in a country of concern.” This text would allow for a US company to sustain an existing operation in a country of concern and to undertake the necessary financial transactions to do so. However, it also appears that this provision allows for a company to expand its investment without constraint if the funds to do so are made available via an intracompany transfer of funds.
   
   Material expansion of existing investments is likely inconsistent with the policy intent of the E.O. If so, the final rule should include clear standards for which intracompany transfers will be considered “routine” and therefore exempt from notifications or prohibitions and which will trigger new obligations under the notification/prohibition regime. The Chips Act guardrails set clear standards around material expansion, with respect to the investments in China of companies receiving Chips Act funding, that could be leveraged for the purposes of this rulemaking as well, at least for covered semiconductor investments.
   
   Similarly, the rule should anticipate scenarios in which a US person invests in a Chinese entity that is not a covered transaction at the time of investment, but, through a change in business strategy, pivots to operate in a covered national security technology or product. This is not a hypothetic exercise: for example, a US person could invest in a Chinese facial recognition software company that plans to develop its products for commercial use, but subsequently the Chinese entity could change its orientation to instead focus on selling its products to the Chinese government for surveillance use. This is of particular concern for cases in which the US person holds a non-controlling interest in the entity, and therefore cannot exert influence to prevent problematic changes to business plans. The final rule should clarify whether US persons are required to notify such investments and/or if the rule would require divestment if entity into which the US person invested subsequently operated in a prohibited national security technology or product.
   
   Recommendation 6: That Congress recommend that Treasury’s final rule for implementing E.O. 14105 further clarify “routine intracompany actions,” and ensure that the rule does not allow for material expansion or operational pivots into covered activities. Any legislation regarding outbound regulation should include similar clarity. 
   
   
7. Differences in corporate supply chain expansion vs. venture & technology startups. As outlined in the section above on trends in outbound investment from the United States to China, there are two types of investment that the US government is most worried could create national security concerns. First are corporate investments, usually made either to execute a global supply chain strategy or to serve the China market. The second are VCinvestments in early-stage companies operating in emerging technologies that may be used for military or surveillance purposes. E.O. 14105 attempts to address both kinds of investments in the same manner, but the differences in the incentives for and structure of corporate operational versus venture investments are substantial. In particular, venture investments are more speculative in that early-stage investment is made before it is clear what the commercial use for a nascent technology will be. Additionally, divesting from a venture capital position is very challenging as early-stage investment is all but frozen until an eventual liquidity event – usually after fifteen or more years of holding the investment position.
   
   Thus, venture investments present three key challenges to policy makers that are usually absent or less relevant to corporate operational investments:
   
   
   • The speculative nature of their technologies’ capabilities and use make it harder to draw narrow bright line distinctions between permissible and impermissible investments.
   • Funding structure flexibility provides VC investors with more opportunities to design their investments in ways that avoid generating reporting obligations or prohibition requirements.
   • Venture positions are illiquid over the medium term, making divestment more challenging.
   
   
   • Given this, it is advisable for the US government to consider additional ways in which forward guidance can help shape the commercial incentives of VC investors in ways that disincentivize early-stage investment in Chinese entities involved in the development of technology that may not be consider national security technology or products at the time of investment but that have a high likelihood of future national security implications. As outlined in greater detail in a report co-authored with Emily Kilcrease, it is advisable for the US government to undertake a set of actions designed to reshape investor expectations about the long-term financial payout to, and the reputational risks associated with, early-stage investments in technologies that are likely to develop into national security technologies or products.
     
     The goal of such actions is to better align investor incentives such that they are less willing to participate in particularly problematic start-ups, thus reducing the need for the US government to be prohibiting transactions or issuing divestment requirements in a heavy-handed manner. Already, the Congress has made steps in this direction by introducing legislation requiring the disclosure by previously exempted investors of their holdings in China and other adversarial jurisdictions. Additionally, the Congress can codify an expanded version of the NS-CMIC list to commit to preventing US persons from investing – even passively – in a set of designated Chinese entities that operate in a narrow set of particularly concerning national security technology areas. Doing so will communicate to VCs that their early-stage investments will not be rewarded by big payoffs during future liquidity events because US investors will be unable to participate in initial public offerings for these companies or private placements. Thus, the value of this approach is its deterrent effect on shifting the calculus of early-stage investors against participating in Chinese startups with technology that are likely to have use cases of particular concern for national security.
     
     Recommendation 7: That Congress, in addition to adopting recommendation 3, further modify the CMIC program to authorize the designation of Chinese entities beyond the current scope to include any Chinese entity operating in sectors important to US national security, as defined through a regulatory process. These sectors may be broader than the three sectors identified for the purpose of the current implementing rules for E.O. 14105 but should be relatively narrow and stable. A subset of the Critical and Emerging Technologies List (CETL) is a good place to start.
   
   
   

Image: Capitol Hill, located in Washington, D.C., is a historic neighborhood and the epicenter of the United States' legislative branch. Dominated by the iconic United States Capitol building, Capitol Hill is home to the U.S. Congress, including the Senate and the House of Representatives. The neighborhood features grand government buildings, including the Library of Congress and the Supreme Court. Beyond its political significance, Capitol Hill is known for its charming row houses, historic streets, and Eastern Market—a vibrant market offering fresh produce, artisan goods, and a lively atmosphere. It is both a symbol of American democracy and a residential and cultural hub in the nation's capital.