Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Glenn Profile picture
Glenn
@GlennLuk
Co-founder/BoD @HealthCareInc | Previously @Catalyte_io | VC/PE @Investcorp Technology Partners — Tech | Econ Development | Investing | China/APAC
17 subscribers
Oct 31 • 11 tweets • 4 min read
Free cash flow is a measure after capital expenditures and incorporates fluctuations in working capital.

Since founding, BYD's modus operandi has been to re-allocate every dollar of operating cashflow + as much capital as it can raise — as non-dilutively as possible — to support the needs of a rapidly growing business.

Frankly, it is financially illiterate to describe re-investment back into a growing business as "losses". Negative cashflow is a cashflow item and — especially if related to CapEx and working capital fluctuations (which I will address below) — is conceptually different from "losses" which is an income statement term. A better approach is to consider how much long-term capital the company has raised an compare it to the scale of operating capacity that capital has enabled.

We can look at this from BYD's latest balance sheet, which I have summarized here: Image
Oct 30 • 12 tweets • 4 min read
For all the flak about "lack of a social welfare safety net", China has one of the lowest pension/retirement ages in the world.

Further, it's hard to imagine that China — a "loud and proud" socialist country — not investing significantly into its social welfare programs in the coming decades, especially as it has officially crossed the "high income" threshold. The problem with the "China does not spend enough on social welfare" narrative-pushers is in part a data one.

A large part of its social welfare spending comes in the form of social transfers in kind (STIK) (~6-7% of GDP).

Oct 24 • 19 tweets • 5 min read
Jonathon highlights what I thought was the most interesting point out of the recent communique.

I tend to look at things from a company/sector perspective, and for me this represented the CCP's effort to adapt the vast administrative bureaucracy to align with the operational and realities of shifting sectoral priorities.

Allow me to explain below in this 🧵 Property and infrastructure development were two of the key economic development priorities from the mid-2000s to the early 2020s.

Both property and infrastructure (especially "traditional" infrastructure like highways and bridges) were highly localized in nature. Land is central to both efforts, and land use falls under the jurisdiction of local governments.

Thus, it made sense for executive power to be decentralized to the local governments: Beijing simply cannot effectively manage land development in Guizhou.
Oct 23 • 6 tweets • 2 min read
From 4th Plenum communique:

Just want to just emphasize here that Chinese policymakers still highlight domestic *demand* not *consumption* This is important because there is a group of people that insist on confusing/conflating demand with consumption in the China context.

These are meaningfully distinct terms: Consumption is just one component of demand, alongside gross capital formation. The distinction is driven by GDP accounting definitions.
Oct 22 • 15 tweets • 6 min read
I can see that folks are already starting to wildly misinterpreting what this chart says and this seems like another one of these Rorscarch tests on China.

Let's nip this in the bud: this is IP share of services exports, which comes from Balance of Payments accounting. That China does not license IP is not an "indictment", it's a statistical quirk that requires some deeper understanding of the BoP and how it maps against real-world trade and investment realities.

Oct 21 • 16 tweets • 6 min read
This was a complex/nuanced discussion on "overcapacity". Thanks for writing it @wstv_lizzi as it is an important topic.

It presented a number of interesting ideas which make sense on their own but I struggled to tie them together under a "grand unifying narrative" related to the "China Model".

The challenge of the "overcapacity" narrative is trying to use it to summarize "China Model" into a neat, compact narrative. But trying to summarize something as complex as China's economy into a neat model is exceedingly difficult.

(as an aside, the piece read like a writer struggling to force-fit an article within pre-defined narratives/framing set by an editor)

Two key problems I've found in the "overcapacity" debate that I'll go into more detail in this 🧵:

1⃣ Unclear/conflated definition of the term "overcapacity"

2⃣ As you drill down down from the macro/national level to individual sector level, you find many sector-specific idiosyncrasies that contradict core elements of "grand unifying" theme around "overcapacity". 1⃣ Defining "overcapacity" itself

"Overcapacity" has become a loaded word, especially when described in the context of the broad "China Model" in the current geopolitical environment.

In regular industrial/manufacturing usage, overcapacity is simply a state/condition where capacity utilization is below a certain "normal" threshold. This threshold may very by sector and different operating conditions.

Standard capacity utilization is defined not only by physical capital stock, but also by an active labor force operating on a normal shift schedule (typically 2 shifts per day, 5 days per week, or 80 hours / week).
Oct 15 • 20 tweets • 6 min read
If you read the official court documents (link in ALT), it is very clear that the underlying reason driving the eventual takeover of the company were the prospect of escalation in U.S. trade sanctions under the 50% rule, which was officially released on September 29th.Image A detailed timeline of the events described in the legal brief clearly show that the entire series of events were instigated by the addition of Wingtech, which indirectly owned/controlled 100% of Nexperia, to the Entity List. Image
Oct 14 • 10 tweets • 4 min read
No. The reason why it has a monopoly today is because China has:

(i) made significant technology and process advances that effectively isolate/mitigate the effects of the environmental damage — concentrated in the up/midstream mining and separation phases — on society, and

(ii) invested in human capital / specialized manufacturing equipment and optimized steps in the downstream processing stages, including deep integration with end-product manufacturing (e.g. permanent magnets which make up the bulk of use cases by economic value)

Whether simply ignorance or worse, inability to recognize — e.g. by implicitly attributing it on Chinese society simply having a higher tolerance for pollution through this type of moral grandstanding 👇 — is frankly one of the key reasons why minimal real progress has been made to address a strategic vulnerability that has been known for decades. I'd once again encourage folks to listen to this podcast from @twittwoods who has been studying the development of China's rare earths industry and was really the first one to clue me into just how much investment has bene made to raise environmental standards, especially since the mid-2010s.Image
Oct 14 • 48 tweets • 16 min read
It's always helpful to understand the "variant view" and I'd encourage you to read Alex's for his.

I DM'ed him why I thought this one was flawed, in supporting the prediction of a 2027-2030 crisis point.

Here are the key points:

1⃣ Systemic risk from the property and LGFV sector have been contained

2⃣ American MNCs make more money off China than vice versa

3⃣ There are more vulnerabilities beyond rare earths

4⃣ Assumption of stasis in China's efforts to catch-up in its areas of vulnerability (advanced chips, global financial system)

5⃣ Last but probably most significantly: ignoring what have actually been China's greatest vulnerabilities — dependence on fossil fuels and iron ore — and the rapid progress China is making to address this — which ultimately affects its geopolitical calculus / internal assessment of leverage. 1⃣ Systemic risk from the property and LGFV sector have been contained

This is the sub-topic where we have had the most back and forth on so I'll just link to previous threads.

Here is one where we were debating the amount of embedded NPLs in the system

Oct 9 • 4 tweets • 2 min read
"Measured in terms of primary energy"

This was obvious when you saw coal >20 PWh (total power demand in China is ~11 PWh). These figures include heat loss from combustion.

It's like accounting for heat loss from fusion energy produced at source (the sun) for solar/wind power. What's relevant is that coal proportion of electricity generation has fallen from a peak of ~79% in 2011 to ~54% in 2025 and will likely fall to <20% by 2035.

Oct 1 • 32 tweets • 9 min read
Professor Setser's contention here is that China's errors & omissions (E&O) are "implausibly" low.

He also believes that declining E&O in response to a change in statistical methodology that explicitly aims to address statistical mismeasurement doesn't make sense and instead offers his own speculative theory.

The former is wrong. The latter is absurd. First some background. There are two categories of E&O:

▪️ Statistical mismeasurement, reporting errors, timing mismatches

▪️ Real capital inflows/outflows that purposely evade official reporting for various reasons (e.g. illicit flows, tax evasion, trade misreporting etc.)
Sep 27 • 16 tweets • 6 min read
I hear over and over again how there are "100-150" Chinese EV firms and "only a handful are profitable".

This is BS and highly misleading.

The people who cite these figures can never seem to name even a fraction of these firms ...

... because they are mindless parrots who are unwilling or incapable of doing their own work to check out basic, factual claims or provide highly relevant context.https://www.ft.com/content/0b32d65b-6963-4e07-bb75-648ff8652a55 This number (which tends to range from 100 to 150) is actually derived from brands — ostensibly pulled out of the CAAM** database. Depending on how high the number is, it could include defunct, retired brands.

It is definitely not "firms".

Just like most large foreign automakers, Chinese ones often have multiple brands under the holdco umbrellas, so they too can target differentiated consumer market segments.

e.g. GM has Buick, Cadillac, Chevrolet and GMC. Stellantis has Alfa Romeo, Chrysler, Citroen, Dodge, Fiat, Jeep, Lancia, Maserati, Peugeot and others.

This distinction is relevant here because — at least as non-insiders — we can only effectively observe profitability at the firm level, not the brand level.

** China Association of Automobile Manufacturers, the standard industry body.Image
Image
Sep 26 • 8 tweets • 3 min read
Historians will look back and identify the Meng arrest in 2018 as a critical turning point by making it "personal" for Huawei, which gets first pick on elite technical talent across multiple engineering disciplines in the world's largest STEM workforce. Up to this point, China's semiconductor progress was halting, at best.

Reality is that (i) SMIC is mediocre and (ii) there was very little financial incentive for the domestic chip ecosystem to move off Western SME and Taiwan foundries.

Sep 20 • 24 tweets • 7 min read
"Involution" (卷) is a complex word.

Perhaps not surprisingly, a certain group of thinktank-adjacent pundits are once again trying to co-opt the meaning of a Chinese word to fit into pre-existing, oversimplified narratives ...

vs.

... really trying to understand its nuances and how Chinese policymakers think — which will help one predict how they will act. Modern Chinese policymakers believe in the idea that competition drives innovation and productivity improvement that sits at the core of economic development.

Too little competition is bad. Hence Chinese policymakers' heavy anti-monopoly bent and allergic reaction to rent-seeking behavior in the private sector.

But too much competition can also be bad if it leads to market behavior that has negative societal externalities.

The goal for policymakers is to find the right balance. And every sector is different, so there is not a one-size-fits-all approach.
Sep 12 • 22 tweets • 8 min read
China's Finance Ministry reviews the 14th 5YP (2021-25) and looks ahead to the 15th (2026-2030).

Source transcript here: mof.gov.cn/zhengwuxinxi/c… x.com/Sino_Market/st… Fiscal revenue grew ~19% in the 2021-25 period compared to the 2016-2020 period (~3.5% nominal growth).

Note: IIRC fiscal revenue excludes categories like land sales, which are part of an auxiliary budget. Image
Aug 31 • 5 tweets • 2 min read
Scholars once worried that China's gender ratio imbalance would lead to a generation of surplus men, fueling crime, chaos, and even war.

What we got instead was ... this. Certain China watchers busy penning missives on how this Paw-temkin Village is merely the latest example of the country's addiction to construction, penchant for capital misallocation and "politically entrenched elites" blocking efforts at structural reforms.
Aug 24 • 10 tweets • 3 min read
Productivity is what ultimately drives per-capita economic growth and increases in living standards over the long run. This concept is one of the pillars of developmental economics.

I’ve come to realize one of the the fundamental issues with Pettis/Setser economic framing is misplaced reliance on accounting identities with little to no consideration of productivity effects.

To wit: nowhere in this thread is there any mention or consideration of how this sectoral shift impacts productivity. In the short to medium run, there can certainly be supply-demand disequilibrium where “weak demand” is an issue.

e.g. in this 🧵 from a year ago I tried to quantify the headwinds from reverse wealth effect impact of the policy-driven pivot away from real estate to manufacturing since 2020 and how they could offset wage growth driven by underlying productivity growth enabled by sectoral shift.
Aug 24 • 22 tweets • 7 min read
In January, I speculated how "the most impactful outcome from DeepSeek's rise may ultimately be closer collaboration with Huawei and other chip designers".

We now have direct evidence of this collaboration, with potential standardization around UE8M0 the first major tangible result.

While some may dismiss this as technical or esoteric jargon relevant only to AI, computer science, or math enthusiasts ... I will try to explain here in plain language some key market and geopolitical implications of this development. First I want to acknowledge others who are much closer to DeepSeek and AI for both raising, highlighting and explaining these recent developments, particularly @teortaxesTex @zephyr_z9 and @Compute_King

In this 🧵 I am merely synthesizing the insights and knowledge gained from following their timelines and trying to add value by layering on market and geopolitical insights.
Aug 21 • 11 tweets • 4 min read
About a year ago, ahead of "Two Sessions", I performed a deep dive into China's economic transition (away from real estate).

Five years since "Three Red Lines", China is now entering the "late innings" of that transition.

▪️ Exogenous macro risks ↘️
▪️ Real estate ↔️ and no longer a major headwind.
▪️ "Animal spirits" ↗️ Exogeneous macro risks are lower than a year ago:

▪️ Trade war-related risk/uncertainty largely cleared w/ better-than-expected outcomes.

▪️ China progressing in broad tech war. It is decisiveley leading renewables & EVs + closed gap in AI & chips.

Aug 12 • 5 tweets • 2 min read
> “The state cannot allocate capital more efficiently than the market.”

An oft-repeated axiom chanted like a religious mantra and accepted by many as a universal truth.

But one that can be easily debunked with a straightforward contra-example from one of the most capital-intensive industries of them all: passenger rail. China Railway (SOE) vs. Brightline (private)

CR HSR:
▪️ 48,000 km of greenfield track, predominantly elevated on viaducts
▪️ Serves 3.6B passengers annually
▪️ ¥550B of revenue on ¥5T of capital investment (9 years revenue payback)
▪️ 42 fatalities over 17+ years and 23B passenger rides

Brightline Florida:
▪️ 376 km of refurbished at-grade track
▪️ Serves 2.8M passengers per year
▪️ $187M revenue on at least $5.5B capital investment (29 years revenue payback)
▪️ Caused 182 fatalities in two-plus years of operation (hint: maybe you shouldn’t run fast trains over at-grade crossings).

Did the private company really do a better job allocating capital here? (rhetorical)
Aug 9 • 12 tweets • 4 min read
Re-invigoration of biking culture in China and the relentless expansion of dedicated biking lanes today can really be traced to the invention and proliferation of dockless bike-share systems starting around a decade ago. Like many emerging industries in China, dockless bike-sharing got off to a chaotic start but within a few years settled into a stable equilibrium.