The VIE Trap: When A Low Price-To-Book Protects Nothing

← All Articles

Never Trust a VIE

If you run all the US-listed equities through Benjamin Graham's Enterprising Investor criteria a strange thing happens at the bottom of the price spectrum. The cheapest companies by the numbers are often the ones you should trust the least.

Take three from the May GrahamGrade screen. FinVolution (FINV) trades at a P/E of 3.7 and a price-to-book of 0.53. Weibo (WB) sits at a P/E of 5.0 and a P/B of 0.52. Trip.com (TCOM) carries a P/E of 7.5. On paper, these look like the kind of deep-value setups Graham built his career on. A sub-4 earnings multiple and a stock trading at roughly half of stated book value is the definition of a bargain bin find.

All three failed the screen anyway. Not because the math was wrong, but because the assets behind those multiples don't actually belong to the people buying the shares. Every one of these companies operates through a variable interest entity (VIE), and that alone is enough to override every attractive number we discussed above.

VIE, Say What?

When you buy shares of FinVolution, Weibo, or Trip.com, you are not buying a piece of a Chinese company. You are buying shares of a shell company, usually incorporated in the Cayman Islands. That Cayman shell company does not own the business. Chinese law restricts or forbids foreign ownership in sectors like internet content, fintech, and online services, so the operating business stays in Chinese hands.

So how does the shell connect to the profits? Through a stack of contracts. The Cayman entity signs a web of agreements (management service fees, loan agreements, equity pledge arrangements) with the actual Chinese operating company. These contracts are designed to route the economics of the business up to the shell, and ultimately to you, without ever transferring legal ownership. You can see the margin of safety deteriorating as you read, right?

That is the VIE. It is an accounting and contractual workaround that lets a foreign-listed shell consolidate the financials of a business it does not legally own. When you see Trip.com's revenue and earnings on a 10-K, you are looking at numbers pulled in through these contracts, not through equity you hold.

What Graham Actually Required

Graham's entire framework rests on one idea: you are buying a claim on real assets and real earnings, at a price low enough that those assets protect you if you are wrong.

That is what margin of safety means. It is not "the stock is cheap." It is "even if the business stumbles, there is enough tangible value behind my shares that I am unlikely to suffer permanent loss." Net current asset value, book value, tangible book value, the Graham Number itself, every one of these is a measure of what you can claim if things go badly.

The operative word is claim. A margin of safety you cannot enforce is not a margin of safety. It is a number on a page. Graham was writing about companies where, in a worst case, a shareholder had a real legal interest in the assets, the inventory, the cash, etc... The protection was enforceable.

Why a VIE means you don't own the business

Now connect the two ideas, and the problem becomes fairly obvious.

In a VIE, your claim on the underlying Chinese assets runs entirely through contracts, and those contracts depend on a Chinese legal system that has never reliably enforced them in favor of foreign shareholders. If Beijing decides the VIE structure violates foreign-ownership rules, the contracts can be invalidated. There is no equity to fall back on, because you never held equity in the operating business. You held equity in a shell company whose only asset is a stack of paper agreements that can be voided on a whim.

So, when Weibo trades at 0.52x book, the honest question is not "how cheap is that?" It is "0.52x book of what, exactly?" If the assets backing that book value cannot be legally claimed by the people who own the shares, the discount is meaningless. A low P/B is only protective if you can enforce the claim it represents. Strip out enforceability and you are left with a cheap-looking number attached to nothing you can hold.

Three Real World Examples

FinVolution: a textbook bargain on top of a structural fault

FinVolution is the most striking of the three because it screens so well. P/E of 3.7. P/B of 0.53. A current ratio above 5. A dividend yield near 6% sustained for eight years and positive earnings growth. If you handed those metrics to a value investor without telling them the country, they would call it a screaming bargain.

Then you read the structure. FinVolution is a Cayman Islands holding company sitting on top of a wholly foreign-owned enterprise and a VIE that holds the actual PRC lending operations. You do not own the lending business. You own contracts that point at it. And the Chinese online lending sector has already lived through existential regulatory intervention since 2020, which means the worst-case scenario here is not hypothetical. It has a precedent.

The earnings flow through contractual fees and intercompany transfers that are difficult to verify independently. Layer on US delisting exposure for China-based ADSs, and the margin-of-safety case collapses at the structural level no matter how good the multiples look. That is why it failed.

Weibo: a cheap multiple on a declining business

Weibo adds a second problem on top of the VIE. The business itself is shrinking. Monthly active users fell from 598 million at the end of 2023 to 567 million at the end of 2025. Revenue has been essentially flat near $1.76 billion for three straight years. Short-video competition (Douyin chief among them) is eating the engagement that drives ad revenue.

So even setting aside the structure, you have a no-growth business. The structure is what really puts the nail in the coffin though. Shareholders hold no direct equity in the PRC operating entities. The 0.52x book value looks like protection, but those assets are reachable only through contracts Chinese courts have not consistently upheld. Dividends require regulatory clearance to leave the country and have already been cut from $0.82 to $0.61 per ADS. The cheap multiple is real. The protection behind it is not.

Trip.com: the healthiest business, same disqualifying structure

Trip.com is the one that really proves the rule. By every operating measure it is the strongest of the three. The travel business is growing, earnings are real, and at a P/E of 7.5 against a P/B of 1.38 the headline valuation still looks reasonable for a market leader. If the structure were sound, this would be a name worth a serious look.

Trip.com runs the same VIE arrangement as the other two: a Cayman Islands holding company that does not directly own the PRC operating entities, with all China-generated cash reaching shareholders only through a multi-step intercompany mechanism subject to withholding taxes and regulatory approval. The detail that makes it concrete is the cash itself. No subsidiary paid any dividend or distribution to the Cayman parent in 2023, 2024, or 2025. Money reached the Hong Kong intermediate holding company and stopped there. Add a January 2026 SAMR investigation and a March 2026 US securities class action, and the near-term legal picture is no cleaner than the others.

So Trip.com fails for exactly the same reason FINV and WB fail, and its case is the most useful of the three precisely because the business at least looks good. A healthy, growing operating company does not make a contractual claim enforceable. You can run the best hotel-booking platform in China and it changes nothing about whether a foreign shareholder can fall back on the assets if the structure is challenged. The quality of the business and the enforceability of your claim are two different questions, and Graham's framework only answers yes to the second one.

Discipline, discipline, discipline

A low price-to-book ratio is a measure of how much asset value sits behind each share. The ratio assumes one thing it never states out loud: you can actually claim those assets. The whole protective logic of value investing depends on enforceability. The moment your claim runs through contracts a hostile regulator can void, the discount stops being a margin of safety.

Cheap is not the same as protected. Graham's framework was never about finding the lowest multiple. It was about finding the lowest multiple attached to assets you could genuinely fall back on. VIE-structured Chinese ADSs break that link, which is why a P/E of 3.7 and half of book value were not enough to clear the screen.

---