The Law of One Price: In Theory vs. Reality
The law of one price is a foundational axiom asserting that two assets representing identical cash flow claims and identical risk profiles should trade at the exact same price. When this law is systematically violated, it exposes the mechanical, regulatory, and structural frictions governing actual market operations.
The Ultimate Laboratory: This dynamic is most prominently observed in the Chinese equity market, where companies simultaneously list “A-shares” on mainland exchanges (Shanghai/Shenzhen) and “H-shares” in Hong Kong. Despite identical dividend entitlements, A-shares historically trade at a massive, volatile premium to H-shares.
1. Foundations of the Liquidity Premium
In frictionless models like the classic CAPM, return is determined entirely by market risk. However, real-world markets have search costs, inventory costs, and asymmetric information. Investors demand compensation for fundamental market volatility and the cost and risk of illiquidity.
Amihud Measure (Price Impact)
Measures daily absolute return divided by daily dollar volume. Assets with high price impacts impose higher execution costs.
Pastor-Stambaugh (Reversals)
Quantifies liquidity based on volume-related return reversals. The pure effect of liquidity risk demands an estimated ~7.5% annualized premium.
The Liquidity-Adjusted CAPM (LCAPM)
Acharya and Pedersen (2005) explicitly incorporated both the level of illiquidity and the variability of illiquidity into asset pricing.
LCAPM Expected Return
Where R_f is the risk-free rate, E(c_i) is the expected illiquidity cost, and λ is the market price of risk. The systematic risk is measured on returns net of transaction costs.
The Four Sub-Betas of Net Systematic Risk
Market Beta
Standard systematic risk. Covariance of gross return with market gross return.
Commonality in Liquidity
How an asset's illiquidity co-moves with aggregate market illiquidity.
Return Sensitivity
How an asset's return responds to shocks in aggregate market illiquidity.
Liquidity Sensitivity
How an asset's illiquidity correlates with overall market returns.
2. Limits to Arbitrage
If identical assets trade at divergent prices, fundamental theory posits arbitrageurs will short the overvalued and buy the undervalued, risk-free. In practice, cross-border arbitrage is highly risky, capital-intensive, and subject to severe structural limitations.
Agency Frictions
Performance-based arbitrage (Shleifer-Vishny): Arbitrage is managed by professionals using outside capital. If a spread irrationally widens, the fund suffers immediate mark-to-market losses. Investors withdraw capital exactly when the trade is most attractive, forcing premature liquidation.
Idiosyncratic Risk
Dual-listed companies (DLCs) across separate exchanges possess immense idiosyncratic risk because shares are rarely fungible. Arbitrageurs face massive horizon uncertainty—sometimes waiting up to 9 years for convergence, enduring holding costs and volatile spread swings.
Short-Sale Constraints
Arbitrage is asymmetric. Going long is easy; shorting requires locating a borrow, posting margin, paying fees, and surviving recall risk. When pessimistic sophisticated investors cannot short, optimistic sentiment artificially inflates the restricted asset’s price.
3. The Chinese AH Premium Puzzle
Mainland Chinese companies frequently dual-list on the A-share market (RMB) and H-share market (HKD). The Hang Seng Stock Connect China AH Premium Index tracks this. Historically, it fluctuates between 115 and 150—meaning A-shares trade at a 15% to 50% premium over identical H-shares.
Capital Outflow Controls
In most emerging markets, the foreign share trades at a premium due to inflow controls. China is the inverse. The AH premium is a direct consequence of strict capital outflow controls.
Domestic retail investors (80%+ of volume) are restricted from transferring capital offshore to buy cheaper H-shares.
Limited alternative domestic investments trap massive liquidity onshore, artificially inflating A-share demand.
The AH premium is effectively the mathematical shadow price of these government capital controls.
Market Microstructure Demographics
A-Share Market
- • Heavily speculative
- • Highly liquid, Retail-driven
- • High turnover rates
- • Detached from fundamental cash-flow analyses
H-Share Market
- • Global Institutional investors
- • Strict fundamental valuation
- • Emerging market “home bias” discount applied
- • Higher global market beta premium
4. Execution Costs & Tax Frictions
Beyond macroeconomic controls, the daily execution environment imposes severe frictions. While dividends declared are identical, realized net cash flows are distorted by asymmetric tax regimes and stamp duties.
| Market Segment | Investor Profile | Dividend Tax Rate | Regulatory Mechanism |
|---|---|---|---|
| A-Share Market | Mainland Retail / Inst. | 0% to 20% | Tiered for long-term holding. <1 mo: 20%; 1-12 mo: 10%; >1 yr: Tax-exempt. |
| H-Share Market | Foreign / HK Direct | 10% Flat | Non-resident enterprises/individuals holding directly face a flat 10% withholding tax. |
| Southbound Connect | Mainland buying H-Shares | 20% Flat | Punitive flat rate regardless of duration, creating a severe tax asymmetry. |
Hong Kong Stamp Duties
Hong Kong levies a substantial 0.1% ad valorem stamp duty on both buyers and sellers (0.2% round-trip), alongside SFC/AFRC levies. For convergence trades targeting narrow mean-reversions, these transaction costs entirely consume the statistical edge. Direct equity pairs trading bears the full brunt.
The Stock Connect Paradox
The 2014 Stock Connect allowed cross-border trading, theoretically removing segmentation. Paradoxically, the premium widened. Why? Asymmetric accessibility (wealth requirements), isolated mainland liquidity spikes, and crucially, no fungibility. You cannot convert bought H-shares to settle an A-share short.
5. Quantitative Execution & Modeling
Given structural persistence, quantitative funds deploy statistical arbitrage via advanced econometric modeling to harvest alpha from mean-reverting properties.
Cointegration
While individual price series are non-stationary (random walks), a linear combination of the two may be stationary. If an underlying long-term equilibrium exists, the spread is expressed as:
When the spread deviates significantly, quants short the outperforming asset and long the underperforming one. Modern funds increasingly use Machine Learning (e.g., Random Forests) incorporating short-term momentum to predict convergence.
Vector Error Correction Models (VECM)
To model long-term equilibrium and short-term dynamics simultaneously, VECM operates on the first differences of variables while incorporating an error-correction term.
The coefficient α is the speed of adjustment. It must be negative for the system to revert. Empirical analyses show the H-share market frequently serves as the fundamental anchor.
Threshold Cointegration & Asymmetry
Standard VECM assumes symmetric adjustment. Empirically, this is false for the AH Premium. Due to costs and short-sale constraints, small deviations cannot be profitably arbitraged.
Prices follow a random walk until they breach a specific transaction cost “threshold.” Furthermore, adjustments are highly asymmetric: markets correct undervalued H-shares much faster than they correct restricted, overvalued A-shares.