ADR Arbitrage and Cross-Listing

An ADR arbitrage trade compares a U.S.-listed depositary receipt with its local ordinary shares after applying the depositary ratio and currency conversion. The basic parity is simple; execution is not:

\[ FV{ADR}={ratio\times P{local}\over FX_{local/USD}} \]

A persistent residual can represent a genuine opportunity, but it can just as easily compensate for time-zone mismatch, conversion fees, short constraints, or a market that is closed. Treat cross-listing as a package of legal and operational claims, not a spread chart.

Construct point-in-time parity

InputRequired control
ADR ratioEffective date, corporate-action adjusted
Local share quoteExchange session and currency
FXTradable bid/ask, not end-of-day fixing
Depositary feesConversion/cancellation schedule
Dividends and withholdingEntitlement dates and tax treatment
BorrowLocate availability and rate on each leg

Use executable bid/ask prices on both legs. A mid-price residual frequently disappears after paying two spreads, FX, and financing. If the local market is closed, build a fair value from index futures, sector proxies, FX, and the ADR itself, then label the residual as a lead-lag prediction rather than locked-in arbitrage.

Conversion is the economic anchor

When conversion is possible, a rich ADR may be shorted and created from local shares; a cheap ADR may be purchased and cancelled into locals. But conversion has eligibility, minimum-size, settlement, custody, and time delays. During stress, these frictions expand and the parity band should widen rather than trigger larger positions.

def adr_residual(adr_mid, local_mid, ratio, fx_usd_per_local):
    fair = ratio * local_mid * fx_usd_per_local
    return 1e4 * (adr_mid / fair - 1), fair

Check the FX convention carefully. If fxusdper_local is dollars per local currency, multiply; if it is local currency per dollar, divide. Unit tests should use known examples because an inverted FX quote creates convincing but catastrophic signals.

Hedge and funding risks

The short leg can be the limiting input. ADR borrow rates, recalls, and settlement fails can dominate expected convergence. The local leg may have foreign ownership limits, capital controls, or different holiday calendars. These are closely related to the availability and rebate mechanics in securities lending.

RiskExample response
Closed local marketReduce size; hedge with futures
Ratio change or splitProcess point-in-time corporate action
FX jumpMaintain explicit FX hedge
RecallPredefine forced-cover procedure
Conversion delayCharge financing and gap-risk reserve

Research design

Measure residuals at aligned clock times and separate open/closed market intervals. Backtest with historical FX bid/ask, actual borrowing availability where possible, and conversion-band assumptions that worsen in volatile regimes. Analyze convergence conditional on liquidity; an average half-life conceals tail episodes in which conversion is impaired.

Cross-listed residuals can be correlated through country, sector, and funding shocks. Position limits should group exposures by economic issuer, not treat each listing pair as independent alpha. That is a standard systematic crowding risk failure mode.

Data-quality checks

Create daily parity exceptions for missing FX, stale local quotes, zero volume, changed ratios, and unconfirmed corporate actions. A single stale local close can create a spectacular residual that survives cross-sectional filters. Compare the implied ADR fair value with depositary announcements and vendor reference data, but retain the original inputs so any override is auditable.

The relevant holding horizon begins at execution, not at signal observation. Charge financing and currency hedging through the conversion settlement date, and report residual convergence separately from the return earned while the local market was closed. This prevents a lead-lag forecast from being misclassified as conversion alpha.

Key takeaways

  • Convert prices using the live ADR ratio and correctly oriented FX quote.
  • Trade executable quotes and a conservative conversion band, never mid-price parity.
  • Closed-market gaps are predictive lead-lag signals, not immediate arbitrage.
  • Borrow, settlement, custody, and corporate actions can dominate spread economics.
  • Aggregate risks by issuer and funding regime before scaling many pairs.
#ADR #cross listing #arbitrage #FX #equity relative value