Paper: arXiv 1611.03110
Authors: Tim Leung, Jamie Kang
Abstract
American Depositary Receipts (ADRs) are exchange-traded certificates that rep- resent shares of non-U.S. company securities. They are major financial instruments for investing in foreign companies. Focusing on Asian ADRs in the context of asyn- chronous markets, we present methodologies and results of empirical analysis of their returns. In particular, we dissect their returns into intraday and overnight com- ponents with respect to the U.S. market hours. The return difference between the S&P500 index, traded through the SPDR S&P500 ETF (SPY), and each ADR is found to be a mean-reverting time series, and is fitted to an Ornstein-Uhlenbeck process via maximum-likelihood estimation (MLE). Our empirical observations also lead us to develop and backtest pairs trading strategies to exploit the mean-reverting ADR-SPY spreads. We find consistent positive payoffs when long position in ADR and short position in SPY are simultaneously executed at selected entry and exit levels.
Complexity vs Empirical Score
- Math Complexity: 6.5/10
- Empirical Rigor: 6.0/10 (capped from 8.0: transaction costs not modelled; no out-of-sample test; see penalty flags)
- Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 8.0)
Why this score: The paper demonstrates strong empirical analysis with robust backtesting of trading strategies. The application of Ornstein-Uhlenbeck processes to ADR-SPY spreads for pairs trading shows a good balance of mathematical modeling and practical application. The findings regarding asynchronous market dynamics and their exploitation are a valuable contribution.
Research Flowchart
flowchart TD
A["Research Goal: Analyze Asynchronous ADR Returns & Develop Trading Strategies"] --> B{"Key Methodology: Decompose ADR Returns"};
B --> C["Data Input: Asian ADRs & SPY ETF (S&P500)"];
C --> D{"Computational Process: Fit ADR-SPY Spread to Ornstein-Uhlenbeck Process via MLE"};
D --> E{"Computational Process: Backtest Pairs Trading Strategies"};
E --> F["Key Finding: Consistent Positive Payoffs from Long ADR / Short SPY at Selected Entry/Exit Levels"];
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper develops and backtests pairs‑trading strategies that go long an ADR and short SPY to exploit mean‑reverting ADR‑SPY return spreads.
Source sentence
Our empirical observations also lead us to develop and backtest pairs trading strategies to exploit the mean-reverting ADR-SPY spreads. We find consistent positive payoffs when long position in ADR and short position in SPY are simultaneously executed at selected entry and exit levels.
Assumes 4
- Both long and short positions are worth $100, resulting in zero net cash outflow. execution
Source sentence
We assume here that both long and short positions are worth $100, resulting in zero net cash outflow.
- Each position is entered and closed only at market close. execution
Source sentence
We assume that each position is entered and closed only at market close.
- Overnight returns are the dominant driver of overall ADR price fluctuations, implying larger magnitudes and higher standard deviations than intraday returns. market
Source sentence
The unique opposite timezone characteristics of the selected ADRs give rise to our initial hypothesis that overnight returns are the dominant driving force of the overall ADR price 5 fluctuations. Hence, overnight returns are expected to have relatively greater magnitudes, and thus higher standard deviations.
- Intraday price fluctuations are mainly attributed to the U.S. market, while overnight fluctuations are driven by home‑market sentiments. market
Source sentence
By doing so, we attribute intraday price fluctuations mainly to the U.S. markets, and those during overnight to other factors, such as home market sentiments.
Data
- ADR price data (opening and closing prices) market-prices public
Source sentence
Using the prices available on Wharton Research Data Services (WRDS) and Yahoo Fi-nance, we collect the daily opening prices, P open i , and closing prices, P close i , from 6/15/2004 to 6/13/2014 (2517 trading days, 10 years) for every sample ADR.
- SPY price data market-prices public
Source sentence
Using the prices available on Wharton Research Data Services (WRDS) and Yahoo Fi-nance, we collect the daily opening prices, P open i , and closing prices, P close i , from 6/15/2004 to 6/13/2014 (2517 trading days, 10 years) for every sample ADR.
Costs
Not modeled
Finds 5
- The return difference between ADRs and SPY is a mean‑reverting time series well‑fit by an Ornstein‑Uhlenbeck process. statistical
Source sentence
The return difference between the S&P500 index, traded through the SPDR S&P500 ETF (SPY), and each ADR is found to be a mean-reverting time series, and is fitted to an Ornstein-Uhlenbeck process via maximum-likelihood estimation (MLE).
- Overnight returns have higher standard deviations than intraday returns for most ADRs, indicating larger price fluctuations overnight. statistical
Source sentence
In Table 2, we show that overnight returns have higher standard deviations than intra-day returns do for most ADRs, except in the cases of the three Indian ADRs (HDB, INFY, WIT) and one Taiwanese ADR (TSM). This confirms that most significant (in terms of magnitudes) price fluctuations tend to occur during overnight periods.
- Intraday ADR returns show relatively high correlation with U.S. market returns, confirming that ADR prices move in line with SPY during U.S. market hours. statistical
Source sentence
12 Most of the ADRs considered herein have positive intraday mean return spreads (XID) with high SPY correlations.
- Backtested pairs‑trading strategy (long ADR, short SPY) yields consistent positive payoffs when entry and exit levels are selected based on the mean‑reverting spread. performance
Source sentence
We find consistent positive payoffs when long position in ADR and short position in SPY are simultaneously executed at selected entry and exit levels. The profitability of such a strategy can be estimated through backtesting with empirical data from June 15, 2004 to June 13, 2014.
- The ADR‑SPY spread series display typical mean‑reverting behavior, with abrupt declines often followed by upward spikes. statistical
Source sentence
The return spreads in Figure ?? display rather volatile movements. In particular, every abrupt decline is often followed by an upward spike, which is a typical behavior of a mean-reverting process.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 30 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-15 · same card as JSON · A language model chose these items: check the source sentence before relying on one.