The institutional grade mean-reversion strategy trades the S&P 100 stocks in short-term long-only mode. Backtests show that the strategy has had robust performance, especially during major stock market corrections and it has outperformed the S&P 500 total return by a very wide margin on both absolute and risk-adjusted basis.
The strategy (MRSP100) is long-only with entry signals generated based on our proprietary PSI5 algo also used in our mean-reversion signal service with no changes in the parameters. A maximum of 100 open positions can be held at a time but this rarely happens. The strategy shows superior absolute and risk-adjusted returns with no losing years since 2000 based on backtest results. Note that this is not a data-mined strategy but based on a mathematical model of stock price action. Risk management is very conservative with position size of all trades set equal to one hundredth of available equity in order to avoid equity curve-fitting and selection bias.
Portfolio backtest settings
Strategy name: MRSP100
Time-frame: Daily (adjusted data)
Strategy type: Mean-reversion, long-only
Entries and exits based on: PSI5
Universe: S&P 100 stocks (current composition)
Backtest period: 01/03/2000 – 08/11/2017
Maximum open positions: 100
Commission per share: $0.01
Position size per stock: Available equity/100
Trade entry and exit: Open of next bar (no look-ahead bias)
|Parameter||MRSP100||Buy and hold|
|Avg. bars in Trade||6.6||–|
The strategy has a MAR of 0.41 (CAGR/Max. DD) versus 0.09 for SPY total return buy and hold and outperforms passive investing in SPY on a risk-adjusted basis but also in absolute returns terms by a wide margin in the order of 3%. Based on the backtest results, the strategy gained about 9% in 2008, 4% in 2010 and 13% in 2011 and has not had a down year in the test period except a -0.3% return in 2003.
Performance curves: Equity, Drawdown, Monthly returns and Monte-Carlo analysis. (Click on images to enlarge.)
According to the Monte Carlo simulation, the probability of a drawdown larger than 8% is only about 1%. (Note that Monte Carlo simulations apply only in the case of strategies based on unique hypotheses, as the MRSP100.)
Disclaimer: No part of the analysis in this blog constitutes a trade recommendation. The past performance of any trading system or methodology is not necessarily indicative of future results. Read the full disclaimer here.
All charts were created with Amibroker – advanced charting and technical analysis software. http://www.amibroker.com
Copyright notice: Any unauthorized copy, reproduction, distribution, publication, display, modification, or transmission of any part of this report is strictly prohibited without prior written permission.