Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Backtesting a Moving Average System on QQQQ

|Includes: PowerShares QQQ Trust ETF (QQQ), SPY

No disclosures

Below is my second backtest using NinjaTrader (the first backtest was on SPY). I decided to start a series of backtests I will be conducting on Scott's Investments with a backtest of Mebane Faber's 10 month SMA system detailed in The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets.

The system is simple: After QQQQ closes the month at or above the 10 month simple moving average, a buy signal is indicated at the next months open price. Note that I required the security to be trading for at least 10 months before the first signal because 10 months of data is necessary before one can calculate a 10 month average. When QQQQ closes a month below the 10 month SMA, the long position is exited on the next months opening price. I used 1 share as the default quantity simply because I am interested in percentage returns and returns relative to buy and hold.

Note: A sell was made at the 9/8/09 close in order to calculate returns to date. This does not mean SPY or QQQQ dipped below their 10 month SMA on 9/8/09!

All data is split adjusted but not adjusted for dividends. I hope to have dividend adjustments integrated at some point, but for now the data is not adjusted for dividends. All tests exclude slippage and commissions, which can be significant depending on the frequency a system has you trading, and do not account for any taxes. So in short these results are not necessarily indicative of what you may experience in the real world but can act as a starting point for the validity of a system.

At first blush, the QQQQ backtest does not look great. However, when compared to buy and holding QQQQ, the outperformance has been clear. The primary benefit of the system is avoiding periods of significant drawdown, which QQQQ has seen two such periods. In addition, the QQQQ backtest shows more trades, which can be expected when a security trades in a range like QQQQ did in 2004-2006.

For some additional technical analysis on QQQQ including some short/medium term projections, Adam Hewison of INO does a great job in this free video.

Performance All Trades Long Trades
Total Net Profit -16.64 -16.64
Gross Profit 19.61 19.61
Gross Loss -36.25 -36.25
Commission 0 0
Profit Factor 0.54 0.54
Cumulated Profit -7.01% -7.01%
Max. Drawdown -25.46% -25.46%
Sharpe Ratio 0 0
     
Start Date 1/1/1999  
End Date 9/8/2009  
     
Total # of Trades 11 11
Percent Profitable 27.27% 27.27%
# of Winning Trades 3 3
# of Losing Trades 8 8
     
Average Trade 1.51% 1.51%
Average Winning Trade 21.50% 21.50%
Average Losing Trade -5.98% -5.98%
Ratio avg. Win / avg. Loss 3.59 3.59
     
Max. conseq. Winners 1 1
Max. conseq. Losers 4 4
Largest Winning Trade 38.50% 38.50%
Largest Losing Trade -12.38% -12.38%
     
# of Trades per Day 0 0
Avg. Time in Market 163.27 days 163.27 days
Avg. Bars in Trade 5.4 5.4
Profit per Month 0.06% 0.06%
Max. Time to Recover 3530.00 days 3530.00 days

Instrument Action Qty Price Execution (at open of month)
QQQQ Buy 1 96.19 1/2000
QQQQ Sell 1 85.19 6/2000
QQQQ Buy 1 93.5 7/2000
QQQQ Sell 1 89.81 8/2000
QQQQ Buy 1 103 9/2000
QQQQ Sell 1 90.25 10/2000
QQQQ Buy 1 25.27 3/2003
QQQQ Sell 1 35 5/2004
QQQQ Buy 1 36.27 6/2004
QQQQ Sell 1 34.61 8/2004
QQQQ Buy 1 36.98 11/2004
QQQQ Sell 1 36.82 4/2005
QQQQ Buy 1 38.03 6/2005
QQQQ Sell 1 36.86 7/2005
QQQQ Buy 1 39.64 8/2005
QQQQ Sell 1 38.93 6/2006
QQQQ Buy 1 40.6 10/2006
QQQQ Sell 1 45.5 2/2008
QQQQ Buy 1 49.92 6/2008
QQQQ Sell 1 44.81 7/2008
QQQQ Buy 1 35.76 6/2009
QQQQ Sell 1 40.74 9/2009

BUY AND HOLDING QQQQ AFTER 10 MONTHS OF TRADING:
Performance All Trades Long Trades
Total Net Profit -55.45 -55.45
Gross Profit 0 0
Gross Loss -55.45 -55.45
Commission 0 0
Profit Factor 0 0
Cumulated Profit -57.65% -57.65%
Max. Drawdown -57.65% -57.65%

Qty Price Month (execution at beginning of month)
1 96.19 1/2000
1 40.74 9/2009