Chaos in Technical Analysis and Backtesting – Part II: More on Close vs. Adjusted Close

This is the second in a series of posts about the impact of data adjustments on technical analysis studies and backtesting, as well as, of the pitfalls of trading system development due to incompatibilities and/or conflicts between data series and analysis techniques. Use of the wrong data series for technical analysis and backtesting can under certain circumstances produce wrong, even chaotic results, as I will show in this second post with specific examples. 

In the first post I demonstrated  that adjusting for dividends results in a non-linear distortion of the  price series backwards in time. An animated image was used to illustrate that visually.

Let us now examine how these adjustments may affect backtesting results. For this purpose we will backtest a simple stop-and-reserve system using adjusted and non-adjusted data for SPY. The system rules are as follows:

Exit short and enter long if: MA(50) > MA(200)
Exit long and enter short if: MA(50) < MA(200)

In other words, this simple system generates a long entry signal if the 50-day simple moving average crosses above the 200-day simple moving average, after covering any open short positions and if the 50-day simple moving average crosses below the 200-day simple moving average the open long position is closed and a short position is opened. Again, this is a very simple system and it is used here just to illustrate the effect of data adjustments on backtesting results. This is not intended to be a trading system recommendation and I just want to make sure this is understood by our less experienced in this area readers.

The initial capital of the trading system was set to 100K in the backtests to follow. The position size was calculated based on current bankroll divided by the entry price.

Case 1: Backtesting results using the non-adjusted price series

The trade-by-trade results of the backtesting of the simple system described above are shown below:

It may be seen that this system generated 17 trades in the price history of SPY. The compounded annual return is show as 9.70% and the maximum drawdown was -27.30%. This system had a win rate of 58.22% in this case.

Case 2: Backtesting results using the adjusted price series

The trade-by-trade results of the backtesting of the simple system described above are shown below:

In this case the system generated 19 trades, the compounded annual return was 11.30% and the maximum drawdown was -32.76%. The new win rate is 57.89%.


It is clear from the results that the two backtests produced different entry and exit points in time and as a consequence, different performance parameters. It is also clear that Case 2 does not produce realistic results as compared to a trade log using this system and SPY actual data as in Case 1, because older entry and exit points of the backtest would depend on future adjustments that have not taken place yet. Thus, the backtest based on adjusted data does not correspond to the performance what would have been obtained had this simple moving average crossover system been used to actually trade SPY since its inception. Thus, the results of the backtest in Case 2 fail to satisfy the requirement of providing a realistic account of past performance.

Therefore, use of dividend-adjusted ETF and stock data for backtesting trading systems is not in general recommended unless there is a well-understood reason for it. The impact of dividends on trading system returns should be evaluated separately.

I want to emphasize again that the conclusions of this study concern only adjustments for dividends. Splits must be taken into account in all backtests otherwise the price series will include gaps and the performance parameters will be distorted, often seriously. In a future post I will deal specifically with split adjusted data price series and their effect on backtesting.

 Check back next week for the next post in the series!

Charting program: Amibroker (Charts created with AmiBroker – advanced charting and technical analysis software.”)   

Disclaimer: The author is not a financial advisor and does not recommend the purchase of any security or advise on the suitability of any trade or investment in any timeframe. ETF, stock, futures, forex and options trading and investing involves substantial financial risks and can result in total loss of capital. If investment or other professional advice is required, a licensed professional should be consulted.

This entry was posted in Technical Analysis, Trading Strategies and tagged , , . Bookmark the permalink.