I reposted your original question above. You specifically asked for which historical market period is best to compare to the current one ? It's an easy answer. You need to pull up the VIX chart and look at the similar volatility surge in the historical past and then compare those reasons to the current reasons involving a country out of Asia and FEDs pending interest rate decision. It only took me two minutes on Google. Now it seems like your question really was about something else involving intraday price patterns and price correlations. Therefore, the golden rule applies. If not willing to do the research and confused about the market conditions...best to go on vacation and stay away from the markets until the dust settles. Then again, by the time the dust settles...there will most likely be another volatility surge lasting 1 - 2 weeks or longer as the historical data shows often. You will be on vacation for many months every year. Then again, your backtesting and research should have traversed in the past the other market conditions. You should at least know how your algo system would have performed like in 2008 and 2009 ? Vacation Activities: 1) If you have kids...do volunteer work at your kids school because that's school time duration. 2) If you have another job...works more hours. Extra money always help. 3) Pick up a new hobby or restart an old hobby. Whatever you do, make sure its not related to the markets to prevent being tempted to trade when you know market conditions are currently not suitable for your algo system. I know an algo trader that hasn't been in the markets for about 3 weeks now because he said market conditions are not suitable to be trading. Yeah, he's waiting for the dust to settle and protecting his trading capital.
thx wrbtrader Backtesting shows it did reasonably well on 2008, 2009. 2015 was fantastic, except for ba da bum August and hit all the criteria for stopping trading.