J, I admire your commitment to your system. It has good track record overall. But you have those few runs that screw everything up. My opinion you need some kind of failsafe stop. 10%? or 5-6% EOD or 10% intraday. Maybe higher 15%? I dont know. Maybe backtest different stops and see what ultimate returns and/or losses it prevents. Even if the market starts falling, Your system might trigger "long" signal way before breakeven on this trade.
Ya, an emergency stop-loss trigger would help. Because... a $10,000 portfolio allocated 60% TZA, 30% UVXY, 10% cash, beginning June 27, would show a balance of $6,426.13 as of last night's closing prices, without considering trading fees, a loss of 35.7%. That is, if you sold. So we don't sell, we'll keep hanging on. It will eventually turn around -- or drop another 35.7%. Methinks, FWIW, the parameters for market behaviour are following a brand new set of rules.
Stop losses aren't a panacea. As J knows, many system backtest worse with stops than without.
And this market isn't "unreal." A robust system should be able to do well in all kinds of markets, including strongly trending ones like this.
The real issue is whether J's system was optimized in a certain kind of market regime and is unable to handle different kinds of regimes. J understands this issue better than anyone, so I'm not criticizing. In fact, I'm merely articulating what is without doubt something that J is thinking about too.
J, I admire your commitment to your system. It has good track record overall. But you have those few runs that screw everything up. My opinion you need some kind of failsafe stop. 10%? or 5-6% EOD or 10% intraday. Maybe higher 15%? I dont know. Maybe backtest different stops and see what ultimate returns and/or losses it prevents.
ReplyDeleteEven if the market starts falling, Your system might trigger "long" signal way before breakeven on this trade.
Ya, an emergency stop-loss trigger would help. Because... a $10,000 portfolio allocated 60% TZA, 30% UVXY, 10% cash, beginning June 27, would show a balance of $6,426.13 as of last night's closing prices, without considering trading fees, a loss of 35.7%. That is, if you sold. So we don't sell, we'll keep hanging on. It will eventually turn around -- or drop another 35.7%. Methinks, FWIW, the parameters for market behaviour are following a brand new set of rules.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteThis is an unreal market. What a recovery two days in a row/
ReplyDeleteStop losses aren't a panacea. As J knows, many system backtest worse with stops than without.
ReplyDeleteAnd this market isn't "unreal." A robust system should be able to do well in all kinds of markets, including strongly trending ones like this.
The real issue is whether J's system was optimized in a certain kind of market regime and is unable to handle different kinds of regimes. J understands this issue better than anyone, so I'm not criticizing. In fact, I'm merely articulating what is without doubt something that J is thinking about too.