I hate to hijack this thread but since you are implying (knowingly or unknowingly) that I belong to the "clueless" camp, let me explain my take on the subject of backtesting a little further. Even though I do not backtest anymore (I did in the past), let me first say that backtesting is not totally without merit. It can serve a purpose in developing a trading strategy (if an aspiring trader doesn't yet have one), however, backtesting also brings along a lot of pitfalls. For one thing, it creates a sense of false security (patterns that worked in the past may or may not work in the future). It creates expectations that may lead to a disappointment if they don't pan out. Backtesting also doesn't account for slippage, it is impossible to simulate slippage by backtesting (forward testing in real market conditions is needed). I ditched backtesting a while back, after I had developed a handful of strategies. Instead, I started to forward test and gradually fine tuned my approach until a light bulb finally came on.
Markets change several times every year via volatility alone. Discretionary traders (intuition or rule base) must change when market conditions change or simply recognize the changed environment and then use the appropriate trade strategy suitable for that new environment. Thus, to succeed...discretionary traders need to either adapt their trade method or switch to one of their other trade methods that's more suitable for current trading conditions. Simply, trying to survive via "one trade method" as if it will fit all trading conditions...it won't be a successful approach to trading. The same is true for the trader himself/herself...they too need to adapt to how they approach their trading or adapt their perception of the market when market conditions change. wrbtrader
I reacted to this statement: "Backtesting is worthless IMO" So you confirm that backtesting is NOT worthless. Which was my point. Pitfalls exist in anything you do. It does not create expectations, at least not if you stay realistic and don't start to dream. Backtesting does account with high probability for slippage as I see the backtested slippage compared to the real slippage over the years. Only if you trade very big size it gets difficult. What I built with the help of backtesting is still working after more then 20 years. Even with forward testing your system can stop to work, this is not a result of backtesting. So you confirm that you used backtesting till you found a handful of strategies. You ditched backtesting AFTER you found these strategies. So backtesting is not worthless and you are not clueless as backtesting gave you at least something to start with (a handful of strategies). When I was still backtesting a light bulb came already on.
With automated mechanical trading you absolutely must backtest, as it can reveal many ways of trading the wrong way faster and cheaper. Also for long term trading you never stop the system at DD as that will remain a loser, and real DD may be worse than BT. You stop systems that obviously are not working correctly or as expected. Sometimes systems stop because third parties change too much, so good to not rely too much on others and diversify too. Also for discretionary trading you must weed out all nuances. So don't stop at scratching the surface.
%% HELPFUL illustration fan27; i just did a gaze @100+year chart of cig smoking in China, Russia, So Korea ,Japan=HUGE UPTREND.....And do i study 100+ years of stock data?? Sure +more than that. I quit smoking, for sure after about 7-8 , may be 10 years ; i hated to keep paying uP to to kill my self early. US smoking Tobacco use chart is down trending.[1965 to 2014 students and adults]. But i do see several tobacco stocks ''HOT'' ,LOL [ NOT a stock tip] Back to original poster question; many of them + i do also, pay attention/use to 10 year charts+ 200 day moving average , but not in a mindless, mechanical sense..............................................................................
Yes. Yes. It's conceptually very simple but can't easily be programmed (maybe an AI system could do it) because it's highly visual - no indicators. I'm looking for clues in the way the bars form. Yes, - although to me now, it just feels like one system (I adjust as the market presents slightly differently each day). I forward tested in a sim for 3 years, acquiring more awareness of things as I went along.