.... The fact of the matter is that backtesting is a piss poor correlation to live trading results.
But a necessary step without which you do not even know where system stands.
.... The fact of the matter is that backtesting is a piss poor correlation to live trading results.
after these few weeks of shake and scoop, why would anyone still want to 'back test'.... it's a rigged game that I have explained play by play in my thread trading is easy and many other threads.
it's a rigged game.
there are no bulls and bears, just the same fcking group of boys manipulating this thing up down up down.
understand their game, everything becomes easy.
Stocks are and always were. Commodities are a different animal all together. Our algos are 100 stats driven and back tested.
your approach has merit as well. Buy, hold and hope. Not my cup of tea.
after these few weeks of shake and scoop, why would anyone still want to 'back test'.... it's a rigged game that I have explained play by play in my thread trading is easy and many other threads.
it's a rigged game.
there are no bulls and bears, just the same fcking group of boys manipulating this thing up down up down.
understand their game, everything becomes easy.
The backtesters will just update their model so it somehow "works" during the latest drawdown. I used to be one myself. Here's a good example.
"Paul has a GEM-like model called DM-COMP which is extremely like GEM, except it uses VEA (developed markets) instead of VEU (whole world ex-US). The key to Economic Pulse is the SPY-COMP V2 risk-on risk-off switch. That's the golden nugget. Once you have that switch, you can use whatever ETF pair you like. He uses the average of 3, 6, and 12 month momentum to score, whereas Gary uses only 12 month. This would make DM-COMP more whippy than GEM, but the risk-off switch is not always activated, so overall DM-COMP is not more whippy. Because the risk-off switch is not always on, he also has a 10% stop-loss trigger. This fired on 12/31/2018. It caused a whipsaw. Lately, he thinks if you have 40% bonds you can skip using the stop-loss."
Use more "whippy" (fast) signals and 10% stop losses...until you get burned the next 4 or 5 times we have a correction and it quickly recovers...and you get whipsawed out of a 10-15% gain. Then go back to a slow, end-of-the month, 1 year momentum signal until it loses over 30% like this year.
you still haven't understood any bit of what I been saying.
i fully understand what you preach. So far it worked, well yet to be seen for this downturn. But your method leaves you in hope mode. Essentially you hope that market will always keep on going up. I can not trade like this.
I can't use Excel for this much data.