Quote from stephencrowley:
I see a lot of typing but nothing that makes sense. Every post you make has the word 'velocity' about 500 times..
You make assumptions about my strategy, I am not going to give that away. It is based on mean reversion but not mean reversion of prices.
Quote from Thunderdog:
Am I to understand that making 1% on full trading account equity per day on a consistent basis is less than outstanding? Are you actually saying that traders ought to be consistently earning up to 4 or 5% on full account equity on a daily basis? And all this against a background of minimal drawdown? Is this what you are actually suggesting?
Quote from Grob109:
Thanks for your response, I appreciate it.
Your collective comments explain a great deal about what you want to do and how you are going to go about it. We all benefit from your commentary.
Treating performance in the context of time does generate a time rate of change measure, that for me, is significant. The two things traders spend are time and money. Time is the more dear since it is not recoverable. And time must be spent in the market to make money. When capital is sidelined it has absolutely no value as an earning power.
One of the most rewarding segments of refinment is optimizing when capital is switched from one application to another based upon its earning power. The measure of that power comes down to the time rate of change at which capital is earning money.
Your target is 1% per day. A day turns out to be a very coarse measure of how time works in making money.
Again, thank you for responding.
Quote from Random.Capital:
The first problem is you don't actaully have a "sufficiently real simulation".
Good luck! Looking forward to seeing the real trading.
Quote from stephencrowley:
I appreciate the skepticism.. but my data feed is very nice and every quote and trade is timestamped by the exchange (accurate to the millisecond) and also as soon as it hits my system, I also do rigorous tests on the round trip times..intradaily seasonal latency, etc. By this I can use bootstrapping to arrive at reasonable confidence intervals of quote speed and execution probability.. etc.