Disliked{quote} in any interview; you will 100% be asked a question like "did you backtest your strategy?", and 100% they will ask you to show them the backtest results. regarding your points above; they are correct points; but I would really re-phrase all of them as: - any strategy will not survive for more than 3 years; regardless of what is it based upon strategies usually need to be improved from time to time; almost on weekly or monthly basis. I have been trading strategies that are based on the same concept for more than 5 years; however due to the...Ignored
Let me ask you one specific question; since you trade MT4; do you record tick data from your VPS server? - I will never use a VPS server. I have a gigabit high speed internet and that seems to do the job. In one of my post earlier, I had described how I free myself from any concerns with latency, slippage, spread, trade context busy etc. by factoring those costs * multiple in my design. To summarize, a strategy is not considered profitable by me unless it recovers 4 times the trading costs (spread, swaps, commissions etc). it's embedded as part of my per trade pnl calculations.
I personally do record every tick in live trading; and I replay them at the end of the day; backtest my strategies and compare the live performance with backtest. Please share what you mean by "record every tick and replay". Is this some feature in MT4 that I am not aware of? Regardless, my universal time action strategy requires that I run it 24hoursx5days/weekx4 weeks/month and possibly 3months/quarter and 12 months/yr but I could not backtest more than one month's worth of M1 candles via my automation so had to limit myself to 4 weeks. Recording tick data across 28 pairs would churn a lot of disk IO and my laptop is already heated up doing all the crunching for 25 seconds at the open of every minute. Would love to know more about this replay thing. sounds interesting.
Mt4 is just a stepping stone for me. Perhaps I will continue to use it only for signal generation and move to a more robust API model down the road. I really don't need MT4 as I don't need any charts or historical candles. Just a handful (or even just the previous candle) open and close price is all I need. I don't need or care about high and low either as I am a 100% indicator-free dude. I don't believe in any measurement of price. Any measurement of a constantly changing value will only result in just another constantly change value. It's useless to me and must be ignored for a strategy to be consistent. Sorry/ I digress.
Edit: for those who understand coding, I use over 28*9x62x8x6 recursive loops arriving to a 'fixed' dot in price vs time using the Inception Method (similar but not the same as described in millipede) hence the churn on my CPU processing power.
Staying in my lane...