I just finished watching one of your series on a VWAP trading system. At the end the difference between the "Realize" button and backtesting is explained briefly, but I am still confused. If i am not entering inputs to optimize during Realization, is it any different than backtesting? Wouldn't "walk forward" testing sometime in the past be the same as backtesting?
Let me know if this page answers your questions: https://www.linnsoft.com/feature/optimization-trading-systems-realization
Yeah, I saw that page. I understand how it can be used to optimize a system by testing on different instruments, but I still don't understand how it would behave any differently than a standard backtest if the variables are kept consistent. If I am not running an "optimization" on any of the variables wouldn't the realization give me the same results as a standard backtest given the same time period? I am optimizing and backtesting my system starting from 2007 in an attempt to include as many market conditions as possible
Optimization (and realization) do require variables, sure. You have to have a parameter that you are adjusting...that's the whole concept behind optimization. Whether that variable be a stop or target, a moving average period, there are 1000s if not 10s of 1000s of things that can be optimized. But if you have no variables, there is no optimization (and no realization as realization runs on top of optimization). Optimization doesn't require multiple symbols....it is often run on a single symbol. But there must be a variable to optimize (or multiple variables to optimize).
Ok, yeah, I think I am clear on Optimization. I was more so trying to understand Realization. In that training video you backtest for 3 months and have good results but then you optimize and backtest over a longer period and have poor results and I just wondered if you had just backtested a farther back period from the beginning if you would get the same results as the realization. And that is where I am at. I am running a strategy that backtested very well since 2007 on an intraday time frame but is underperforming since I began simulated trading it a month ago. I wondered if somehow backtesting left something out that the realization process didn't. Thanks