Відео

Glad to help

it is available in all platforms. Linear Regression Linear Regression Channel Linear Regression Slope etc.

Thank you for the feedback 👍

having a template in builder, force the builder to behave a certain way. for example, keeping the signal the same and only add filters using a template will force builder to generate same strategy with only different filters.

I actually use Multicharts. when testing simple ideas you need feedback with variable data printing, debugging, so while it is not shown in my videos, but that how I make sure code is working properly. SQX plays a different role, it is a lot easier to build, test, iterate with automation, but you need to know the proper workflow. for example running 1000 different filters in SQX is a lot easier implementation than MC, but you need to know what filters and what lookback periods, etc. At the end of the day, they are all tools, and if you know what you want to achieve, then you can apply the same concepts in any platform, but each has its own +/-

of course any filter to have an effect it will reduce number of trades. the portfolio I tested had +1800 trades, and after applying the filter it had +1000 trades, so about 40% reduction in trades, but you still have a healthy number left, specially since the filter is the same for all strategies.

you can watch difference between TS and MC over here; ua-cam.com/video/zWY_YeDCuNs/v-deo.htmlsi=vgt4afvV7Yu_5tcb as for SQX it is a different platform, it is more for data mining. all platforms do overlap and each have their own plus and minuses but they are all tools, and you can achieve what you want in any of them to a point.

volatility is expansion/shrinking of range so range, ATR, Bollinger Bands are all can be used to measure volatility

you can do it on open trades, but in the video it is based on profit and loss, so if strategies are producing profits at the same time, then they will be positively correlated.

short term lookback works best, so for range, atr, bolling bands, 5-15 is the best spot

😊

the data copied from Multichart and the graphs done in GPT (using a python library)

Its not about a special filter, as there are hundreds to try, its just the concept to grasp is really important. Try smaller range, smaller ATR, smaller BB width, even lower vix , they all will work

you can join the community go.statoasis.com/join or take the course go.statoasis.com/atm

there are other ways to mitigate risk

thanks for the feedback. will add it to post ideas

thank you for the feedback. running strategies, switching them on/off, portfolio capitalization is all in module 5 of Algo Trading Masterclass. you can register here to learn more go.statoasis.com/atm

without context, it means sell at the close, most likely it's an intraday trade.

exit after number of bars works very well with most strategies

Thank you! I'm glad you found it helpful. Keeping strategies simple can indeed reduce the risk of overfitting.

Glad you enjoyed it

Glad you like it, Simple works 😊

Are other EAs working properly? Arr you using same timeframe and session time? Honestly there are many things that can go wrong, best is to tackle one variable at a time until you find solution

yes, will post it in go.statoasis.com/join when ready

you can't compare a strategy over 4 days with a strategy over 14 years, regardless of the number of trades. this is like coming to a conclusion that every day for the past 25 years is summer because you saw last 4 sunny days in a raw with 30C temperatures. More over, there is nothing of value that comes from testing last 4 days to trade the future.

you take the weekly profit and loss of every strategy and if every week strategy A and strategy B are winners then they are positively correlated and vice versa. low correlation means that their win/loss are not happening at the same time, when measured on weekly basis. you can do this on any timeframe, but these strategies don't trade everyday of the year.

Vix is mean reversion short using one of the 8 patterns. Its not important, i just want convey the idea to start with market edge and then choose strategy style and direction. So you can easily use any oscillator to short the vix and it will work.

Thanks.

This pattern only compares the close to previous close. Larry compare high to previous high and low to previous low

too many acronyms ☺ if you mean trading your developed strategies automatically, then you have many options depending on the language you develop in. for Retail traders, most popular are TradeStation, Multicharts, NinjaTrader, Python platforms like QunatConnect etc. and C platforms like Zorro etc.

@@StatOasis is there ones already developed and back tested? For certain market conditions I.e. Bull market I don't trade, but would like exposure to something automated

if you mean developed strategies, there are many services out there but I highly advice against trading a single strategy, it is much better to trade a portfolio, which again many services available. I have different portfolios in go.statoasis.com/community that trade on monthly signals, Long only, liquid ETFs only, using different flavors of dual momentum

@@StatOasis OK cool yeah I'll check that out! U seem like a vet and know who is actually good out here haha. If u have any other that are trusted let me know. Liked your info and video!

Thank you for watching

of course you can, but its a tool, it is not the reason you will be building a profitable strategy. what you need is a solid workflow that guides you step by step on building robust profitable strategies. that can be done with excel for simple strategies and with any other platform including SQX for more demanding strategies. My ATM is opening in couple of weeks, join the webinar to see if it is a good fit for you go.statoasis.com/atm

@@StatOasis I will. Thank you very much.

My pleasure

on an index like the S&P500 there is no down, I mean there is, but is not usual at least historicaly , specially in the daily timeframe, of course I'm not saying it will stay like that forever ( just look at the japanese index, nothing last forever, but until that moment comes, long term long only positions its the easiest way to go.

most markets have what I call "Market Edge", this could be direction or strategy style, or both. SP500 has both Long Direction and Mean Reversion Style. check this video to see you can even win with random entry using above info ua-cam.com/video/ItUeBDEwhsc/v-deo.htmlsi=Z_SlBfK-sNoSTM5L check also this video: ua-cam.com/video/9G5p9suCHcA/v-deo.htmlsi=CXPvy2lzrraPhqfS In my go.statoasis.com/atm I teach students how to find market edge for any instrument, and I also give a database of market edge for most liquid markets.

i'm a crypto algo developer, so i was wondering about that for crypto Futures. for now i guess i would take the EMA200 as a "hint" for the direction then. Thank you both for the reply!

😂 that is amazon glitch, but the book was published in 1990

Glad you like it!

I don't know the platform you're using, but ATR(7) of 0 candle is < than ATR(7) of 7th candle is not correct, besides you don't really know the value of ATR candle 0, thats why you should check the value of ATR of the previous closed candle, and also, the reason to use ATR is to check the range of the candles, without averaging it, so it should be something like ATR(1) of previous candle < than ATR(1) of the lowest low of the previous 8 candles ( 8 because you are starting to check the lowest value starting from 2 candles ago ).

ATR(7)[0] is the average true range of the past seven bars at the close today. ATR(7)[7] is the average true range of the past seven bars at the close seven bars ago. Btw, NR7 is measuring the range of today bar at the close and comparing it to the previous range of the past seven bars individually and not the average true range. My ATR indicator is flexible were I can change the calculations to be based on Range or True Range, and in this case it was showing the range

you can build a custom block range < range[1] and range < range[2] and range < range[3] and so on .... then you can use that block in Algo Wizard or Builder

try again, I removed the location requirement

no point in using genetic on less than 5000 variations, and no point in using random on above 100000 variations. So between 5k and 100k is where you will need to make a choice, but either way SQX implementation of SPP is only application available publically and it is much better than nothing

there are literally hundreds of variations, nothing is perfect, is all about what you are looking for from the behaviour of the strategy in certain market regime and its equity profile.

you can develop a strategy using multiple timeframes, even multiple instruments in SQX, and you can transfer that strategy to MT, but you still have to create the sub charts for the other timeframes in MT just like they are setup in SQX so the strategy can work.

ASP is good for everything you see on the channel please use my link go.statoasis.com/MC, thanks.