DrBourse
If you don't Ask, you don't Get.
- Joined
- 14 January 2010
- Posts
- 888
- Reactions
- 2,092
YUP, BB System IMO.
Skate, your preferred "Selected Period "seems to be 3 weeks.
So can you enter, or program a WMA 15 into the system (15 = 15 trading days in the 3 week Selected Period).
Maybe a WMA 21 would be ok.
"don't let the simplicity of this system fool you".
you got it completely wrong @DrBourse hereOK, so unless you can select your own "weighted average setting for the selected period", then it's a "BB System", Dangerous to say the least - I have always avoided BB Systems.
With all Software Trading Platforms, the crucial thing is that the user "Must be able to Drive EVERYTHING", otherwise we have no idea how their formulas are constructed.
That is, they could be built so that the Software gives a result desired by those that constructed it in the first place.
In other words, they could have programed the WMA to always show results that are below the Black Box Systems favoured EMA & SMA.
Years ago I challenged a couple of BBSystem Providers, they all refused to explain their "in house formulas" - anyone with an enquiring mind would have to as 'WHY the reluctance to be truthful'.
Count me out of this discussion, Personally I would not place any creedence in any of their results.
Sorry M8
fully understood, no grief, it is good to have some differing and sometimes confrontational (is that english?) opinionsOK qldfrog, never used Amibroker, so I was flying blind with most of my questions, just wanted to make sure it was totally driveable, and not a Black Box System.
NP.
Cheers
Misunderstanding i am sureCaptain Obfuscation (aka the bull headed one) inferring Amibroker is a blackbox system? That's rich, must be why the Amibroker user manual is 1478 pages long, after all its such a simple black box.
That ratio of consecutive losers to consecutive winners of 2.5x would be mentally very tough to tradeView attachment 135950
The 2 year backtest below is for comparison that includes the COVID period
The results below use a "3-Week MA" of the "High" & the "3-Week MA" of the "Low" forming a channel equal to the average weekly volatility for the 3-Week period with slight alterations to the original code.
The re-coding of the original strategy was slight but extremely time-consuming
Both strategies incorporate a Stale Stop, Trailing Stop & TakeProfit Stop but the most recent version of the strategy is the one on the right. The improvements were worth the effort. The original strategy (the one on the left) I wouldn't consider trading but with the improvements, I would. The low channel exit as suggested by the strategy developer has been removed from the "Stale Stop Exit" using my default setting instead was where the biggest improvement came from. The "system drawdown" is now within my limits.
The slight re-coding has changed the metrics
As a mechanical system trader, I find it difficult to trade a system knowing that strategy improvements always come at a cost that I have highlighted in red.
View attachment 135952
Trading a 3-Week Moving Average Channel
Overall it was a worthwhile exercise & I thank the member for passing on the information.
Summary
Would I really trade the updated "3-Week Moving Average Strategy"? - Nah, not really.
Skate.
People’s personal approach will often dictate what system metrics are important to them. For me personally my day-to-day approach to trading is not particularly sensitive to overall win/loss %. By that I mean the win/loss % doesn’t mess with my head when it comes to the challenges of getting up every day and placing my trades day in day out. However, what really does mess with my head is the consecutive number of winning or losing trades—I find it really tough to place a long sequence of losing trades no matter how well I know the system. So I tend to only live trade systems that have a much greater number of consecutive winning trades than losing trades. Whether a system has 40%, 50% or 60% overall winning trades doesn’t really bother me too much.Very true MA. As a generality (not commenting on Skate's backtest specifically), many of us initially think below 40% win rate is palatable, but is actually pretty "miserable" lot of the time over the long run. Doable, but lots more discipline and gumption required in the long run.
Must confess I personally watch % winners closely in backtests, but will have to keep a better eye on the "consecutive win" or loss parameters during backtesting.
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?