CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. -- % of retail investor accounts lose money when trading CFDs with this provider. You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.

SUG SuperDarwin @bendex

Mechanisms are in place to detect low liquidity and trading is automatically halted when this occurs, the lower time frame strategies are more sensitive and will halt at a smaller reduction in liquidity than the D1 for instance. (The performance of D1 strategies are less prone to slipping a few fractions of a pip or even a few pips.) This is to minimize slippage and to prevent entry during low liquidity around news etc. With the safety mechanism no unexpectedly large losses because of slippages have occurred in the last several years. My EA produce slippage logs and stores slippage data in GV which enables me to track accurately.

No scalpers. A collection of fully automated breakout strategies H1 to D1.

No martingale or grids here. All trades always have a reasonable SL and TP plus alternate exit signals including trailing stops. All trades use a stable risk amount based on the strategy, total equity in account and current volatility. I do NOT increase risk in any way because of losses or short term winning positions, profits or streaks. I have made some anti-martingale and illusions of profitability posts on various forums complete with simple mathematical explanation as to why the risk vs reward is always better when you use a stable risk. Cav’s post sums it up nicely here.

As of 2020 Jan, SUG is a 6 time prize winner of the Darwinia Contest and currently are awarded 377,921.65 € Darwinia funding. (Most recent prize win)

About 8.2 Risk Stability, the score itself is reasonably stable at this point over time. Darwinex does not seem to differentiate between risk fluctuations from hourly strategies to the D1 strategies. One will be a little larger risk size but normally held for only hours with a tighter stop-loss and one will be smaller risk size but held for possibly several days or longer. Some strategies are slightly larger positioned because of their slightly higher edge. Another reason for fluctuating leverage is that some trades from different markets or different time-frames do overlap. The different strategies across different time-frames sometimes have an overall small amount of correlation but this is factored in while I’m selecting strategies and calculating for risk sizing. Each strategy has a static stable risk size but since Darwinex does not differentiate by Magic Number. I have to settle for a non-perfect Risk Stability score. (At this time, a 10 for Risk Adjustment though!)

About 5 or 6 for consistency of return and duration. Multiple time frames, trailing stops and additional exit strategies seem to be penalized. (https://community.darwinex.com/t/consistency-investable-attribute-r-r-dc/347) Each strategy has preset static exit strategies but since they are all on the same account the over all timing and P/L of exits are varying and the scores reflect that.

And lastly the OS and CS scores, another attribute is quite low due to a penalized trading style. Some breakout entry levels may appear to the Darwinex system to be “late” but the strategies are designed to enter later once a breakout is more certain to minimize fake-out losses. Even a simple trailing stop hurts the CS. I think it is the accuracy, risk and performance that matters most (PF,EX etc). Here’s a supporting thread about os/cs.

These are the notes but I leave the judgement to you if SUG is a good fit for your portfolio:
https://www.darwinex.com/darwin/SUG.4.2/

Please feel free to post any questions, comments or concerns.

11 Likes

Hi @bendex, welcome to the Community.

I agree with your comments on Risk Stability and share your ‘sadness’ that Darwinex has trouble identifying the actual risk a trader is taking due to not taking in account the StopLoss. It kind of ‘forces’ the trader to trade with fixed pip values, instead of fixed Risk per trade, but that’s a discussion to be had in another thread :slight_smile:
To get a better score here, you would need to split up your strategy over multiple accounts where each account only trades one timeframe, but that’s a lot of effort for little reward I think.

A poor OS is just inherent to breakout strategies, or any other strategy that waits for some kind of confirmation before entering the market. You can’t really argue that it would be more profitable to have entered sooner :wink: so the poor OS is justified. But smart investors know this too, so I wouldn’t worry too much about the OS score.

Also, the good thing is that Darwinex actually adjusts the weights of attributes depending on the type of strategy. For example, if Darwinex detects your account is ‘scalping’ with short term trades, OS and CS are more heavily weighted and LA is less weighted. With longer term strategies, like yours, the LA (and possibly some others) becomes more important and the OS and CS become less important to the overall D-score.

All that said, keep up the good work! Your strategy seems to be a healthy one from my perspective.

3 Likes

I’d say in HINDSIGHT having entered at a better price would have always been more profitable but I think we can argue something like “Waiting for confirmation can lead to a better performance over-all for some methods, strategies or people trading live or OOS.”

OK, you’re right, I’ll try not to sweat it anymore. Thanks for the vote of confidence, hopefully investors do know this. I guess investors that like breakouts will be used to the low OS. Darwinex generally does an awesome job of giving investors information and they seem to be improving. I guess this one is so curious to me because waiting for confirmation seems to be so common.

This is fantastic, I didn’t know this. Thanks for the heads up and kind words.

1 Like

That’s how the OS and CS scores work :slight_smile: They look in hindsight.

Yup, I agree completely. This will show in your PF score, which imo is a lot more important than OS and CS (and Darwinex agrees, as it also gets a lot more weight in the total D-score, at least for long term strategies).

2 Likes

Investors do not care about OS CS R+ R- …
SYO always had a very low OS and reached 4 millions.
Same story about THA .
These rates are cool but have not predictive value.
PF is good when you come from a lucky semester. :smiley_cat:

4 Likes

5 posts were merged into an existing topic: Filtering out marti-grids

Hi @bendex
I want to congratulate for your trading results and also for you quality contribution to this community.

Can you share something more about your background and your backtesting/optimization process?

Thanx! :wink:

2 Likes

Thanks.

In short I’m a programmer turned poker player turned trader. Before you make any snap judgments, I took my poker career seriously. Some notable career advancements came from studying work by Dan Harrington and David Sklansky as well as joining poker schools including PocketFives, CardRunners and DeucesCracked. I studied every day for hours either at one of the schools or analyzing my millions of hands contained in my Pokertracker database looking for holes in my own game. I quickly developed the discipline to properly manage a bank-roll, walk away from tables that weren’t (very) profitable and to stay out of stakes I wasn’t ready for.

I started dabbling in trading around 2007 first as a manual trader, I took the standard route of jumping from book to book and system to system trying different things. I studied the james16 thread at forexfactory and even joined their school for a short time. Eventually I decided to take advantage of my programming capabilities and by 2012 I had mostly stopped manual trading and began developing my first promising EA based on breakouts. I have completely rewritten the EA 3 times since then but the basic concepts still remain.

As for optimization, I don’t do that for these breakout strategies. After hundreds of walk-forwards I was not able to find any system of optimizing and then walking forward that could outperform the base strategies (in further out of sample data.) That’s not to say I’ve given up on walk forwards and optimizing for all my endeavors but for these breakout strategies in general, I don’t actually optimize.

As for tools I currently use for SUG quants: MT4, TDS, Quant Analyzer and Excel. I must give some credit to StrategyQuant as well for some of the generated strategies which inspired some ideas for components in my own EA. I can also give credit to the two authors who were the most influential to my quant trading: Keven Davey and Robert Pardo. Also credit to @TradeSignalMachine for his fine walkforward software WFP which I’m sure I’m not done with, not to mention for him guiding me to check out Darwinex!

12 Likes

That’s funny because that is pretty much my own career path :laughing:

8 Likes

And how do you decide the parameters ( stops and targets )?
Is this a parameterless strategy?

I think that you mean that you don’t re-optimize, but you need to find the values for the base strategy…

1 Like

Some might call it that but I do not consider the strategy selection stage to be an optimization even though sometimes (but not always) the optimizer is involved.

3 Likes

I love this response… excellent statement capturing a complex subject in one sentence - well done :slight_smile:

Great to see your contributions here on the forum @bendex :muscle:

3 Likes

The best time to add SUG (once you have already decided to buy) to your portfolio is at the first moment you can find it FLAT (not having any trades open - this should be somewhat frequent.)

Timing this Darwin on recent wins or losses would be futile. This Darwin is somewhat complex/diversified via different markets, types of breakouts and time-frames. A small DD or run-up is luck and noise.

2 Likes

After a lot of analyses I have decided to drop the max spread on all the strategies in the portfolio, this should trim about 2% of the “slipperiest” trades and should not effect the edge of any of the strategies. This is in light of the large divergence experienced in some recent AUDUSD trades and also some GOLD trades to a lesser extent. End result should be significantly less chance of large divergence.

3 Likes

Hey Bendex,
congratulations for the great results. Ten consecutive months with a profit, heading for one entire year in the green. Impressive :clap:

I wanted to ask you two questions:

  1. For the first 7 months of your Darwin, you had a good profit that was taken away from the market. You started again from 0, at least in terms of profits. Since then, you had impressive results. Is it just variance or there were some changes in the system?

  2. Why did you stop playing poker? Just curious. I always found fascinating the similarities between poker and trading. I think that being a former poker player helped you tremendously in being successful in trading. Trading requires someone to understand that being good doesn’t mean winning every single trade (hand) and every single day. It requires someone to understand money management. It requires someone who can manage “tilting”, to use a poker term :slight_smile:

PS: not a poker player, so I don’t challenge you :smile: Just passionate about the mathematics and psychology behind it. I read Harringtons’ and Sklansky’s too.

5 Likes

Thanks!

  1. During the first 7 months it was realized that 3 of the strategies were probably stale, it was also realized around the same point after copious testing that a couple of key things could be tweaked in regards to my strategy building process. So I’d say a combination of both “expected variance” (I expect strategies to go stale or do poorly sometimes) and some changes to the building process.

  2. I had a friend who was in need of a developer and he somehow got it in his head that I am the only person he would trust with this contract. Reluctantly I cam up with a number that would convince me to commit to a mostly working at home contract and he agreed. I tried to do all 3 (trading, poker, dev) but just didn’t have the focus or time once the dev work started to increase. At the time I saw the dev job as a means to grow my trading account a little faster than poker could. I still play live a few times a year for fun and occasionally flip through one of Harrington’s newest books. I bet the games online are pretty good these days with everyone being locked in their houses and all. I secretly hope my dev career ends so I can have an excuse to play again FT! Also, I agree with all of your additional points, well said.

5 Likes

Hi @bendex! I am a SUG investor and I have observed that the DARWIN has been inactive for a few days. Was wondering if there is any important reason for that? Greetings!

2 Likes

Hello, thanks for the question. Everything is definitely functioning normally. The reason for no trades the last few days is that there have been no signals for any of the break-out strategies in the SUG strategy portfolio. It is unusual but it does happen from time to time.

7 Likes