Yep, its a fun platform and it seems to like my signal. I have a pretty good ratio of 2.77 and Darwinex actually increases the size of my trades (larger than my C2 trades) to normalize the risk to 6.5% (since Aug12 my DD has been <5%).
@BlackOpzFX2 what a creative and interesting take on Martingale! Hereās hoping youāre on to something, looks good so far! 
1 Like
@ChrisPage Hey Chris! Iād been reading your posts since you joined C2. Welcome To The SMARTY-Verse. An upside-down world where Martingales dont BLOW UP accounts and leave you in the unemployment line behind the Wendys but instead actually promise profits without excessive risk. Buckle Upā¦
Have you tested this martingale with a simple Head or Tail game that pays even money?
What was the maximum drawdown after, say, 10 000 spins?
Just curious, from a mathematical point of view.
Yep and the results were OK but not impressive. My improvement that BOOSTED profit potential and allowed me to reduce risk and still earn was my dynamic RR system. I accidently disconnected it and saw my returns plummet. Its part of the āsauceā. What you describe are the types of systems and testing I developed on the way to this one.
1 Like
I figured that much, because from a mathematical perspective no martingale, no matter how sophisticated, can make money (let alone beat the market) if the trading system itself has no edge to begin with.
Election Day: SMARTY - Smart Martingale had an interesting election day and night. I decided to let the algo run since I felt I had enough STOP code to limit losses to the downside (I did wish I could reduce lotsize 50% but C2 has a minilot minimum). The night started out with a HUGE Black Swanish price spike to the downside for the algo.
I actually felt comfortable with the $-560 BUY loss since when the price spikes wildly its easy for your SLās to get badly gapped. I got slipped just a little but I was pretty happy with the more than recoverable 1% loss. A few hours later the algo flipped sides and entered a SELL trade that resulted in good profits. This week and next will continue to be volatile but I feel pretty good about letting SMARTY continue to run without any manual intervention.
Also between the larger gain/loss trades the algo played pretty good defense by using its defensive āexit & re-enterā strategy. When the currency moves from strength to weakness the algo has the possibility of closing the trade and limiting larger losses and the trade record shows multiple times where this code executed and limited losses to $30 or less. As always after events Iām looking for improvements but so far Iām pretty satisfied with how the existing SMARTY algo executed the strategy.
Update: I did program a modification to the loss-mitigation code that has the possibility of creating āsofterā landings when trading with higher risk lotsizes. The edit aims to reduce trade exposure vs profit gain. Trading profit potential for increased safety. (just the way I like it)
For fun I fed my SMARTY-Smart Martingale stats into Claude AI and asked it to explain and evaluate my numbers. I was pleasantly surprised with the stats when explained. They hit my design goals for the algo of minimizing risk and drawdown which in the end will give me the headroom to increase profits with manageable risk in the future. Having the AI confirm my goals just by reading the stats is a nice surprise.
C2 Stats: W:L ratio - 1.46:1 Sharpe Ratio - 2.55 Sortino Ratio - 5.11 Calmar Ratio - 7.124
Iāll break down each of these trading performance metrics and evaluate them:
1. Win/Loss (W:L) Ratio - 1.46:1
- Meaning: For every losing trade, there are 1.46 winning trades
- Evaluation: This is a solid W:L ratio. While not exceptional, anything above 1:1 means youāre winning more trades than losing. However, W/L ratio should always be considered alongside average win and loss sizes, as one large loss could wipe out multiple small wins.
2. Sharpe Ratio - 2.55
- Meaning: Measures risk-adjusted returns, comparing excess returns to volatility (both upside and downside)
- Evaluation: This is an excellent Sharpe ratio. Generally:
- 2+ = Very good
- Above 3 = Excellent Your 2.55 indicates strong risk-adjusted returns with good consistency.
3. Sortino Ratio - 5.11
- Meaning: Similar to Sharpe but only considers downside volatility (harmful volatility)
- Evaluation: This is an exceptional Sortino ratio. Generally:
- 2+ = Very good
- Above 4 = Excellent Your 5.11 suggests the strategy handles downside risk very well.
4. Calmar Ratio - 7.124
- Meaning: Measures return relative to maximum drawdown (peak to trough decline)
- Evaluation: This is an outstanding Calmar ratio. Generally:
- 3-5 = Very good
- Above 5 = Excellent Your 7.124 suggests exceptional returns relative to the worst drawdown period.
Overall Assessment: These statistics indicate a very well-performing trading system with:
- Consistent winning trades (good W/L ratio)
- Strong risk-adjusted returns (Sharpe)
- Excellent downside risk management (Sortino)
- Superior returns relative to maximum risk taken (Calmar)
The most impressive aspects are the Sortino and Calmar ratios, suggesting particularly good management of downside risk. However, remember that past performance doesnāt guarantee future results, and it would be worth looking at these metrics over different time periods to ensure consistency.
Ummm Sirā¦ This Is A MARTINGALE!?
SMARTY - Smart Martingaleās Darwin Qualified for my 2nd 30K today (60K Total). Iām up 2.37% but really should be more than 5% (my Darwinex trade-copier was offline and of courseā¦) so no off-switch for at least the rest of the year. C2 didnt have an issue and caught all of the move+ (5.8%). So far even with the volatility of the election weāve been riding pretty smoothly so far. Typically Nov/Dec are good for robots that like an active market and SMARTY is no exception.
SMARTY - Smart Martingale has continued chugging along and this is more a ādiaryā post than anything else. It was a pretty quiet economic news week and I expected SMARTY to not do too much. Thursdays āJobless Gainsā and āHome Salesā news saved the week and closed an open GBPUSD progression for +2% setting me up nicely for the coming news filled week. My Darwin also progressed from Top 15% to Top 10%. SMARTY prob wont advance to the 35K tier in Nov but next month I should be setup to smash though quite a few funding tiers.
Data-wise after a couple months running the āAlgo Completeā version (with numerous updates) Iām pretty confident that SMARTY would mostly likely be able to recover a mitigated max-loss within the typical positive month. In my backtesting I frequently see my max-loss mitigated from 8% to one or two percent. If the mitigation reduces that 8% to less than 5% I consider that a win since the trading record proves that its possible to profit in a technically ābustā month then continue gaining.
Next week should be Fun!
Happy Thanksgiving!! SMARTY - Smart Martingale is done for the month and did quite well on C2 for a nice percent gain of 7.8% vs 6.73% for the source account (positive slippage). This is the first month in a while where I havenāt had an idea to improve the ātradingā portion of SMARTY in some way.
I can tell its mostly code-complete when my minds starts drifting to āquality-of-lifeā improvements. SMARTY has a few key properties that can be tested within the algo. So I programmed an in-code FXScan module which shows a FX strength rating that at minimum excludes weak currencies and at best highlight possible strong winners. Its already revealed a few strong potential currencies for further diversification.
In closing my Darwin made 4.19% for Nov (even excluding the trade-copier miss) which placed it in the Top 10% on the platform and also locks in a second 30K allocation. It also makes a third December allocation very easy to achieve with only 0.5% required. Of course, Iāll be shooting for more so December should hit additional funding levels.
SMARTY - Smart Martingale is READY for December. This month ended cleanly with the 1st being Sunday, the first trading day of the week. Just completed my monthly optimizations and as-per-usual one of the currency parameter sets usually carries over. I always feel good about that since it shows robustness. Not only did the USDJPY settings work on October unseen data (best set to use in Nov) but now also worked on November unseen data.
GBPUSD has improved parameters that outperform the November set. These are 6-month backtests that project 3%+ returns but for the last few months the actual LIVE results have out-performed the backtest. December should be more of the same since its the EOY month. Typically, January starts out slow as trading resumes so that month it might stick closer to the plan (even though Trump inauguration will probably boost the markets).
My Darwinex account gained 4.19% in November and almost activated my 5% Permanent 100K allocation. Only 0.81% left for activation. I like the USDJPY settings since they are more risk-averse, so I feel comfortable rolling -them-over for another month. This week Iāll continue running GBPUSD optimizations to test for improvements, but this GBP/JPY set looks pretty solid.
NO NEW TRADE DAYS: Dec 25 - Xmas Only, Dec 27 to Jan 1 - trading resumes Jan 2, 2025. (open trades will continue to work but initiating new trades is restricted).
FINALLY! - SMARTY - Smart Martingale hit a rough patch. Since I implemented the algo on Aug 12th its been fairly smooth sailing. The robot likes volatility and for an extended period this past week the market was indecisive and triggered repeat SLās. I wasnt worried since I have plenty of safeguards and āstopā code to limit downside loss well below 8%. Worst case this scenario looked like 5% which is within same month/next month recovery per testing and optimization.
BUTā¦ Instead of downside SMARTY recovered 99.69% of the dip. GBPUSD only executed mitigation and positive closure on the last trade to close the sequence when normally multiple trades in a losing sequence would post minor losses or small profits. This sequence is about as unlucky as I can get without hitting my expected 5% of 8% max. (TOTAL: 1.92% Loss)
At extended levels the robot is more concerned about mitigating loss than net profit and getting out of this trade at an easily recoverable 1.92% would be fine with me if I were trading GBPUSD alone butā¦
Iām NOT - GBPUSD is paired with USDJPY which was chosen for its uncorrelation and this episode proves it. So while GBPUSD was struggling to recover, USDJPY was just chugging along printing profits. Filling the GBPUSD gap even as the hole was being dug.
Nicer point is that if their fortunes were reversed GBPUSD would net much more profit per pip that USDJPY and the net gain between the two would be larger.
My backtests tell me to expect an extended trade sequence every 4 or 5 months and this one has appeared on schedule. I never had to touch the robot during this sequence and just let it do its thing. In hindsight the one time I even thought about touching it to ālock-in-profitsā would have been a mistake. I āthink about itā but NEVER intervene and prefer to fix the bot if it needs improvement over trying to manually trade on top of it.
Even though I DID develop a positive ātweakā for SMARTY after analyzing the GBPUSD trades. Its a small minor code update but in this sequence it would have prevented at least ONE of the GBPUSD loss trades and thats good enough for me. SMARTY is built as a trade-management and ārecoveryā bot and so far has performed exactly as designed. Next week is Inflation and Jobs - Should be a Nice Ride!
Hmmmā¦ Month has been pretty quiet for SMARTY - Smart Martingale. Got a couple of percent ahead then the reverse in Drawdown. Have dipped and pulled out of it 2X. Actually, Iāve been feeling pretty good about seeing how well my ārecoveryā systems are working to prevent serious loss. So far the system is demonstrating exactly what I wished for - nice safe lower-risk profits when going āgoodā and mitigating losses to near breakeven or small loss/profit when things are going ābadā. So far so good. Even the losses at their greatest were <3% and in-month recoverable. Just what I wanted from a ābest-caseā scenario of my downside systems.
This dip DID give me a chance to program a possible long-term code enhancement but for the most part I didnt have to touch anything. Posting today because at the mid-point (12/14) and checking my Darwinex.com rating I was shocked to see its in the top 15%!? - WHAT!?- Looks like it been a rough month for most and Iām not the exception. It appears that no-matter the system the market has been chaotic enough to screw up most peopleās plans. Next week is FULL of high-impact events so I expect a nice positive holiday bump heading into Xmas since my algo prefers volatility.
Hi - Just read through the thread and Iām still not super clear on whatās the claim on the algo. Iām not very familiar with martingaling outside of the fact that itās bad and donāt do it in trading, so itās very possible that my ignorance is at work here. However, isnāt the whole point of martingaling to increase the pace of recovery (as you suggested in the thread)? But how is that done without increasing the risk if the subsequent trades do not work? Thatās the part that Iām confused by.
The claim is that I can build a martingale-āishā system that leverages some of the ārecoveryā advantages of martingale without the risk of blowing your account. I use a martingale āstyleā progression and initially 2X sizing after losses but higher in the progression of loss trades I divert from ātrueā martingale which will easily blow trading accounts of any size. At that stage I use multiple methods to get out of the danger zone even if it means takes a small loss of a couple percent.
Google martingale systems and watch a few videos. They always BLOW the account. Most martingales have a beautiful 45 degree slope (many small wins) with huge losing streak drawdowns until the eventual day the drawdown blows the account. Watch a few Youtube videos on them, the gains can be amazing! Those systems are the majority of robots sold since they work until they dont.
My edits to this system increases the survivability of running this type of algo. When my systems is profiting it mimics the slope of true martingale while also demonstrating that during periods that would cause horrible drawdown for typical martingale systems (like NOW) I seem to be ācoastingā until the market hits āsmooth-sailingā and we resume the slope. Lately even the S&P
has been taking a hit and currencies are tossing around but SMARTY - Smart Martingale is riding the volatility. Its the 'Better" Smarter Marty.
In addition Iāve taken a system known for HEAVY drawdown and tamed it 6.4% max lifetime and <5% since the algo was fully implemented on Aug 12th (the start of the rise).
Hi - thanks for the response. It would be helpful to see a real example of what this means. Iām trying to wrap my head around how this works but just canāt see it, so if youāre willing to share Iād love to learn more.
I ask because in my head when i read a strategy is a martingale strategy itās screaming āRUN AWAY!ā There have been managers on C2 who have gone the martingale route when theyāre in a losing position and itās never turned out well (as youād expect), so you may face skepticism for a while. Regardless I wish you all the best with the strategy and will keep an eye on the trades. Thanks!
1 Like
Iām curious too. I just looked back at all your November and December trades, and they follow this pattern:
- Put on a position.
2a. If position goes down, close position (possibly with a stop loss order, canāt tell)
2b. If position goes up, add a small additional position, and put the original position on a very short stop loss.
2a results in a fair number of small losses.
2b results in a fair number of modest gains.
I donāt see a single example of a trade adding on additional position when it has gone down.
Any sort of Martingale Iām familiar with is focused on trying to turn losses into gains by ādoubling downā, except not usually actually with doubleā¦ more like leveraging up. The main strategy mechanisms here seem to be (a) cut losses quickly, and (b) lock in gains as they go up. I like what I see being done, but I donāt see the implied relationship to a smarter Martingale style.
I use martingale sizing BUT i dont ADD to losing trades. If you notice I go from 1 to 2 to 4 to 8ā¦ on losses. But youāre right I close dem losers FAST! I take the risk of ādoublingā my lotsize but not my dollar risk. I attack losers with increased risk but control downside losses to fit my ā fraidy catā risk tolerance. (Iām a risk-hawk. My theory: If I can keep downside risk low while generating outsize reward/risk then I can always dial-up the lotsize multiplier for even more profits)
Theres a LOT going on under the hood. Traditional martingales keep every losing trade open until profit. I dont. Been there, done that, paid the swap fees and have the T-Shirt, sometimes generating negative swap on multiple large trades for weeks until closure. Nope. Bye. Bye loserā¦ I can still recoup without carrying the sequence losers.
Most martys are DOUBLE or DIE. Instead of account wipeout at DIE. My algo can āsurrenderā at max 8% loss or less. An amount thats within one month recovery.
@Babbage_9010 - Concerning point 2b - I dont anti-martingale and add to winning positions either. C2 has a unique way of executing partial trades that makes it 100% compatible with any broker but obscures trade actions. It closes the entire trade and re-opens just the partial size. So closed trades are the sum of all trades closed. Example: 0.50 trade = 5. If I close 0.20 (2) C2 will close the 0.50 (5) and reopen a 0.30 (3) but reports profit/loss on the 0.80 (8) when the entire trade is closed (5 p&l + 3 p&l). It can make decoding the trade record a bit tricky. You might think I opened a 5 then an 8 but nope. Just C2 lotsize accounting.
Thanks! That makes sense.
1 Like
