CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. 64 % 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.

FLI - SteadyPips

Hello Darwinex community,

I want to introduce my DARWIN FLI to the community as it is featuring in the top of the October 2019 DarwinIA leaderboard and I feel the time is ripe to give my strategy some exposure.

First let me tell you a little about myself and how I ended up here. I am a 29 year-old econometrics student in the final months of finishing my master’s degree in Quantitative Finance. I have been trading some CFD’s since 2011, blowing some of my poker winnings on Plus500 and other bucket shops. In 2014, I stumbled into the Bitcoin market and built up a comfortable bankroll for a student. Only to to lose the majority in the January 2015 Bitcoin crash, trading with 2-3x leverage. Fastforward some and I have managed to build a good bankroll once more, meanwhile studying Quantitative Finance, going into the 2017 Bitcoin peak. Realising that the easy money would be gone soon, I decided to cash out a significant portion and to study the ‘big-boys’ market more, as I like to refer to FX and equities trading as opposed to the crypto cesspool. I was introduced to Darwinex by the promise that consistency, durability and openness could be rewarded with investor’s money.

As I set off, I managed to do OK for a bit, until I hit a streak of bad luck and start overleveraging and blow up. After messing about with three discretionary strategies (two I actually made a DARWIN for, which I left up for transparancy purposes), I basically stumbled upon a surprisingly simple way to get a statistical edge predicting the intraday price move of the EURUSD pair. I turned this into the automated strategy FLI.

The main reason I run this strategy and decide to show it to the world is because I want to prove that I can make things work if I research my ideas well enough and am able to stick to those, even in periods of drawdown, which the strategy obviously experienced in the first 6 months of running it. Additionally, I am looking for feedback on what makes this strategy interesting to invest in or, more interestingly, get dismissed.

Some more information on the strategy itself:

  • Every morning it makes a prediction on the direction of the EURUSD market for that day, and it tries to enter as advantageous as possible.
  • The strategy does not hold more than 1 simultaneous trade at any time and never holds trades overnight.
  • The strategy adjusts position size to assume a fixed leverage.
  • It was backtested for 15 months, or 336 trades and yielded 204%, with a max drawdown of 22%. It achieved a max D-score of 84.4. Some more details seen below

  • The main reason for the larger drawdown in the live strategy is seen well when comparing the daily return distributions of the backtested and live strategy. The left side of the distribution in the live environment (bottom) has several more -5% to -6% days than the distribution in the backtest environment (top)
  • Some of these large down days could have been prevented by applying some discretionary intervention as they mainly occured during expected market-moving events such as NFP/FOMC/etc.

I hope this gives a nice overview of me and my strategy and I am happy to receive any comments or feedback and answer any questions you might have

Cheers!
SteadyPips

8 Likes

@Livenemaxx
This is the tipical case of promising Darwin snubbed by investors and prized by darwinia, a robust growth without a manipulated inception on a low that is often acheived with trackrecord farming.

Keep on @SteadyPips ! :+1:

3 Likes

I agree it is good one for DarwinIA and small investment. But it will struggle with investment over $1M.
But maybe it is enough for this trader.

2 Likes

Welcome. I like you honest intro, finally someone who is able to speak about tough beginnings!
+Darwin looks promising. Good luck :slight_smile:

3 Likes

Thanks for the good vibes so far!

I think it is important to be open about your previous dealings as to show your intentions here. I believe there are very few succesful traders that haven’t blown accounts or screwed up in any other way. I often see this as a strength, as people who have failed in the past and can be open about it show that they have taken something from it.

As for the Capacity problem mentioned by @Livenemaxx, I agree and have been getting to grips with what influences divergence. Although I do not remotely think I will reach ~1M$ in AuM any time soon, I have done some research on how to limit divergence. To see which trades introduce the most divergence I have invested a small amount of money into my own DARWIN. It will probably help a lot if I slightly tweak my time-based executions, which I now do at the first tick of a new hour.

Cheers!

5 Likes

Hi Steady pips and welcome - can you please confirm exactly what date you started this new system, has it been from the beginning of your FLI trade history? What do you think made last month so successful? A key element for me in any Darwin is consistency- no nasty surprises within reason, I am not sure how any mathematical model can predict the movement of any currency with so many news based spikes but wish you luck!

1 Like

Hi marketserpent!

Thanks for dropping by! Let me answer your questions and elaborate slightly

  • I did indeed start my system on the the 19th of September 2018. I think it was the beginning of that month that I had an aha-moment and I started tweaking my inputs. The backtests I ran do actually go until 19th of September as well. Actually, when I started the system I was pretty disappointed to go into 40% drawdown in 6 months, but analyzing the trades it seemed that I initially ran into a streak of bad luck, as can be seen in the return histograms in the first post, where we see an extraordinary amount of relatively large losses.
  • Last month has been succesful due to two reasons, the main one being no more than chance. Last month had a 75% strike rate which is pretty good when flipping a 54% biased coin 23 times. The second reason is that I did introduce some discretionary closing ‘rules’, which I applied yesterday for example. While the system dictated me to close at 22:00 GMT, I closed a little earlier when the price started to bounce. I intervened 6 times this month, but I would have to check what influence that had exactly.
  • As for consistency, my system trades relatively big and a string of bad luck could quickly eat into profits. But on the other hand, a string of luck can quickly turn good profits. There is no telling in advance.
  • However, I believe that there is two upsides to my system, one stemming from the fact that it keeps relative risk constant. This will mean that a constant investment in the DARWIN will expose the investor to a constant absolute risk. The other upside is that it takes every decision without any regard for previous trades, minimizing risk of blowup.
  • As for the news spikes, when they are expected by the economic calendar, I do allow some manual intervention since this month to set certain take-profit or stop-losses for the day. Not all news-based volatility can be predicted however, which is something the system has dealt well with for the past 2.5 years (50/50 backtest/live).

I hope this clearly answers your questions, happy to answer any more if necessary :slight_smile:

Cheers!
SteadyPips

1 Like

Many thanks for the comprehensive response - investors would have had to be very patient with that 6-month spell of bad luck;) Seriously, it is good to see the last 6 have been much better and your recent improvements are obviously paying off. The issue investors have when investing in relatively new Darwins is whether the latest downturn is a recoverable drawdown or collapse of the entire system to join the many languishing in the 95% DD category, as I am sure you can appreciate, of which only time will tell…wish you luck!

Yep, so was I, but the system being automated and the money itself not being a big deal helped me endure the drawdown. I actually think that it is very good that this spell of bad luck is in there so investors know what they can endure when backing the system. If they can stomach the variability shown, they deserve the rewards. I dare to safely say that the probability of going into 95% drawdown is close or lower than the probability of the strategy hitting 1900% profit eventually, given that we do not end up on the wrong side of a ‘January 2015 CHF-type’-event.

Coming back briefly to your earlier question on what made this month so well, I did a tally on when I intervened and it shows that mitigating some of the worse days proved profitable. In the end I managed to improve the month’s results by 20% 2019-11-01%2012_05_36-Window

In other news, the strategy made it to #7 in the October 2019 DarwinIA and has been allocated an extra €76’385.52, bringing the total AuM to approximately €120k. It will be a good test to see whether I can keep the same level of trust in my system with some significant backing.

Cheers!

2 Likes

WOW…
These are words from a true trader…
I have to resist the temptation to buy FLI and stick to my rules. :wink:
Great job and keep on .

2 Likes

Pareto rule : 80 : 20 ,
80% of the profit is done on 20% of time.
The only way to turn a 99% random market ino a straight 45° line is a martingale.

While naive investors look to return experienced ones look to trading journal .
FLI has a perfect risk management and a 100% consistent ruleset.

1 Like

I dont’ want to steal your recipe but can you add any details?
Candlesticks, indicators … something else?

Sticking to the rules, I like that, I read somewhere that you require Darwins to be at least 2 years old, am I right?

Couldn’t have worded it better :slight_smile:

Can’t really go into much depth other than that it is mainly a reversion strategy. This is especially reflected in the Os attribute. I have identified some sweetspots where price has a larger than average probability of reversing, both in time and price. Hope you understand I do not want to go deeper than that :wink:

Cheers!

2 Likes

200% right.
Return is significant only on 2y timeframe, everything less is flawed by luck (either good or bad), so you need 2 years native and public, no hindsight and survivorship bias.