Announcement

Collapse
No announcement yet.

Momentum Capture Signal

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Momentum Capture Signal

    Hi Everyone,

    I decided to offer my momentum strategy via simpletrader.com:

    https://www.simpletrader.net/signal/...m-Capture.html

    It is also possible to invest at Darwinex:
    Meet DARWIN index BAX and add it your portfolio of liquid alpha from €/£/$ 200. Broker and asset manager regulated by the FCA (UK) and the CNMV (Spain).


    It is a fully automated trading system. Only in very rare cases will I close positions manually.

    No martingale or averaging is used. Every position has a clearly defined stop loss, usually between 20-100 pips.

    The EA trades during the hours after speeches from central bankers or other central bank events. It does not trade very often, but historically it has been good in catching the momentum afterwards without too much risk in periods of low volatility. After detecting strong price action and momentum, it will open positions in trend direction as well as pending orders for a possible reversal.

    Therefore, the EA can be seen as a "long-term volatility bet".

    The Following symbols will be traded:

    EURUSD,EURJPY,EURAUD,EURCAD,GBPUSD,USDCAD,USDCHF,G BPJPY,AUDJPY,USDJPY,AUDUSD,EURGBP,EURCHF and CADJPY

    The historical maximum drawdown in an 8-year portfolio backtest starting 2010 was about $311 per 0.01 lots and 16 max orders, so $933 (or 19%) for the current settings on a $5000 balance. The maximum stagnation period was 79 days in the portfolio backtest.

    In (optimized) backtests the current settings would have more than 60% a year, but considering optimization bias 30-50% is probably more realistic. Of course, none can predict the future. We can only judge from historic tests and live trades.
    The backtest can be downloaded here: https://c.mql5.com/6/799/MomentumCap..._portfolio.zip

  • #2
    Hello Corni

    Looks interesting. Very good stats on the Darwinex link.

    Few questions.

    1) Could you please explain why there is a 32% drawdown in the MQL5 signal?
    2) Why other EA's/traders would you suggest it "could be similar to" in trading style

    KR,
    BRM

    Comment


    • #3
      Am I the only one who can't see Corni's reply to BRM?

      Comment


      • #4
        I wrote a lengthy reply. But after modifying it, it was labelled as spam. I wanted to wait a bit in case it gets accepted by administrators, but it seems they are busy with other stuff.
        So here another answer:

        1) I started the account with $100, because at that time I had no intention of selling the EA or offering a signal and it was just an experimental account. As also explained in the signal description, the 32% drawdown occurred when the balance was low, somewhere between $100 and $150, so the account was much too small for multiple orders and the actual drawdown was also rather small, $40 maybe.

        To have a "clean" signal, I started a new one at ICMarkets: https://www.myfxbook.com/portfolio/m...arkets/2380299
        But I also continued with the old one to have a longer live history.

        Interestingly, it seems that both myfxbook and simpletrader do not calculate the drawdown accurately. It was only a short intraday drawdown and maybe they only use end of day data, not sure. But this is also an important point if you want to judge martingale-type strategies. On some of those the actual drawdown might be much larger than the one shown at simpletrade ...

        2) The strategy is a bit similar to "SFE Price Action" if you only used it during specific time periods and a bit more aggressively (in terms of entry level, not lot size). Therefore, I would also use SFE as a "benchmark" to compare performance against. If you look at the SFE MAMM, it is down about 10% since Aug 2017, while my EA made about 20% since then. The time period was rather bad for momentum systems but my systems was historically good in reducing the risk as much as possible while still having most of the high-probability opportunities.
        One important difference is that my EA has a clearly defined maximum risk with leverage < 10 for all symbols taken together. SFE on the other hand sometimes triggers with a lot of symbols, especially on USD events, so if it is lucky it makes more than my strategy, but if it isn't it loses more. So overall I think the return / risk is better on my strategy. Of course, I can only judge from historic bakctests and performance. Any system can fail in the future.

        Comment


        • #5
          Hello Corni, congratulations for the excellent performance and low DD.
          May I know if you are in the habit of optimizing your EA every six months to better adapt to market changes?

          Comment


          • #6
            Thanks.
            Actually since I published it, I optimized every week to improve the parameters and also add changes to the logic. I am also always open for interesting suggestions for improvement.
            But from the portfolio backtest below you can see that the performance is not getting worse in recent years. I think this type of market behavior does not change in six month or even a year. It will probably only change when a lot of volume trades the same way.
            Of course, the backtest has some optimization bias and therefore the future profit will not be the same (the portfolio below is on 0.01 lots with max 16 positions at a time). But I changed all parameters up to 20% and it was still very profitable, so it is not an extreme over-optimization.


            portfolio__2.png

            Comment


            • #7
              Your trading looks very impressive Corni, keep up the good work!

              Comment


              • #8
                corni
                Is the graph the result of a Monte Carlo simulation?
                Are backtests made with a 99% modeling and with what database ?

                Comment


                • #9
                  No it is not a MC simulation. It is the equity curve of the actual portfolio backtest.
                  The backtests were performed with Tick data suite using tick data from Dukascopy. I also used 5USD commission and 150ms execution delay to simulate slippage.

                  Comment


                  • #10
                    corni

                    Today many breakout systems (such as SFE Price Action, SafeHaven, TitanTrader) have traded on eurusd and gbpusd where there has been a strong upward movement.
                    How come your ea did not do eurusd and gbpusd operations instead ?

                    Comment


                    • #11
                      Usually that would be the case if the movement was not within the period allowed by the time filter. But I am not 100% in this case. I will check it tomorrow.

                      Comment


                      • #12
                        I can confirm that it did not trade because of the time filter. The strong momentum that would have triggered otherwise was at 12:00 GMT. Yesterday I was a bit confused because I thought Darwinex used GMT+0 and it showed the strong momentum at 15:00 (so Darwnex is actually using GMT+3). At 15:00 my system would have triggered trades because of the speech of FOMC Member Bostic.

                        True, it is a drawback of my system that it will miss some of the opportunities compared to continuous momentum strategies, but because of this filter it will also miss a lot of false breakouts and therefore limit the risk/drawdown in periods without clear trends. SFE had a very bad period from Aug 2017 until the recent recovery because of many reversals, many of which my system has been able to filter out.

                        I am not saying that you should put all the money in my system, but I would recommend to use two (or more) momentum strategies at the same time to have a good diversification.

                        Comment


                        • #13
                          Originally posted by corni View Post
                          I can confirm that it did not trade because of the time filter. The strong momentum that would have triggered otherwise was at 12:00 GMT. Yesterday I was a bit confused because I thought Darwinex used GMT+0 and it showed the strong momentum at 15:00 (so Darwnex is actually using GMT+3). At 15:00 my system would have triggered trades because of the speech of FOMC Member Bostic.

                          True, it is a drawback of my system that it will miss some of the opportunities compared to continuous momentum strategies, but because of this filter it will also miss a lot of false breakouts and therefore limit the risk/drawdown in periods without clear trends. SFE had a very bad period from Aug 2017 until the recent recovery because of many reversals, many of which my system has been able to filter out.

                          I am not saying that you should put all the money in my system, but I would recommend to use two (or more) momentum strategies at the same time to have a good diversification.
                          So what session does it trade during?

                          End of London/Start of NY? Ah wait no it's linked to the economic calender to when CB's speak only, yes? That's quite a different/cool take on things if so.

                          What countries' CB's?

                          Comment


                          • #14
                            Correct, as you wrote it has no specific session. It starts Monday morning and closes potential open positions Friday 20 GMT. During that time it will only look for entries after important events and speeches for currencies:

                            USD, EUR, JPY, GBP, CAD, AUD, CHF

                            Somehow NZD events were not that great in backtests, but I might have to run some more analysis on that. CHF also had only very few trades in backtest. The 2015 EURCHF crash was not traded because my EA has a "maximum allowed movement" variable that filters out such extreme price movements.

                            Be aware that I am usually sleeping during AUD events, so if you subscribe to my signal you have to trust the EA to do its job alone or monitor the positions during those events yourself.

                            Comment


                            • #15
                              So price action this week was not very helpful. This strategy works best if there are some surprises in fundamental news, but the perception of central bank positions did not change much this week. So most trades were closed break even.

                              The equity decline came mostly from trading USDCAD on Wednesday, which had the worst-case scenario where first one direction triggers then the other and both end in stop loss even though in the end the initial direction would have been correct. But it is hard to automatically judge where to go from this price action. I will check on historic backtest whether it is generally better to wait a bit longer after news on USDCAD before entering a position.

                              But overall such periods were normal in the past. As long as we get enough surprises/trends in a given period the strategy should stay profitable.
                              So have a great weekend.

                              Comment

                              Working...
                              X