Multi Timeframe Forex - strategy details and results

Hello,

I am a leader of Multi Timeframe Forex strategy. I start this topic for everyone interested in this strategy. In this thread I would like to provide you with details regarding my trading approach as well us some statistics and trading results from the past.

I am trading using this strategy on my live account since Jan 2015. I have 100% TOS badge, so all the trades from my live account are replicated. The algo was not modified since 2015, all trading decisions are done by algo itself. Strategy performance is monitored on a daily basis by a team of traders.

Strategy characteristics

  • Instruments E/U, E/J, U/J (higher volumes) A/U, A/J (lower volumes).
  • Low to moderate margin use
  • Doing 1-2 trades per week on average
  • Short to Medium term trend following strategy
  • All positions have Stop Loss to protect the value of the account
  • Position scaling - geometric. When assigning position size for a trade, the strategy takes into account average account balance from the last 4 weeks period. This is a smooth profit capitalization algorithm.
  • Trading 24/5 between Monday 6:00 am GMT to Friday 8:35 GMT. Trading disabled during Christmas - New Year period and Easter Period, due to low market liquidity.
  • Closing position before the end of the week.
  • Strategy operates well in both low and high volatility environments. (No trading in extremely high volatility above levels of 2008-2010 period, as it could not be backtested).

Strategy goal
To deliver steady risk adjusted returns for 10+ years period in every trading conditions

LIVE TRADING RESULTS - Interactive Brokers Account

Chart below is generated by SmartTrading application - an in-house application to design, test, analyze and trade with algo strategies which operates with Interactive Brokers API.

Basic statistics for the period 2015/01/01-2019/07/12 (including trading fees and commissions):
CAGR = 22,58%
Maximum draw-down = 10,2%
Chart is showing unrealised + realised cash balance with 5 minutes time intervals.

More to come:

  • Key Alpha driver of this strategy. Low Profit Sensitivity to changes of values in strategy parameters - means that risk of this strategy is very well controlled and is immune to over-optimization (one of the key risk factors in algo trading).
  • Examples of heatmaps presenting backtested results for a set of strategy parameters.
  • Results for period 2004-2015, additional charts and results per currency pair
  • Backested results charts and heatmaps for other currency pairs (to show why we trade those)
  • Additional statistics for the periods (Sharpe Ratio etc.)

Pricing - 59$ for everyone subscribing this year. 1 month free trading coupon available for everyone interested.

Have questions? - Shoot here, or contact me via private message.

MS

Hello,

Just a quick post to share some more backtest results with you.
The first one that I shared was live trading one. This one consists of two pieces.
2013 01 to 2015 01 is a backtested result done using IB data (real tick data that IB is providing for live trading)
2015 01 onwards is a live trading data that you have already seen. Both joined into one chart to give you a better perspective. Using tick data from the broker which you trade is the most accurate for backtesting.

CAGR = 19,76%
Maximum drawdown 2013 - now - 11,1%

More backtested results to come on longer periods from tick data from other data providers.

Don’t be shy if you have any questions.

More details to come soon.

MS

What is the minimum percentage scaling and minimum capital required to trade your system - if at all possible?

Refreshing and interesting to see a very different trading style than so many strats here. The few published trades here take a single big position at the opening, and then use many small trades to get out of the big position. Only one of the trades added to the original position, and this one did average down a little; well, actually it averaged up on a short position, just two minutes after the first short was filled. But nothing at all like the Martingales we see around here from time to time. A fine looking start.

Hello v1Trader,

Thanks for your kind words,
Indeed the strategy works that it takes a position according to the market direction, and only rarely adds up to the position (this happens mostly when trending is strong). Then it looks for best moments to unload and book as much profit as it can. In case when trade goes the other way the strategy will also try to unload partially to minimize losses (unless SL is hit).

MS

1 Like

Hello Traders,

Providing you additional info about HEATMAPS - key concept and tool utilized in Forex Multi Time Frame (MTF) strategy. Below you can see an example of heatmap presenting average trade value for EUR/JPY pair based on 2004-2014 historical simulation period.
Few hints on how to read it:

  • There are two key parameters in MTF strategy K(X axis) & RH (Y axis).
  • Each tile on the heatmap represents a value of average trade that the strategy achieved during
    historical simulation for given values of parameters K & RH.
  • The warmer the color is the higher the result.
  • The highest value is $5132,31 was achieved for K=4, RH=85, the lowest equals to $1540,39 for K=3,25 RH=78.
  • Such heatmap can be generated for any type of statistic which might be relevant (Sharpe Ratio, Max Drawdown, etc.)

How do we use it in our project?
We analyzed dozens of strategy with this tool and it allowed us to identify the best concepts.
What we were looking for was obviously a positive steady return with low drawdowns in long term.
But the most important thing is to have such result for a wide spectrum of parameters.
In other words, no matter which parameter set will you choose, the strategy perform well.
Some of the combinations will be slightly better, some slightly worse, but this is just a how statistics works. Anyway even the worst case parameter combination here is still pretty decent (positive average trade of $1.5K)

For a given currency we are using multiple sub-strategies, each one with different combination of parameters trading independently. So all the EUR/JPY trades are a total result of multiple sub-strategies. We do this because it is not possible to predict which pair of parameters will be best in the future, that changes and depends on multiple market conditions. What we know however is that any of them should in long term generate positive returns. This is our insurance for over-optimization problem. We just need to ensure that parameters provide relevant information to the strategy. A good parameter will be valuable in wide value range (given that the range is logical).

Another example of a heatmap, this time for Sortinio Ratio, below.

Happy to hear questions if you have any.
If you like the concept, please subscribe to Multi Time Frame Forex strategy.
Current Price $59 valid until the end of 2019.

More information will come soon in this thread so stay tuned.

Happy to answer to any of the questions.

MS

I have rescaled Multi Timeframe Forex strategy.
Rescaling multiplier of 15% has been used.
This is to make it easier for smaller accounts to subscribe.

MS

Hello,

Some more backtesting data for you for Multi Timeframe Forex Strategy
Time period 2004-01-01 to 2013-01-01 (9 years)
Max drawdown 15.89%
CAGR 19.06%
Currencies tested: EUR / USD, EUR / JPY, USD / JPY, AUD / USD.
AUD JPY - which is used in the live version is not included here, due to some issues with test data. (I will try to solve it soon and post an updated screen).
As you can see there are no excessive drawdowns during high volatility time, as the trading style adjusts to the volatility level.

Good luck with your trading!

MS

I have rescaled back to the previous equity size for Multi Timeframe Forex.

Below I have prepared table with proposed scaling for different investor’s risk profiles and equity size as per today. I will try update it as the equity of strategy changes.

Dear subscribers,

A short update regarding results of MTF strategy during recent week and some outlook for the next months.
We managed to capitalize substantial profits from directional moves that took place at the beginning of August. Since then we were looking for continuation of those moves, however market stalled in new ranges and any further stronger moves were chaotic and quickly faded, causing some minor draw-down on our strategy.
Despite that our outlook for the next months is optimistic, since medium term volatility picked up slightly and we expect some stronger trends to develop in the coming months. This is our base scenario. Should that materialize the strategy will be in a good position to continue generating significant profits.
Alternatively the market will continue to be ranged for some more time. This is our my opinion less likely as market sentiment is shifting due to rising uncertainty related with FED and ECB actions and Trade War developments.

In the meantime our results are improving. We are proud to notice that only after 3 months we managed to climb to a pretty decent position in the C2 leaderboard, being classified as #55 among all strategies and # 7 in forex sub-class. We fill that there is a substantial room for improvement for us and we will try to challenge the top C2 performers in the coming months.

In terms of pricing we will keep the current price $59 price till the end of the year and we aim to deliver great value for that price.

Flexible scaling mechanism on C2 allows you to subscribe to C2 strategy even with $10k on the account. For more scaling information, please check this thread or contact me directly.

MS