Many of us regularly spend long hours looking at price charts trying to get some clue what the hell the market is doing now and where it will move next. But how much do we actually know about the market? Most traders tend to rely on some subjective conclusions based on their own trading experience and/or dubious wisdom picked up on various forums and chats. Only few traders make attempts to analyze the market basing on some well-defined criteria, and from those who try, only a few get any useful results that can be applied to the actual trading.

Today I will present an example of the market analysis that might be useful for anyone who trades any kind of grid martingale system. Moreover, I will provide a tool that can be used for further investigations of any currency pair you might wish to analyze. It might be that your analysis will be more interesting than mine. Let me know! I will be glad if you share your conclusions and observations on our forum.

My analysis today will be done from the point of view of a grid martingale trading. The grid martingale strategy is quite popular; many EAs use different variants of this trading system. Good examples of such are the commercial EA Forex Warrior and the freeware/donateware EA Fiddler. Experienced users of such EAs know well that some currency pairs are better suited for grid martingale than the other. In particular, GBPUSD is widely considered to be a good choice whereas the JPY pairs are considered to be rather dangerous. Is this really so? And what about other pairs? What parameters are better suited for different pairs? Today I will show you the instrument that will allow you to get answers to these and other questions yourself.

Specifically for this article, I wrote an Expert Advisor **StatCollect**. This EA does not open any trades but collects information about the price movements during a given period of time. **StatCollect** has just two parameters: the grid step and the take profit. When started, EA opens a virtual order, e.g., a long. If the price moves against the order by the grid step, a new virtual long order is opened. If the price moves in the direction of the order(s) and reaches the take profit of the last order, EA stores the number of the grid levels in the basket and starts a new basket in the same direction. In the end, EA will be able to tell us how many baskets of long orders with different grid levels the price has made during a given period of time. In order to complete the picture, EA simultaneously “trades” both in the long and the short directions and reports the total number of baskets opened with a given number of grid levels.

So, the settings of **StatCollect**:

• **StepSizePips** – grid step,

• **TPPips** – takeprofit,

• **Buy (TRUE/FALSE)** – trade long,

• **Sell (TRUE/FALSE)** – trade short,

• **nBasket0** – number of basket levels (EA prints out the number of baskets with number of levels >= nBasket0).

Expert Advisor **StatCollect** is available for free downloading in the end of this article.

Now, let us back-test StatCollect on some currency pair. After the test is completed, EA creates a text file named “StatCollect.currencypair.txt” in the tester\files\ folder. For example, backtesting EA on GBPUSD with grid step 30 pips and takeprofit 40 pips over the period of 3.5 years, we get the following report file:

In the above report, we see how many baskets with different grid levels we got, the total number of baskets (4214), the mean square average of the basket level (3.07), the probability of baskets with different grid levels, and the probability of a basket with 11 or more grid levels (0.38%).

Now let us repeat the same back test for different currency pairs, with the same parameters and the same time period, and analyze the results. First, we can compare the total number of baskets for different currency pairs. The total number of baskets is connected with the volatility of the currency pair (but not only). It is also directly connected with the total profit we can get with from this pair (as each closed basket brings some profit into our pocket). Compiling the results, we obtain something like this

This picture shows us the potential profitness of different pairs in grid martingale trading. This is a useful information. However, an experienced trader knows that just profit has little meaning in trading. What matters is a profit-versus-risk balance. So, the next question we need to address is: “Which pair is safer to trade?”.

Let us now compare the probabilities of baskets with different grid levels for different pairs. Plotting the probability against the grid level for several of our results together, we obtain the following picture.

An interesting feature of this comparison is that the probability of baskets with 1-3 levels is very much similar for all currency pairs. On the contrary, the probabilities of large baskets are very much different.

In order to make the comparison between pairs easier, I decided to select just one parameter for each pair. Which parameter is the most important for a trader? It is the probability to get a large basket beyond a certain level threshold. Below I plotted the probability of a basket with 11 or more levels for different pairs, which can be interpreted as a “risk level” for the pair. The ordering of the pairs is the same as on the figure with the number of baskets: the most profitable pairs are on the left, the less profitable are on the right.

0

Now, it is getting interesting. What do we see? Clearly, the two best performing pairs are GBPUSD and EURUSD. We see also several low-volatile pairs with low risk level: USDCAD, AUDNZD, AUDCAD. It is quite surprising to see that GBPCAD, which is often considered to be well suited for martingale trading, shows in fact a very high risk level. Predictably, most pairs with JPY show high risk levels too.

Summarizing, in this small and simple analysis I obtained quite a clear picture of suitability of different currency pairs for a grid martingale trading. Most of the conclusions agree with the “common knowledge” known to most of experienced martingale traders. However, some of the conclusions, in particular, about the high risk of GBPCAD, were quite unexpected. I think I learned something from this small exercise. I hope that you learned something from it too.

Author: Vladimir aka loopsider