Sometimes strategies that have been successful for years suddenly stop working. This can be caused by a number of different fundamental changes in the market.
When strategies stop performing, many times we are left scratching our heads trying to figure out why. It is important that we are aware of this possibility and guard our capital closely in these situations.
Tim Black from Winner’s Edge Trading published a piece last week discussing a Forex strategy called Asia Mirror that used to be incredibly profitable for him. He built an Expert Advisor around the strategy and was able to take consistent profits out of the currency markets for multiple years before the strategy suddenly stopped working.
In his post, Tim suggests that there might be something to gain by breaking down the strategy to analyze why it stopped trading profitably. His hope is that a team of quantitative traders can determine the cause of the strategy’s failure and perhaps find an adjustment to fix the system.
He starts with a brief background on the logic behind the strategy:
During the slow time between the close of the New York session and the open of the Euro zone session, we anticipate that the market will drift in a retracement or “mirror” of the prior day’s trading.
We use this drift to set ourselves up for the Euro zone and New York traders to push the price back in the same direction as the prior day.
Ground Rules
Then he continues by listing the ground rules that the strategy follows:
- We define the “prior day” as 5 pm yesterday to 5 pm today, NY Time (Winter= GMT-5, Summer=GMT-4)
- There must be at least 20 pips of movement from open to close on the prior day to provide a definitive direction.
- We enter the pending positions 10 minutes after the Tokyo market open unless there is news at a later time. If there is later news, enter positions 5 – 10 minutes after news release.
- The price action must not be in or near the “Strike Zone” when we enter the trade.
- We take the trade Monday – Thursday nights (NY Time.)
Trade Execution
Tim then discusses the execution of a trade:
- Mark an hourly chart with vertical lines at 5pm yesterday and 5pm (all times are NY Times) today.
- Measure the distance in pips between 5 pm yesterday’s open (NY Time) and 5 pm today’s close.
- Pull your Fibonacci in the same direction of the open-close from the high of the day to the low of the day. In other words, if the day closed higher than it opened, pull the Fibonacci from the low of the day to the high of the day. If the day closed lower, then pull Fibonacci from the high of the day to the low of the day.
- Mark the “strike zone” – the area between the 38.2% Fibonacci level to the 61.8% Fibonacci level.
- Calculate your position size (we’ll discuss that a little later.)
- Place your pending orders a few pips ahead of the 38.2%, 50% and 61.8% Fibonacci levels in the direction of the prior day.
- Set your stop loss on all your pending orders 4 or 5 pips behind the 78.6% Fibonacci levels.
- Set your take profit on all your pending orders 4 or 5 pips ahead of the prior day’s low/high (depending on trade direction.)
Remove any unfilled pending orders around 9 am NY time on the next day. If you get one or two positions filled and they hit the target, close the unfilled pending orders at that time.
The Most Consistent Currency Pairs
Tim also covered which currency pairs worked best with this strategy:
The pairs we used initially for this strategy were:
AUD/JPY, NZD/USD, CHF/JPY, GBP/USD, NZD/JPY, USD/CHF and USD/JPY.
The most consistently profitable pairs were GBP/USD and NZD/USD.
The Discretionary Element
Tim continues from there by explaining that there is a discretionary element to the strategy when multiple trades are triggered at the same time. Traders may be forced to choose between one trade or another. They may also have to determine if it is worth it to exit one trade when another is signalled.
If we could find a way to make this strategy profitable, we could likely develop rules to avoid this type of discretionary judgement as well.
Risk Management
The post also spends a great deal of time breaking down the risk management of the system. Basically, the goal is to risk 2% of capital on every trade. Tim does a great job of explaining exactly how to figure this out and convert currencies if necessary.
Of course, the big problem here is that the strategy is no longer profitable. It doesn’t have the same edge that it used to have. If Tim and his team can determine why that is, they may be able to correct whatever has fundamentally changed about this approach.
Jean-Francois says
I believe that a strategy that is based on other human behaviors indeed will eventually fail because others will learn of it and make it valueless. A strategy based purely on indicators has much better chances of success in the long run because the definition of indicators do not change.
Andrew Selby says
I understand where you are coming from, and I prefer to use indicators as well in my own trading. However, I was quite fascinated by the concept that these guys are working on. I doubt it would ever fit with what suits me personally, but I thought it was worth discussing either way.
Harm Bodewes says
I fully agree. Indicators don’t discriminate between stocks, bonds, Forex, etc. They are just working with data. I guess you have to do something with volatility in conjunction with risk management & stop loss but than I think your chances are higher to have a more stable and wide range EA.
But I must say that I respect these guys. They seem to know what they are doing and have build a profitable EA which has been working for years. Nothing is lasting for ever on this earth. just time for an update. Good luck!
Andrew Selby says
Thanks, Harm!
MTLibrary says
Great Commentary! A great strategy from my own experience has the following in common. It’s repeatable. Can be used on any instrument – and will always result in the same end result. When you figure out how price works (why it stops where it does to then turn and reverse for example) – you will then understand the rhythm found in all markets. It may take you 8 years to find it though. Most traders stop trading after 3 years (the majority of your so called 95% that lose).
Repeatable and always the same result!! – yes possible. If trading like so is not possible, then a trader must ask the question why? The key is when a trade fails, go back and find out why it failed. When you think you found the reason, you then apply the new rules again, and repeat as needed.