How to Manage Your Trades Part 4: Managing Systematic Pullbacks
In today’s blog post we continue our exploration of trade management techniques and their impact on your bottom line. In case you missed the first three installments, you can find part 1 here, part 2 here and part 3 here.
In today’s post we are going to adopt a simple pullback strategy described by Andrew Abraham in his “Trend Following Bible”. Andrew Abraham is a systematic trend-following CTA, so we are going to use his setup on a combination of FX Majors, Crude Oil and Dow and applying 5 different management techniques on the same trades.
Once again, you will notice just how much the management can impact the bottom line. But we also uncover some other “nuggets” in the process.
Abraham Systematic Pullback Rules
Abraham made his rules quite clear. Essentially we are playing daily pullbacks into a weekly trend.
Here is the weekly chart of EURJPY, with the MACD applied. We will use this for the following execution screenshots (Daily charts).
If going Long:
Weekly MACD (Standard settings) is above zero and increasing.
On the Daily time frame we then wait for a pullback using a retracement oscillator. Abraham utilizes the Elder Force Index but since it utilizes volume (which in the FX spot market is not attainable), we are using an RSI instead and there is no statistically significant difference in the results anyhow. when it crosses below the zero line
Place a buy stop one tick or 0.001% above the high of the prior two bars
In the chart below, the green lines are where long trades would have been placed. I added the weekly 26 and 12 moving averages on the chart in order to give an idea of the weekly MACD stance. Evidently, the 12 must be above the 26 and both must be sloping up in order for the MACD to be positive and increasing (for longs).
If going Short:
Weekly MACD is negative and below the zero line.
Daily time frame there is a pullup via the retracement oscillator when it crosses above the zero line
Place a sell stop one tick or 0.001% below the low of the prior two bars
Regarding the stop loss:
If Long: One tick or 0.001% below the two-bar low
If Short: One tick or 0.001% above the two-bar high
However, Mr. Abraham did not specify what multiple of the 39 period ATR to use, so we have gone with a simple Supertrend Trailing Stop because it is a trailing ATR level, albeit from the candle mid-point. Our work tests various ATR multiples.
Tests were performed in MT4 on daily charts of WTI (2010-2018), US30 (2009-2018), AUDUSD, EURUSD, GBPUSD, NZDUSD, USDJPY (currencies 2001-2018).
Initial Setup Assessment
Once again, this work would not have been possible without our programmer Tony, who was kind enough to put in some hours through the X-Mas vacations in order to get these (& other) tests done. If you would like to either code or test some systematic rules in either MT4 or Ninjatrader environments, feel free to reach out to him at [email protected]
Using Profit Factor from MT4’s report for quick assessment, all with trailing SuperTrend(2,7).
It turns out that the MACD filter on the Weekly chart produced the best results. The original rules specified MACD > 0 and rising (for long), so it’s not surprising that’s the best filter. Also very good is the weekly Momentum Market Type > 0 (for long).
To have an idea regarding the frequency of this kind of Pullback, there were 26 trades on the Dow (so just over 2 per year on average), 59 trades on the AudUsd (around 15 trades per year) and 73 trades on USDJPY (around 18 trades per year). Naturally, this is a long-term systematic strategy so it does require diversification across asset classes in order to work. Actually, Mr. Abraham did suggest trading the following basket of instruments:
- Currencies: Australian Dollar, British Pound, Canadian Dollar, Dollar index, Japanese Yen, Swiss Franc
- Agriculture Commodities: Corn, Cocoa, Cotton, Soybeans,
Soybean oil, Soymeal, Sugar, Wheat, Livestock Commodities, Lean hogs, Live cattle
- Metals: Copper, Gold, Silver, Energy Commodities, Crude, Heating oil
- Stocks: S&P500, Nasdaq
- Bonds: Eurodollar, 10-year bond, 30-year bond
What we have found out (which is also mirrored in Robert Carver’s book Systematic Trading), is that diversification is the only free lunch available. Not only should strategies be applied to multiple markets, but there is room to diversify within the same strategy (more on this later).
Trade Management with Abraham’s Pull-back
From this initial test of Abraham’s rules, we can see a few things:
- Taking profit at 2R, despite being advantageous on paper, isn’t feasible in real life without additional filters. This gives more substance to our argument from the first article in this series: set & forget might help overcome psychological biases that usually work against us, but we cannot count on the mathematics of risk:reward alone if we want to profit in the markets. In reality, positive results over time are a product of market type + risk:reward + win rate.
- The best management technique for this pullback model (which has a built-in trend filter) are naturally trailing stops and not fixed targets.
- Between the 2-bar trailing stop and the SuperTrend(2,7), the latter is probably better because it leaves more room for pull-backs.
Just to refresh your memory on the other management devices we used:
- Mom 2: We told the system to exit any trades when the day showed slowing momentum. To us that means ATR < 80% of it’s 7-day average value and/or a neutral close.
- 2-bar: The second method turned out to be a bit smarter. It’s a variation of a trailing stop which follows momentum along with market structure. We trail the stop below 2-bar highs (if selling) or 2-bar lows (if buying) and the stop is trailed ONLY if we make new highs. Essentially we’re trailing the stop but at the same time trying not to give too much back when the market starts to run out of steam.
Since we’re dealing with pullbacks within a broader trend, it shouldn’t be too much of a surprise that trailing stops prove most effective. In trends, the key is to let your profits run. When the profits are capped at fixed targets, the system can’t make up for the (usually low) win rate it has.
Now that we know the Supertrend (2,7) is the most adequate management vehicle, let’s go back to see what happens if we attempt to filter the market trend in different ways.
With proper management, the equity curves are “fairly” similar, which means that they are identifying broadly the same trends at the same moments in time. So this is also instructive: so long as there is a trend in place, there isn’t one single best way to identify it! If you’re spending time searching for the “best trend filter” or something fancy, then please stop! It’s a waste of time. If the market isn’t moving, it’s not moving – period. If it is moving, then any trend filter will pick it up.
The interesting fact is that the SQNs (which in our case is the same as a Sharpe Ratio) are quite a lot higher than what systematic traders like Robert Carver expect from single models. Carver said that each single model usually has a Sharpe Ratio of around 0.2. Well, we’re up to 0.55! However, a model with 0.55 isn’t yet tradable with confidence.
This begs the question: what if we diversify within the same model? By diversifying the filter (MACD vs. Weekly MT) and the time frame (3.0 vs. 2.0 ATR in the trailing stop), can we obtain a signifcant increase in SQN?
It turns out that even the same model, with small changes, does in fact add diversification. Of course we wouldn’t trade this combination because we’re doing the same thing in 3 similar ways – there’s not much common sense in this approach. But imagine being able to combine systems that identify different behavioual traits and combine them together! The diversification effect would be huge.
Diversification really is the only free lunch – as Ray Dalio and other gugus have said for ages. We’re simply going through the numbers and finding out how accurate that statement is.
Over to You
We continue to find the same pattern over and over again: trade management makes a huge difference. It can transform a volatile equity curve into a viable risk-adjusted curve. However, it takes experience to know where to look and what venues to explore, and this is the “art” of trading. There are always trade-offs to consider and without a tad of experience, it is difficult to not get blinded by rogue results.
Our conclusions this time round do not differ significantly from Part 3:
- filtering for trends is necessary, with this pullback setup;
- trailing stops are better than fixed targets in trending markets;
- diversification benefits are cardinal when applying a systematic model.
One thing to add is that we did “play around” with the inputs to see whether we could squeeze any more juice out of the pullback model and the fact is that the trend filter overrides everything. This simply means that for any given strategy, there may be a physiological “cap” to the performance in any given month/quarter/year contingent upon the trends (or lack trends) in that timespan.
About the Author
Justin is a Forex trader and Coach. He is co-owner of www.fxrenew.com, a provider of Forex signals from ex-bank and hedge fund traders (get a free trial), or get FREE access to the Advanced Forex Course for Smart Traders. If you like his writing you can subscribe to the newsletter for free.
The post How to Manage Your Trades Part 4: Managing Systematic Pullbacks appeared first on FX Renew.