Compression trading

Trading Systems
1

In my Tickmill UK review thread I was asked about my trading system. Rather than derail that thread, I thought I’d post an overview here.

From Dr Elder’s The New Trading for a Living (emphasis mine):

Chaos Theory has achieved prominence in the recent decades. Markets are largely chaotic, and the only time you can have an edge is during orderly periods. In my view, markets are chaotic much of the time, but out of that chaos, islands of order and structure keep emerging and disappearing. The essence of market analysis is recognizing the emergence of orderly patterns and having enough courage and conviction to trade them. If you trade during chaotic periods, the only ones to benefit will be your broker, who’ll collect his commission, and a professional day-trader, who’ll scalp you.

The key point to keep in mind is that once in a while a pattern emerges from chaos. Your system should recognize this transition, and that’s when you should put on a trade! Earlier we spoke about the one great advantage of a private trader over professionals—he may wait for a good trade instead of having to be active each day. The chaos theory confirms that message.

This is probably the most profound thing I have ever read about trading, and it underpins my entire strategy. I look for orderly periods, and open positions based on the statistically probable consequences of that order. There are many ways to do this, but my primary tools are Bollinger Bands.

Overview

I use a basic Bollinger Band squeeze strategy to find periods of low volatility (ie, order) and then place pending orders on either side of the price. When the breakout occurs, my pending order triggers, capturing the price movement. Note that Bollinger Bands are not the only way to do this; there are other volatility indicators, such as envelopes and the ATR, which could just as easily be used. However, I dislike off-chart indicators, and I find Bollinger Bands are visually very effective for me.

I’m trading the EURGBP, GBPCAD, GBPJPY & GBPUSD H1 charts, typically the London session, simply because it’s my local timezone. I don’t hold positions overnight, and dislike holding positions for more than 3-5 hours. I’m a proponent of knowing a handful of pairs very well, and these are the only pairs that I currently trade. They are all correlated, which means I’m only ever trading one of them at a time.

My chart

This is an example from earlier this week - my first trade of the year actually, and my first trade with my new broker.

GBPUSDH1.png800×600 18.6 KB

There are two sets of indicators on my chart - those with a solid line, and those with a dashed line. Solid line indicators are trading signals. Dashed line indicators are trade/risk management signals.

I get a positive trade signal when the Bollinger Bands (blue) contract to within a certain distance of the price average as marked by a set of envelopes (green). This signals a valid price compression, or a moment of relative order in the markets.

Once I see this signal, I can set pending orders on either side of the price. The inner gray line is my entry point, and the outer gray line is my profit target. The red line is my stop loss. I continue to tweak these levels, set by envelopes in reference to the average price, to improve my results. When I first started, for example, my stops and my profit targets were much wider, forcing me to stay in trades far longer than I liked, and with only a 1R ratio. Now my stops and targets are much closer, so my trades finish faster, and my ratio is up to 1.5R. I expect to continue to improve this over time.

In this chart, one of my positions has been triggered, and the unused pending order has been deleted. But it was essentially a mirror image of the triggered order, which ended up giving me 27.7 pips over the course of about 4 hours.

Risks

There are two primary ways to go wrong with this strategy.

First, you can have a valid breakout start in one direction, but then reverse in response to a news event. You minimize the risk of this by not trading around scheduled news. You are looking for periods of order, and want to avoid anything that is likely to increase the chaos of the markets.

Second, you can have a false breakout that nevertheless reaches your stop order, triggering your position, before reversing and stopping you out, forcing you to absorb a loss. While you can mitigate this by moving your entry point farther out, the larger your spreads, the more likely this is to happen. Moving to Tickmill’s low spread, commission based ‘Pro’ account was all about reducing the risk of this second issue.

Stats

As the system is a work in progress, take all stats with a grain of salt. However, over the last three months I have averaged:

  1. 12-18 triggered positions per month
  2. A hit rate ranging between 40-60%, depending on the month
  3. All three months were quietly profitable

As with any good system, the overall strategy is simple, but fine tuning your entries, stops, and profit targets makes a huge difference in the final outcome. As I continue to refine the system, I hope to improve my results without altering the underlying mechanics.

6 Likes
2

Thanks very much for starting off such an interesting and promising thread.

I used to know someone, a long time ago (an institutional trader, in France) who used a very similar system but on faster charts.

Do you mean that the level of the outer grey line at the time of opening the trade is your target, or that the target is the position of the grey line wherever/whenever the price reaches it? I ask because (for example in the case of the short trade you’ve shown on the chart above) the grey line is going to move down somewhat as the price approaches it, isn’t it?

So I’m wondering whether your target is static or dynamic?

Again, thanks for the thread, which I’ll be following.

2 Likes
3

It’s a static target. I determine my stop and my target at the time I set my pending orders. If a squeeze is prolonged, and those indicators shift somewhat, I might adjust my pending orders to compensate (although in practice this rarely happens, as price is usually flat during a compression), but I don’t changed them once an order has been triggered.

That’s why they are in light, dashed lines - they are only relevant when placing my pending orders, and I don’t want to be distracted by them the rest of the time. They just save me calculating distances and positions manually.

2 Likes
4

Thanks - that makes complete sense.

This method’s certainly based on something realistic, and it’s no stretch to imagine that it can have a decent edge, though as you rightly said, how decent is always going to depend on fine tuning your entries, stops, and profit targets.

Have you tried it on any faster charts? I’m wondering whether increasing the trading frequency might help, overall, even though the win rate might decrease slightly.

2 Likes
5

I settled on the H1 simply because I work full-time, so I can set my positions in the morning when compressions are most likely, following the Asian session, and then just monitor the trade on my phone. I can’t see any reason that it would not work on faster (or slower) charts, although the settings would need to be changed.

2 Likes
6

I like it :sunglasses:

Do you have any stats as to how many of your trades turn out to be “fakeys” where they turn around and push higher after "sucking you in " ?

I’m also wondering whether you could bet twice as much and close a half as you do now, but let the other half run on to say 35- 40 pips ? I guess “stopping it out” could be a problem here unless you used perhaps a trailing stop loss for that part, if you need to be out whilst it is active.

Just a thought.

2 Likes
7

That’s an interesting question, and I’ll review my trade journal.

I have experimented with a more open ended approach that essentially uses a trailing stop, but the results were less consistent. Because my availability to manage trades varies, fixed targets makes sense for me.

Saying, ‘I want to be a full-time trader’ is practically a cliche, but I would love to be able focus on the markets throughout the day.

1 Like
8

This is the kind of situation where an element of discretion comes into play.

GBPJPYH1.png800×600 18.8 KB

My position had been triggered by a bearish impulse move that came tantalizingly close to hitting my target within the first two hours. Then price reversed to around my entry point and started moving sideways for a couple of hours.

My options were:

  1. Accept the impulse move is over and the eventual direction is outside the scope of my strategy. Close the trade while in profit.

  2. Accept that the markets do not move in straight lines and that the move may well not be over. Leave the order in place.

  3. Accept that I can’t know either way, and split the difference. Move my stop up to just inside of ‘profit’ and let things play out, knowing that my risk is now zero.

I went with option 3), and eventually the bearish move resumed, handing me 32.6 pips. Chalk this one up to the advantage of trading only a few pairs; I would probably have made a different call on one of my less volatile pairs, but the GBPJPY can be particularly explosive. I also dropped down to the M5 when considering my options, so I could get a clearer look at how the price action was playing out.

2 Likes
9

Just a quick update. Finishing up the month, my adjusted trading plan produced 12 trades with a 75% hit rate (9 wins & 3 losses). Obviously, this is a small sample size, but it’s still encouraging.

That puts my total profit for the month at 9*(1.5R) - 3*(1R) = 10.5R. I believe I can improve on this by adjusting my profit target, which does not seem ambitious enough.

Although this is fewer trades in a month than my previous average, it is a much higher hit rate, suggesting that I have managed to exclude primarily losing trades, which is fine. In answer to @Falstaff, only one of my three losses this month is what I would consider a fake-out. One was simply a trade that started moving sideways (and then collapsed) before reaching my profit target, and the other was an avoidable loss.

It was the day of the BoC rate announcement, and I had a trading signal on the GBPCAD. I expected to be in and out of the market long before the announcement, but what I had not appreciated was that the market was unwilling to chose a direction prior to the announcement, so price action was full of spikes, but flat overall. I was caught by one of those spikes and suffered a loss. In retrospect, I simply should not have been in the market on that currency on that day.

1 Like
10

i have a question and a suggestion;

Question: what is your moving average band setting ?

suggestion: maybe you should split your trade into two trades, the first one your’ll set your usual tp and extended it for the 2nd one, and breakeven on the this latter when the TP1 was hit

11

I’m being deliberately vague when comes to settings. Partly, that’s because I’m not trying to ‘sell’ my system to anyone - I’d rather people innovate, and then come back and teach me a thing or two - and partly that’s because the numbers keep changing as I tweak and review my trading plan. But it’s also because I really don’t want to be responsible for anyone else’s trading decisions.

Of course, anyone who really wants to know can just take a look at my charts and tweak their settings until they match.

Your suggestion has been made before, and it is a good suggestion. In near term, though, that requires more trade management than I can reliably devote to my trading while working full time. If I am ever able to devote more time to my trading, though…

2 Likes