Weekly 80/20 Setups

Quoting Flyjetz

Disliked

This system has now lost five in a row, and six of the last seven. This is starting to hurt the statistics a bit, but I have not lost faith in the system. I have to remind myself that I've only been doing this live for four to five months now and I will continue to monitor how the system progresses over time.

Ignored

Quoting mathematician

Disliked

I found this thread last week while I was looking for a "set-and-forget" system. I must say that I think this is a great idea you have! After years of struggling to find the right system for me, I have slowly realized what kind of a trader I am and what I am looking for in a trading system that will suit me well. I think that this is it (or something very similar, based on this). I particularly like these features:

1. It's simple. There is no visual interpretation of indicators or chart patterns. It is simply a calculated projection of entry and exit prices.
2. It's set-and-forget. There is no waiting in front of the chart for something to line up. Just set the orders and check in occasionally. This, I find particularly important since it eliminates so much emotion.
3. I like the core concept: A linear projection based on the slope of the midpoints from the past 20 weeks. Very elegant.
4. There is room for error. You aren't looking to buy/sell tops/bottoms. You just need that middle 60% of the projected weekly range.

I have just started trading this system. I'm only working with a limited set of curency pairs, so far, but I caught about 250 pips on EUR/JPY this week, which was nice.

However, one thing that bothered me right away was the fact that there is no stop loss. Even though the TP exit target is updated every week, having no stop loss still leaves open the possibility of a huge loss. For example, if price turns against your position and never looks back for a long time. In that case, the updated TP effectively becomes a trailing stop in a negative trend. Nevertheless, I decided to trade this system as is, without a stop loss, for now because I want to really get a feel for this system and learn how it works in its unmodified state.

For the purposes of research, I will be keeping track of drawdowns on my trades. More specifically, I mean the maximum negative extension from the entry point that occurs before the trade closes. By gathering statistics on these drawdowns we might be able to determine a practical stop loss strategy.

I am also interested in trying different combinations of currency pairs. I will try to balance my positions and reduce exposure, as you do. Perhaps the system is just fine how it is, it simply need to be diversified by trading it with the right basket of currencies.

In closing, don't lose faith. Sure you have a string of losses, but every system does. Your track record from forward testing is relatively short, but what's there looks fantastic! You've made thousands of pips, that's more than most can say. I think this system just needs some tinkering in the back end with some solid risk management techniques; the front end is very simple and I wouldn't want to change that too much. I will do my best to research and develop some risk management strategies for this system. I'll let you know what I find out.

Ignored

Quoting Flyjetz

Disliked

As I've said before, I only post here on 10 currency pairs, but actually am trading closer to 25 pairs in my real account.

Ignored

Quoting Flyjetz

Disliked

Thanks for the kind words! Conventional FOREX education says you will get killed by trading without stop losses. I initially started this system with stop losses equal in size to the amount of profit I projected I would make. In other words, if the entry to exit was 200 pips, then the stop loss was set at 200 pips below the entry. Although I only traded this way for few weeks, I determined that this system needed a lot more flexibility and more time to allow the currencies to actually draw out the weekly candle. After I eliminated the use of Stop Losses, the success really exploded. I've pounded the table repeatedly on the importance of risk management and you mention it as well. If you are going to trade without stop losses, you need to be using a tight risk management strategy. I have been experimenting with different size trades from the beginning of this system. I've always used 50:1 leverage, and I started out with individual trades that equalled 3-5% of my total capital. I quickly determined that was too risky for me with this strategy. I cut back to 0.5% to 1% for quite awhile. Right now, I am trading 0.6% to 1.2%. This seems to be my "happy spot" for the moment. I feel that by trading a large number of currencies, you are protected to some degree of any one currency moving against you. As I've said before, I only post here on 10 currency pairs, but actually am trading closer to 25 pairs in my real account.

Ignored

Quoting mathematician

Disliked

I was wondering how you determine positions sizing. You said you are risking about 0.6% - 1.2% per trade, but since there is no stop loss, how do you know how many units to trade?

What I have been doing so far, is basing it off the average losing trade (according to your statistics), which was about -160 at the time. So, to play it safe, I rounded this to -200 pips per losing trade. Then I based my position size on that assumption, i.e. I wanted to risk 1% per trade, so I chose the number of units such that a 200-pip loss would be 1%. Is this what you do, base it off the average losing trade?

What about currency pairs that naturally have a larger daily range, such as GBP/JPY, GBP/CHF, and some of the exotics? Do you reduce the position size on those pairs? I suppose you could keep track of the average losing trade for each pair individually and size accordingly.

Ignored

Quoting Flyjetz

Disliked

First off, I use OANDA, which lets you choose any position size you want. It doesn't trade in standard "lot" sizes but uses "unit" sizes. Under my trader preferences, I can set the value of the standard trade at "X"% of capital. Right now I have it set at 1.2%. So anytime I enter a trade (long or short), it automatically calculates the correct number of units based on my limit entry price for that specific currency pair. To answer your second question, I limit my trade size to 0.6% for all the exotics and metals, to further limit my risk. I do this by simply cutting in half the suggested number of units to trade.

So technically, I could be risking more or less than the 0.6-1.2% that I mentioned. But I have found that right now, those starting points are working well for my personal risk tolerance.

To give you a better idea of how much risk I am currently taking on...
I track 32 total pairs, and currently trade in 27 of them. Of those 27 pairs, I have an active position right now in 16 pairs. My margin used is 18% of my margin available. Oanda's margin rules are slightly different (safer) than other Forex brokers. At Oanda, you can only lose up to half of your account before margin calls close your positions. So if you considered entire account balance, I really only have 9% of margin used. To give you an idea, when I first started with trades at 3-5% of overall capital, there were times when my margin used was 60-70% of margin available. I felt way too vulnerable at that level, which is why I cut back to where I am at right now.

Ignored

Quoting mathematician

Disliked

Yes, I use OANDA too. I like it. I think they have the right idea with their platform by encouraging (forcing) safer practices. I probably would have gone bust many months ago if I were with some other dealer. (By the way, I also use 50:1 leveraging.)

As a user, I understand OANDA's fractional lot system. Actually, it's one of my favourite features of OANDA. Their "X%" option in the trading preferences doesn't make too much sense to me. It's a percentage of your margin available, and for non-USD based pairs, the units are different depending on the current price of the base currency. There is no real connection to what percent of your account you will lose on a losing trade. For example, I set the preferences to 1%. On some pairs, I will lose 1% if a trade closes at about -200 pips. On other pairs I will lose 1% if the trade loses over 300 pips.

As I mentioned before, I have been using a fixed number of units based on the idea that a losing trade of say 200 pips will cost me 1% of my account. The formula I use to estimate the number of units is:

Units = (Account Balance)*(Risk %)*10000/(Risked pips)

So if I want to risk 1% and I assume that a losing trade will lose about 200 pips, then:
Units = (Account Balance)*(0.01)*10000/(200) = (Account Balance)*0.5

Of course since there is no stop, there's no guarantee that a losing trade will be exactly 200 pips, so I could lose more or less than 1%. Also, since I use the same number of units for each pair, I am likely to be risking a bit more on pair that move faster such as GBP/JPY. At the same time, those pairs will likely make more profit on winning trades, so it doesn't bother me.

Once I have gathered more data and become more experienced with your 80/20 system as it was originally posted, I will start to do some tests. I'd like to run some rudimentary Monte Carlo tests to determine an optimal position sizing scheme.

Ignored

Quoting Flyjetz

Disliked

So far, all the fish have been about the same size (roughly 150-pips).

Ignored

Quoting Flyjetz

Disliked

mathematician,

Also, I wanted to say, that just looking at your name, I can tell that you and I think along the same wave lengths. Your first post on this thread was right on the money (no pun intended) as well. Let me share my thought process in developing this system.

The knowns:
- Every week, I know a candle will be drawn.
- The candle is formed by price movement either up or down or both.
- Carving a slice out of that candle will result in a profit

The unknowns:
- Where next week's candle will be drawn
- How big the candle will be

The "Two Great Leaps of Faith":
- I can estimate the "where" by looking at the prior long-range trend of candles.
- I can estimate the "size" by looking at prior week's candle sizes.

Rules for Profit:
- Your estimate is only an estimate based off the best data you have available.
- Trying to capture 100% of the candle will only work if your estimates are 100% accurate, or the candle range is much larger than expected. (In other words, rarely).
- Capturing a "slice" (60% for me) will enable more trades and allow room for error at both ends of the candle estimate.
- RISK MANAGEMENT, RISK MANAGEMENT, RISK MANAGEMENT!!!

Ignored

Quoting dim13

Disliked

I have some suggestions. I did alittle backtesting and these may be more profitable.

- for Buy/Sell bands (in place of the 80%/20% bands):

High Band (Sell): Average of prev 3 Highs + (ATR3 * .5)
Low Band (Buy): Average of prev 3 Lows - (ATR3 * .5)

These bands get hit less often but are more accurate. You might want to switch to daily charts since these band will get hit less often.

(or)

High Band (Sell): Average of prev 3 Highs
Low Band(Buy): Average of prev 3 Lows

For signals, I found it more accurate to use just use the Slope Change. If the Slope is down today(or this week) - Sell, if the Slope is up today (or this week) - Buy. It might be better to just follow the daily/weekly trend and enter on pull backs, instead of using a 20 period average - this will create a lag.

Ignored

Quoting mathematician

Disliked

Yes, my education is in mathematics, you too?

...

Speaking of the percentage of winning trades: It is worth noting that your percentage of winning trades has not varied much over the history of your trades. We've seen it over 70% and as low as about 55%, but that really isn't much variation considering there have only been...what? 55 trades now? There will naturally be higher variation in the begginning, but this will quiet down as the number of trades grows. It seems to be settling on 2/3, which is very nice.

Ignored

Quoting skyline

Disliked

I like this system Flyjetz, maybe I'll look to code an EA for it since it's completely mechanical (no user guess , discrection or whatever)

Ignored

Quoting Flyjetz

Disliked

I don't currently use EAs, (maybe someday). But I'm sure it would be helpful for other followers of this system. Thanks!

Ignored