Very simple USDJPY scalp

I have been experimenting (successfully so far) with a very simple scalping method on USDJPY, which I would like to present for discussion here.

The entry points are very easy to identify, but I have yet to find a good exit and stop-loss rule. Thus, I could also not yet introduce money management to this method, which would be necessary to call this a complete system.

Here is what I do.

I take a broad look at the market to form an opinion on the general direction that USDJPY is trading in. This includes daily and hourly charts, checking support and resistance points, and major stock indices, in particular S&P500 and the Nikkei 225 because this pair moves a lot in line with general risk appetite or aversion.



I then open a one-minute chart of USDJPY. Indicators are

1. LWMA(5)
2. LWMA(45)
3. Stochastics(5,3,3)

Entry is on a cross of the two LWMAs. Stochastics I use as a filter (was overbought above 80 before I go short or oversold below 20 before I go long). The Bollinger Bands use standard settings, they are not used as an indicator but help me visually to gauge the market situation.

In the attached chart I went short at the vertical red line. Price was wobbling a bit after that as it had actually already dipped into oversold territory on the oscillator, but then started going down.

Exit is the difficult part. I usually put a fairly tight trailing stop of 10 pips to lock in profits, but that of course gets me booted out of the trade quickly. I also have the "trendchaser" EA on the chart, but it usually does not come into play. Frequently I exit by discretion after spike candles, if the pair starts ranging, or when I feel the trade is going against me.

I use this on USDJPY because I find it fairly predictable on very short timeframes and because the pair moves during the Asian and European sessions, which are the ones I trade.



What I would love to get some input on is,

1. Has somebody experimented with a similar setup on other pairs?
2. Which sessions work for those pairs?
3. What is a good exit rule?
4. Where should an initial stop loss be placed?
5. Because this is on a short timeframe, would it make sense to just stop-and-reverse regardless of overall direction?
6. Etc.

Looking forward to everybody's opinions!

Jesper

Add to this, I generally prefer going short (on any pair) because of market psychology.

Just went short USDJPY again, but not in profit yet. EURUSD with low spreads across brokers should be promising, too...

I know that MA crosses are a lagging indicator, but do believe that on a time frame as short as this that shouldn't matter too much.

Which MT4 broker has the lowest spread on USDJPY???

Hello,

I am using a similar strategy.

EMA close 5, 13, 34

and

Macd 11 12 1

This Macd settings are perfect for your system and confirm every cross of your MA's.

Let me know what you think.

Holger

Hello,

Me again. Just forgot. I am scalping 5 min on EUR/USD, USD/CHF, AUD/USD.

The 5 and 45 is not bad, but I changed the 45 to 40 on my chart. Now Macd and moving average cross looks very good. If the 5 is above the 40 you are in an up trend and if it is below you are in a down trend. I use the 13 and 34 for this. This helps me to get an idea of the trend without watching different time frames and instead I am watching 2 more currency pairs.

Let me know if this helps.

Holger

Hello Holger,

Thanks for the input. I applied the MACD with your settings to my chart and you are correct, it tracks the price movements very closely. The MACD bars cross over zero pretty much where the MAs cross.

I also underlied a second MACD at (144, 233, 1), not for signals but to help me look at the overall direction (I'm reluctant to talk about "trend" on the M1 timeframe).

I guess setting the longer MA to 40 or 45 makes hardly any difference. The point is that the two averages are fairly far apart, thus giving less signals (and noise) than for example 5 over 12.

In general I don't believe in MA crossover or oscillator signals. They are by nature lagging indicators and make you chase the market. However on this short a timeframe I don't see this as a problem.

The attached MACD is a custom indicator that changes color with direction. To the best of my knowledge it is in the public domain for free use (I got it off a forum that Eaty started).

Cheers,

Jesper

Holger,

I forgot to ask in my previous mail... How do you deal with exits? On reverse signals or at predetermined levels, or is it fully discretionary? Do you set initial stops?

Also, one more thought on using the two MACDs. If the slower one is consistently staying in one color, indicating strong directional push, the faster one can generate (re-)entry signals. Say the slow MACD has been red for a while, but the fast MACD turned blue for at least two bars. This indicates a retracement while the direction is still intact. When the fast MACD turns red again the retracement is over and price starts pushing in the main direction again.

Incidentally, this is quite close to the approach taken by the LMT "low maintenance trading" system that was spam-promoted by everybody and his little sister a couple of months ago. I got a Clickbank refund for that because it really didn't do much for me that freely available indicators couldn't.

Enjoy the weekend!

Jesper

Hello Jesper,

Instead of using a second macd you can use another LWMA 100 to get an idea about the direction of the trend. 5 and 45 above 100 = only long, 5 and 45 below 100 = only short. Because you are on a 1 minute chart (more noise) it is more difficult with exits. You have to look for small profits only and have to make more trades per day. But the 100 will help to filter out some less profitable trades. If you try to let a trade run, you have always the risk that it reverses. Take always small profits between 5 and 10 pips.
I like more the 5 min (less noise), because after crossing of the macd I get usually at least 10 pips. If the market is choppy or in a range I go for 10 pips, otherwise I go for 20 pips. If the market is trending well, I let it run. There are less crossings on the 5 min, but it makes it less stressfull to wait for it and to watch 3 or 4 pairs. If I have 2 - 3 crosses of the macd on 3 pairs I am happy with the profit.

Hope this gives you some ideas.

Holger

Hi Holger,

Sure, any trend indicator with a longer setting will help to point out overall direction. I just personally prefer using different types of indicators for signals (here the MA) and filters (here the MACD).

I usually don't find it wise to use oscillators for entry signals, especially not in the "standard" way of selling when they're overbought and vice versa -- they can stay there very long, and there is a lot of discretion in deciding what level is overbougth or -sold. So I think a better use for oscillators is as direction filter, MACD over/below zero, RSI or Stochastics over/below 50, etc.

The MACD with two timeframes works for me in trend-following systems. For counter-trend entries I would overlay two RSIs in one indicator window (refer to the standard "Momentum" template in MT4) and wait for them to separate significantly before the quicker one converges back to the slower one.

As for profits, I usually try to lock in 5 pips after the trade went for 10 pips in my direction, then incrementally trail manually. On USDJPY I can often get around 20 pips per trade.

It's already Friday night where I live, time to hit the pub.

Cheers!

Jesper

Of course, MACD is nothing else than difference of two MAs. So when you have same or similar MAs on the chart (say MA 7, MA 21) and MACD(7,21,1) then they cross at the same moment as MACD crosses zero line, i.e. when difference of the two MAs is zero...

From specs:
MACD = EMA(CLOSE, 12)-EMA(CLOSE, 26)
SIGNAL = SMA(MACD, 9)

You might want to try adding a third indicator such as a WMA 8 or 10. Use that only for possible exits, when it crosses the 5.

Here you have a MACD (very visual) withch can help to see + support and resistance. Hope this can help.