One day Dow breakout trades

[tomorton]

I'm going to log here the results of these potential trades on the Dow. I will actually take as many of them as I can using spreadbets but the log will show the theoretical trade gains and losses, not my actual results. Trades will be opened by pre-set order but closed manually: I won't always be able to post here in real time when I set an order or close a trade but will report all results.

Method is highly simplistic: before the New York open -
set buy order at yesterday's high with stop at the day's low
set sell order at yesterday's low with stop at the day's high
whichever trade is triggered, close it manually when New York closes next day.

Backtesting over the last 3 months suggests a win rate of 56% for a potential gain of 3,600pts (not accounting for slippage and spreads).

I can't yet truly believe this can work, so this forward testing to show up pitfalls.

Comments welcome.

 

[tomorton]

Further backtesting for April-July indicates a 65% win rate for +1,300pts - seems quite rational as a much more trending period than the more recent three months. But glad to do this test deomstrating reliability through varying market conditions.

 

[tomorton]

Set up for session 12/10 -
Buy order 17111, stop 17027:
Sell order 17027, stop 17111:
Risk = 84pts
If one order triggered, don't cancel the other.

 

[paszkman]

Hi Tom, interesting idea and I like concepts like this. I guess that as soon as one order opens, say the buy order, you cancel the pending sell order?

Would you run it as 1$ per point, or would you use position sizing so that each position, no matter how many points, is equal to every position?

 

[tomorton]

Hi Tom, interesting idea and I like concepts like this. I guess that as soon as one order opens, say the buy order, you cancel the pending sell order?

Would you run it as 1$ per point, or would you use position sizing so that each position, no matter how many points, is equal to every position?

Cheers paszkman - I would leave the untriggered order in place so that I automatically get a stop and reverse if my first order is stopped out. Outside days, when price breaches both the high and the low of the previous day, are not that common (looks like 18 in the last year on the Dow) and 12 of these closed outside the prior range also, raising the possibility that a stop and reverse would have generated a profitable second trade as the first was stopped.

Position sizing is an individual issue but I would run this with varying amounts of $ per point so that the total risk for the trade remains constant: I would base this figure on a %age of my account size. But I can see that there might be the possibility of increasing/reducing $ per point in certain patterns as probability varies, but I just don't have data on these yet.

 

[paszkman]

Cheers paszkman - I would leave the untriggered order in place so that I automatically get a stop and reverse if my first order is stopped out. Outside days, when price breaches both the high and the low of the previous day, are not that common (looks like 18 in the last year on the Dow) and 12 of these closed outside the prior range also, raising the possibility that a stop and reverse would have generated a profitable second trade as the first was stopped.

Position sizing is an individual issue but I would run this with varying amounts of $ per point so that the total risk for the trade remains constant: I would base this figure on a %age of my account size. But I can see that there might be the possibility of increasing/reducing $ per point in certain patterns as probability varies, but I just don't have data on these yet.

is your research based off say yahoo historical data (so real cash trading session high/low/open/close) or off spreadbetter daily high/low/open/close data?

 

[kalott]

http://www.marketwatch.com/story/making-gaps-in-the-sp-500-chart-work-for-you-2015-07-15

 

[tomorton]

is your research based off say yahoo historical data (so real cash trading session high/low/open/close) or off spreadbetter daily high/low/open/close data?

Historical data from Sharescope to give me high, low and close prices. These are restricted to the actual NY session times whereas the SB charts show their Dow market virtually 24 hours a day so e.g. the SB high could have been printed during the Asian session while NY is closed - so I'm aiming to get orders in as late as possible before the actual US open so that Asian volatility doesn't cause false entries.

 

[tomorton]

Today's result -
Buy order triggered at 17111, closed manually at 17131 = +20pts
Win rate = 1/1 = 100%
Total gain = +20

Tomorrow's orders -
buy at 17140
sell at 17064

 

[tomorton]

A frustrating experience today. The spreadbet quote on the Dow opened well below yesterday's Low so there was no chance to enter the trade short - you can't trade a breakout through a level that price has already passed.

At that point I had thought this would be a no-trade day as price was unlikely to climb all the way back through yesterday's High. However, it did do so and then fell all the way back to close at 17082.

So I'm going to count this as a failed Long -
Buy order triggered at 17140, closed manually at 17082 = -58pts
Win rate = 1/2 = 50%
Total gain = -38pts

Ho hum.
Tomorrow's order levels -
Buy at 17173
Sell at 17034

 

[stockanaya]

Hi Tomorton,

I am newbie to this forum as well as to Spread betting, I really like your idea...infact I tried something similar not exactly the same. I will test your system in my demo account for sure.

Regards

 

[tomorton]

Cheers and good luck stockanaya.

I'll be slightly amazed if something so simple can work but its worth a try. Please be aware that this method goes 180 degrees against much conventional trading wisdom - running your winners, buying only in an uptrend, risk reward ratio etc. Anyway lets have some fun and see where it takes us.

 

[tomorton]

A good result today. Price fell through yesterday's Low of 17034 and closed at 16924.
Gain = +110pts
Win rate = 2/3 (66%)
Cumulative gain = +72pts

Helpfully, the SB quote at the US Open was right in the middle of yesterday's range so there was no problem entering both buy and sell orders and I was able to make an actual profit.

Tomorrow's order levels -
Buy at 17112
Sell at 16887
Risk = 225pts (will be adjusting £ per pt accordingly obviously)

 

[paszkman]

Helpfully, the SB quote at the US Open was right in the middle of yesterday's range so there was no problem entering both buy and sell orders and I was able to make an actual profit.

So have you thought about how to deal with gaps? i.e. if open price is outside of previous day's range? Will you set limit orders? or ignore until price returns to range?

 

[tomorton]

Was thinking about gaps when I woke up this morning. Best I can plan is, if SB quote is above (below) the prior day's range, I would not set a Buy (Sell) but would go ahead and set the Sell (Buy) order. If I had screenwatching time, if and when price touched the mid-point of the range, I would then set the Buy (Sell) order as well.

The lowest risk option would be treat it as a no-trade day. But Outside Days on the Dow are not that common and two-thirds of them end in at least some profit at the close from the breakout trade opposing the gapped open.

 

[tomorton]

Modest win today, showing +30pts from 14/10's High to subsequent Close. The UK SB price at the US Open was obligingly right near the middle of the range but with such a wide daily range in points terms I suppose a major breakout was unlikely. Still, a win's a win.

Gain = +30pts
Win rate = 3/4 (75%)
Cumulative gain = +102pts

15/10 range was 17144.5 to 16933.5 (211pts). The SB quote is currently a little above that but anything could happen between now and tomorrow afternoon.

NB: University of Michigan consumer confidence comes out soon after the US Open. It is forecast to be a bit up on the previous. This release can sometimes move the market around unpredictably. Hold off your orders until later if you want to avoid this volatility risk.

 

[stockanaya]

Indeed a 30 point gain. I will do this again in my Demo account. I was trying to put these numbers (open/high/low/close) in an excel sheet to work out a trend but seems like this could be a hit or miss depending on volatility. I guess we need to have a clear understanding of news which affects the market you are playing in.

 

[tomorton]

Awkward situation on the Dow. I happen to be screenwatching as we arrive at the US open and the SB quote is already above yesterday's high. Therefore, I have entered only my Sell order at yesterday's Low. I will enter a Buy order at yesterday's High only if price drops back to halfway through the range - at around 17039.

 

[tomorton]

Update on Friday.

The EOD charts appear to show a comfortable gain of 71.5pts from Thursday's high to Friday's Close. In practice however, the SB quote "broke out" above Thursday's high immediately after Thursday's Close and never fell back far enough for me to enter the buy order. So I have to count Friday as a no-trade day.

If you had taken the premature signal for Friday's session, the trade would not have been stopped out but this is a matter of luck, not driven by the actual US market itself.

 

[tomorton]

I'm going to continue this next week but further reports may become sketchy - being around to enter Dow orders just before the US Open isn't always possible for me. However, I'm going to try this on the FTSE100 too, exact same rules.