Hikkake swing trades on the FTSE100

[tomorton]

I have been tracking Hikkake patterns on the FTSE for some time but have only occasionally traded them. It seems a reasonably reliable signal, I have also backtested it across charts covering several years for the S&P, Dow, Dax etc.

The strategy is set out in the attached article by Dan Chesler. The real reward comes when you can make a good exit, way beyond the average return.

This is an EOD signal and I will start to upload these asap after they occur as from here on. This looks a simple signal and strategy and it would be nice to develop some guidelines that will maximise its return - please feel free to post your own outcomes and refinements of this strategy here.

 

[Hoggums]

Hi - looks interesting - I'll do a proper backtest for you.... results sometime next week.

 

[tomorton]

Hikkake FTSE100 2010 Results

I have taken stop to be the extreme high of the pattern for Shorts, the low for Longs: e.g. the 04/02 Short trade is triggered by breach that day of the Low of the Inside Day 29/01, 5148.6: I have used the highest High of the upward move of 01, 02 and 03/02, 5305.4, as the stop: as the profit target, I have used the standard 2 x risk (for no better reason other than that it is standard) i.e. 2 x (5305.4-5148.6) = 313.6 (target therefore 4991.8). (please note that where the breach occurs later than the 4th session after the Inside Day, I don't take the trade)

04/02 Short 5148.6 Stopped out 18/02 at 5305.4 Loss -156.8
15/03 Short 5594.7 Stopped out 17/03 at 5646.7 Loss -52.0
12/04 Long 5782.3 Stopped out 22/04 at 5684.5 Loss -97.8
27/04 Short 5666.0 Exit 05/05 at 5396.6 Win +269.4
14/05 Short 5257.1 Exit 25/05 at 4899.3 Win +357.8
07/06 Short 5072.5 Stopped out 17/06 at 5262.5 Loss -190.0
22/06 Short 5239.4 Exit 25/06 at 5059.6 Win +179.8
06/08 Short 5353.5 Stopped out 09/08 at 5416.8 Loss -63.3
20/08 Short 5228.6 Stopped out 01/09 at 5351.4 Loss -122.8
22/09 Short 5536.5 Stopped out 30/09 at 5635.7 Loss -99.2
06/10 Long 5660.0 Exit 04/11 at 5851.0 Win +191.0
26/10 Short 5724.3 Stopped out at 5794.3 Loss -70.0
10/11 Short 5841.3 Exit 12/11 at 5719.7 Win +121.6

Net gain of closed trades = +267.7
Performance = 5/13 (38%)
Av win = +223.9
Av Loss = -106.5
Av r: r = 1:2.1

The r:r isn't bad but the points score is not amazing. As ever, there may be ways to improve the profitable exits and I would be happy for comments here on this especially.

 

[Hoggums]

I programmed up my backtester with rules for Hikkake to see how it would do over a long range of time. I am using FTSE yahoo EOD data.

As with many so called strategies there is no exit criteria - so thus is it incomplete as a trading strategy. So I have programmed a buy and hold strat - the system will hold the trade until it is either stopped out or and new signal is generated. Whether the signal is the same or different as before it will close the trade at this point and open a new trade.

For realism I have programmed the minimum stop size to be 0.25% of the index (currently about 13-14 pnts), the system will not allow a stop smaller than this. Risk per trade is 2% of trading account.

Full results attached (Use wordpad to view - it contains unix linebreaks) - summary below:

RESULTS, 1 runs
Run 0, from 15/05/1984 00:00:00 to 06/05/2010 00:00:00,
FTSE, Spread:1, Margin 60, Interest 0.0

RATIOS:
Run 0 (FTSE:I0.0 ). Total: 652 Winners: 139 (21%) Losers: 513 (79%)
Average: Winners: 139 (21%) Losers: 513 (79%)

PIPS:
Run 0 (FTSE:I0.0 ). Total Pips: 4558 Avg Win: 100 Avg Loss: -18 Profit Factor: 5.48
Average: Total Pips: 4558 Avg Win: 100 Avg Loss: -18 Profit Factor: 5.48

TRADES:
Run 0 (FTSE:I0.0 ). Max Trade Win 778 pips. Max Trade Loss -119 pips.
Max Trade Win Unrealised 916 pips. Max Trade Loss Unrealised -185 pips.
Average Trade Length 4D 16H 11M. Longest Trade 72D 0H 0M. Shortest Trade 1D 0H 0M
Max conseq winners 3. Max conseq losers 34

DRAWDOWNS:
Run 0 (FTSE:I0.0 ). Max Single Trade DD: -2% Max Unrealised Single Trade DD: 2%
Max Total DD: 60% Max Unrealised DD: 60%

RISK and MARGIN:
Run 0 (FTSE:I0.0 ). Min Margin Req: 1% Max Margin Req: 2%
Max Risk from Single Trade: 2% Max Total Risk: 2%

BALANCE:

Starting balance for all accounts: £10000.00
Run 0 (FTSE:I0.0 ). Min Balance £9265.64 (93%). Max Balance £2671885.85 (26719%) Final Balance £1191848.11
Min Unrealised Balance £9160.73 (92%). Max Unrealised balance £3374464.45 (33745%)

ANNUAL BREAKDOWN:
Run 0 (FTSE:I0.0 ).
1984 22% increase, range -5% to 35%
1985 -7% increase, range -24% to 4%
1986 30% increase, range -8% to 49%
1987 252% increase, range 0% to 295%
1988 12% increase, range -13% to 15%
1989 37% increase, range 0% to 77%
1990 39% increase, range -7% to 39%
1991 5% increase, range -13% to 17%
1992 -5% increase, range -19% to 1%
1993 -25% increase, range -25% to 0%
1994 139% increase, range 0% to 155%
1995 11% increase, range 0% to 22%
1996 39% increase, range -15% to 44%
1997 80% increase, range -5% to 87%
1998 59% increase, range -18% to 64%
1999 42% increase, range -2% to 69%
2000 -10% increase, range -21% to 5%
2001 12% increase, range -14% to 26%
2002 -8% increase, range -22% to 0%
2003 83% increase, range -2% to 99%
2004 7% increase, range -26% to 10%
2005 -17% increase, range -25% to 14%
2006 -4% increase, range -11% to 14%
2007 19% increase, range -8% to 47%
2008 -4% increase, range -20% to 0%
2009 -43% increase, range -43% to 0%
2010 1% increase, range -10% to 1%

 

[tomorton]

Good work Hoggums, thanks.

The results look good too, though I'm not familiar with backtesting software. However, it looks to me like -
worst case outcome over such a long term was 7% loss
best case scenario over period was 33,000% gain
average r:r 1:5
worst drawdown 185pts, best gain +900pts

This all sounds pretty good. The test does suggest the pattern is worth some work to optimise exits. Have I read your results correctly?

 

[wackypete2]

Very interesting. I've seen a similar article before but never really looked into trading the pattern. Good job, Hoggums, with the testing.

I'm seeing this though, which raises flags:
Max Total DD: 60% AND
Max conseq losers 34

There are also 2 consecutive losing years a few times with an average 24 trades per year throughout. I agree that optimizing exits may help, but what about using 4h bars to give more trades?

Just my thoughts.

Peter

 

[tomorton]

The short signalled at 5841.3 on 10/11 has now retraced almost to entry point. The trade would still be live, as the stop would be 5902.1, but the best available price so far was the low of 30/11, 5519.2 for +322.1.

 

[tomorton]

Inside day today, with just a 51pt range. This sets up the possibility of a hikkake Friday, Monday or Tuesday. Can't say direction yet, tomorrow's session will be the clue: if tomorrow is an outside day, I would ignore that bar and look again Monday. If no signal by the close on Tuiesday, I would ignore the signal.

 

[tomorton]

Day 2 of the pattern is encouraging. A positive close on yesterday but certainly not a bullish pattern - hopefully not enough to push much higher but enough to suck in some more longs and bring in some close stops.

Now looking for confirmation for a short - if tomorrow's session has a lower high than today's and breaks Wednesday's low. Stop at today's high.

 

[tomorton]

Day 3 continues the bearish set-up but has not yet breached Wednesday's low. Possible short entry Monday or Tuesday.

 

[tomorton]

Today's rising session puts 100pts between today's high and last Wednesday's low. It seems unlikely that tomorrow's price will drop through the low - even if it does, the risk of 100pts may be too much. If no short entry signal tomorrow, scrap this pattern and wait for the next.

 

[tomorton]

No confirmation, no trade.

 

[tomorton]

Go to http://www.trade2win.com/boards/uk-indices/112778-ftse100-swing-trades-2011-a.html#post1368872 for further hikkake analysis.

happy new year all!