Buy at 4pm and sell at 4:30pm strategy

awg said:

keep up the good work

if you are keeping data, what would be interesting in addition, would be downside volatility. ie R/R for the 7.5 up points

Click to expand...

Thanks mate. I can tell you that the max gain is 57.5 and max loss is -38 in Nov. I wouldn't apply a R/R to the 7.5pt average, because it actually include the average of the down days as well. The reward one targets is definitely not 7.5 points, and a stop of 4 points (targeting R/R of 2/1) will have you stop out more often than now. I would probably trade this with a stop of some 1.5-2 std deviation of the range across the data set. Alternatively, you can simply take the max loss as say 1% of trading capital and size your position size accordingly.

Anyway, my cunning plan is for everyone to follow this strategy and have it become a self-fulfilling prophecy

awg said:

can u remember what the post title was for that dude who was trying to scalp 1 point with a 20 point stop?

Click to expand...

https://www.aussiestockforums.com/forums/showthread.php?t=12163

skc said:

Anyway, my cunning plan is for everyone to follow this strategy and have it become a self-fulfilling prophecy

Click to expand...

ha ha, I doubt ASF will make a diff.

A 200+ lot gets clipped in a second in this cowboy session.

Today was down. Everything works in phases, so if you are going to use it, prob should start soon.

Update on December numbers.

Total trading days = 19 (Excluding the half days on 24th and 31st)
Total days which SPI went up btw 4-4:30 = 10
Total days which SPI went down btw 4-4:30 = 9
Up days ratio = 53%

Total XJO change in month = +20.2 points
Total change btw 4-4:30pm = +88 points (Excluding the mad minute on 23rd where the SPI rose ~200 points)

Average change btw 4-4:30pm = 4.6 points per day
Average rise in up days = 23.8 points
Average fall in down days = -16.6 points

Max rise = 47.5 points
Max fall = -43 points
Max range = 60 points

I trade this move every day. I have also written my own backtesting software and have back tested the 1 minute data for the last year based on going long at 4.01 every day - no indictaors or anything like that I just go long every day at 4.01. I have found and my testing has confirmed that tight stops will flick you out but if you have your stop wide enough it is possible to have 48% winners with almost a 2:1 profit : loss ratio so it makes money but is nerve wracking. Going long at 4.15 increases to 52% winners still with around 2:1 win loss. Even though its slightly less reliable for some reason I prefer 4.01 over 4.15. I havent tested 4.10 but will give it a go and post results.

So if you can handle a loss every second day you can make money from this move.

Just to clarify I set a target at 2x stop but exit regardless at 4.29. So I dont get 2:1 every trade but my testing to date (still searching for better) says that 2:1 is the best initial target. I have tried to optimise my stop placement and win loss ratio and this is the best I have found to date.

At the end of the day its all about position sizing and money management.

As for why it happens I dont care - all I care about is can it be traded.

Thanks for the affirmation. Can I ask how you place your stop? I use around 1.5x standard deviation of the range between 4:00:01 and it's low, but I have never tested whether that is the optimal.

It is rare for the price not to go below the 4pm level before shooting up, so I have been playing around with increasing the position size even when it falls initially. I know it's against conventional trading to increase the size of a losing trade, but if the statistics is on your side then may be it's OK. Obviously the overall position size is still within my risk management strategy, and it also means that the returns on a net up day would not be as good.

I dont use any indicators on this trade, especially atr as volatility changes too much.

I just use a 13 point stop, 26 point target or exit at open of 4.29 bar because that's what my backtesting showed was optimal. Keeps it simple.

Today was a scary start but got the 26.

Been busy but will update data files and post some current test results later.

I updated to yesterdays data and re ran the trade. Looks like 15 : 30 is better now that there is more data in there. 4:10 entry looks ok. Problem is though that because you exit at 4.29 if target not reached a lot of the trades dont reach full profit. The profits are based on $25 per point. So you cant live off this, its just a bit of fun at the end of the day. Dates are 1-2-08 to 15-1-09

Time Profit Percent wins stop points target points
4:01
$5,975 41% 10 20
$9,025 45% 11 22
$9,175 45% 12 24
$10,450 47% 13 26
$11,275 47% 14 28
$12,225 48% 15 30
$10,625 48% 16 32
$9,175 48% 17 34

4:10
$3,425 43% 10 20
$7,625 46% 11 22
$8,700 48% 12 24
$8,800 49% 13 26
$10,325 51% 14 28
$15,050 54% 15 30
$14,000 54% 16 32
$12,350 54% 17 34

That doesn format too well in the preview but you shou be able to figure it out.