I tested by curiosity a strategy found in the last Kaufman Perry’s book, here is the rules :
It’s a intraday momentum divergence strategy, on equities, buying when the price is falling but a momentum indicator is rising from a low point. Opposite for sales.
The momentum is the 10-day RSI
The standard action is to sell and set a profit-target based on volatility. We can exit with a loss on a new price high or a momentum low; we can exit with a profit using a price profit target or a momentum high. That gives us natural exit rules without using a stop-loss.for buy signals, the opposite needs to occur. The price lows need to decline at a slower rate.
When the momentum of the highs is declining and the momentum of the lows are rising, we have the sell condition, and when the momentum of the lows is rising thile the momentum of the highs are falling, we can buy.
Buy signal :
low(t) > low(t-1) > low(t-2). The lows of the price bar are rising.
The RSI of the low prices is less than 20. The price is oversold
The difference beteween the RSI of the high and RSI of the lows must be greater then 15. This assures that the high and lows are moving differently.
We exit the long trade when the RSI of the high >=95
Separating by 15 points two momentum indicators descrease the number of trades but increases the percentage of profitable trades. Results on 2016, 5min Timeframe on SmallCap ETF : 29 shorts, 14 buys, 78%Winrates
You now got all. I allow myself to share this strategy, because, like many book’s strategy she is not complete. We don’t have every sells conditions and something disturb me in the rules : the RSI period.
Btw, i created this, note it’s only a draft and i maybe bad coded rules above:
Kaufman Div
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
DEFPARAMCumulateOrders=False// Cumul des positions désactivé
n=1
//LOWS
c1a=low[0]>low[1]andlow[1]>low[2]
c1v=high[0]<high[1]andhigh[1]>high[2]
//RSI
RSI1l=RSI[3](low)
RSI1h=RSI[3](high)
RSI1a=RSI1l<20
RSI2a=RSI1h-RSI1l>15
RSI1v=RSI1h>80
RSI2v=RSI1l-RSI1h>15
c2a=RSI1aANDRSI2a
c2v=RSI1vANDRSI2v
CONDBUY=c1aANDC2a
CONDSELL=C1vandc2v
//ENTREE LONG
IFCONDBUYTHEN
BUYNSHARESATMARKET
ENDIF
//EXIT LONG
exit1=RSI1h>95
IFExit1then
sellatmarket
ENDIF
//ENTREE SHORT
IFCONDSELLTHEN
SELLSHORTNSHARESATMARKET
ENDIF
//EXIT SHORT
exits=RSI1l<=5
IFExitSTHEN
exitshortATMARKET
ENDIF
Results on many stocks are interesting but the stratgy need more works.
I see the attached backtest as “accident”. I took the code above, optimised c2a and c2v on 1min Dow. But it shows us that we can get something from this strategy. If u plan working on it, Kaufman is using 5min timeframes, not 1min as i did.
RSI high/low difference is supposed to be a filtrer for false signals.
Where i’m not sure on RSI period. There is a problem on the system description, are we supposed to use RSI of the 3 lasts low/high or RSI[10](dclose) ? MTF could maybe answer this question. Gonna try…
Where i’m confidence is when i see the result after 1 hour of work and 50 codes lines. If we dig we certainly can found gold.
To help us continually offer you the best experience on ProRealCode, we use cookies. By clicking on "Continue" you are agreeing to our use of them. You can also check our "privacy policy" page for more information.Continue
.
