Главная » Периодические издания » Last minus sign
2 3 4 ... 24
last minus sign
(X) On Chart 3 which represents the open. In suanary, you have a lover close relative to the previous day followed by another lover close, folloved by a lover open. In this case a long position vas taken on the open of the third day and exited on the close of the third duy. in all cases trade entry is on the open (shown as the last minus or plus sign in the pattern) and trade exit on the close* Vo stops vere used in the tests.
The dates of entry on all possible buys after two lower closings and a lover open (*5) and all sales after two higher closings and a higher open Iti2) are listed on chart c. on the chart itself each daily bar ehowe the open (entry) vith a da h on the left of the bar. The dash on the right of the bar is the close (exit).
I conclude that a bias does exist from open to close, as evidenced by the fact that many Of the patterns produce 60% or better results Regarding the syatoms, gross profits are high enough on some to warrant a system trade. Howeverj that is not the best application- Better systems do exist and I recommend further research usinq the most profitable patterns as the focus of the research. Note patterns 5, 9. 12, 13, 2ХЛ 24, 25, 29, 30, 33, 36, 44, 45, 52 and 53 all of vhich vill provide a good start in system development.
Some practical app1ications are to coordinate the patterns vith яupport/resistance/ angles and trend lines. For instance, an open on an important angle of support vhen a strong upward bias ie indicated increases the validity of the support Hatch action around the open for confirmation of the bias before enterinq a trade. Another application is to place stops to enter just above or belov the open in the direction of the bias (ORBP - refer to ORB article also in this issue for a complete explanation of this technique). This should be done early in the session to anticipate early (urgent) entry on the part of market participants. Early entry (Defined in Concepts Section of this letter) usually results in a big move for the day.
The patterns can also be used to enter when a trend has been defined- For example, pattern *-18i a buy after three up daysr a lower day and a lover open reflects a strong up market vith a one day counter move, The 1 over open provides an
JUNE 33 EURODOLLAR
V-i * j+ S.* *4iT* (*V
Cif. Eip: Jen i3 Ср:!зп Е>э; Jun ;3
9360 9340 9320 9300 9280
9260 £240 922C 9200 9580 9160 9140 9120 160
it- ;a ;a ? :i ;i * ч i с f я г* т i*
excellent chance for lov risfc entry. The objective is new highs for the move. Note the high percentago profit (63%). I have observed that most of the best profits in this pattern come in strong uptrends. In trading ranges it is not as profitable.
I do not use these systems eethodically in ay trading but find them to be a useful starting point in systea development. Increased and narroving daily ranges can be applied successfully to the basic close/open patterns to form some very reliable systems. In fact, dally range
differences relative to the previous day, or days, may be one of the most important concepts in system testing.
prтск pattern studies ii close to close patterns
tn the previous chapter, l presented all possible tvo thru five day open to close KQ/C) patterns- in this report i have tested all tvo tnru five day close to close (c/c) partems* tntry vas assumed on the last close of the pattern vith exit on the next days close. stops vere not used.
A listing of the patterns is given in table в. also a chart of june oonas is displayed highlighting pattern *24i (-++J.
t nave three objectives in cms study. tne first is to determine vhethor a bias exists from close to close. the second is to isolate profitable яуз tens ♦ the third is to use the patterns vitn a decisive bias to structure trads entry with other techniques for further system development.
table в shows, from loft to right, the pattern number* cno pattern itself, vnether a buy or sell vas taken, the number of trades of the pattern, percentage profit, and tne gross profit.
the june bond cnart shovs five patterns (a thru e) with а, с ana e profitable vnile b and d vere losers. tne left hand corner of the chart shovs an example of tne pattern; ana the
entry point. on table r, ? 24, under tile pattern column tne pattern snovs tvo lover closings
followed by tvo higher closings all relative to tne close immediately preceding it. entry vas tasen on the close of the last cay. in this case, it vas higher and a sale vas indicated, тле trade was exited on the next days close i the x on insert chart),
refer to the * profitable column for bias and verify vith the gross profit column- tne larger the numbers the stronger the bias. large gross profits vlll also give some indication of the systems validity, but slippage and oommision must be subtracted from tne profits to get an accurate reading. this is accomplished oy taking the number of trades and multiplying by commission and estimated slippage per trade. knen a strong bias is indicated further system development is pursued.
the results for these patterns are less inpressivo than thoee for tne open to close patterns in the previous chapter- because of the marked aitrerence, i conclude tnat tne bias breaks down the longer the position is held- Risk increased dramatically as indicated by the substantial drop ln gross profits. on the other
hand the C/C pattern* do provide a viable oeginning for further systea development. In fact those pattern* are the brainchild of the 0/c pattorns.
ifnen an 0/C entry is preceded by u high profitability C/C pattern there are some interesting results. For example, pattern *24 [-) vas profitable 5b!*S of tne time vnen selling- vith Jib,/30 gross profits* On its own tnis IS not that impressive, but, if you wait for the next days open and sell it if it is nigher and exit on the close tne results improve to profitability and $22,24 gross profits. In adaption- if tne day before entry, the last day of the c/C pattern, a vide spread day occurred and the sen was taken on tne higher open there is a dramat1с improvement. вЪ% profitability vith
$U,a57 eross profits on just a tnird 01 tne trades of the previous pattern -
An winning system is qui te an
accomplishment. A variation of the last pattern is a sale vith a / tic stop above tne open. Results vere as follows* э2% vinners vith S1U,5U0 Grose profits ana average vinners vere 4 times tfce si2e of average losers, This is the beginning of some excellent lov risk entry techniques- It may difficult to appreciate the value of a system vith only и--1, winning trades but vith as small a risk as / tics, bU4> is an extraordinary result.
Tn summary, tnese patterns can be used for indications of bias every day. it is prudent not to trade against tne bias unless iniormation clearly suggests otherwise* These basic patterns provide a sound base for further researcn. Any high percentage profitability system vili result m great revards vnen experimenting vith tne variables.
РЛ Г Г£йН I РЛТТЕЕИ 8/S f ?ЯА0£5 ЖП 5*0SS PROf ГТ
PRICE PATTERN STUDIES SOYBEANS OPEN TO CLOSE
In this report i nave tested all combinations of four closings with either a nigne: or lower open the day after the last close. The data utt11 zed was Soybean open, niqn, low ano close prices from 197U to 1988. mis resulted in tnirty-two pattern variations. The primary objective of this study is to determine if a bias exists from open to close given the previous price pattern. The secondary objective is to define profitable trading systems.
A Table (C) of the patterns is provided that includes- beginning with the left-nana column, the System Number, the Pattern, ICntry, Mumber of Trades, Percentage o£ Winning Trades. the Average Dollar Amount of all Winning TradQS, the Average Dollar Amount of all Losing Trades and the Total yross Profit. In the Pattern column, ( + J symbolizes an up closing, ) a aovn closing relative to tho previous day. The exception is the last sign on tne far rignt of tne pattern which represents the open of that day and the point of entry for tne trade*
Tnere are five days involved in each pattern, тле aoy of entry is Day I- Tne day imneoiately preceding the day of entry is Day Z. Each previous day is Day 3, 4, f etc.. (Refer to Diagram IJ. For example. Pattern 30 shows that on Day ь there is a lover close tnan tne previous day l-J (Day b not shovnj. Day 4 hae a lover close tnan day 2 Day i has a lover close
than Day 1 (-J* Day 2 has a hiqher close than Day J t+J. Day ls open is lover tnan Day Zs close (-) A sale is assumed and entry is taken on Day ls open with trade exit on Day le close. A positive Cross Profits figure is the on!у criterion used to determine trade direction (Buy/Sell). Consequently all patterns shov a profit - Tne preieraoie entry is shown in Table С in the column labelled н/S- In some cases this produces a percentage profit Dolow rtfty percent but because of the nigh average win, Gross Profits are positive, some cases also show a smaller average win than average loss but the percent profitability is very nigh producing a positive Gross Profit. These are borderline patterns that may or may not nave any preaictivc value. You should consider all of these factors (Percent Pror1t. Average Kin to Average Loss, and Gross Profits) when assessing the validity of a pattern.
Л natural limitation exists in Soyoeans that
aoes not exist in т.Bonds- That is, the potential mtraday price move is larger in T.Bonas than Soybeans 154.000 to ?J,000 respectivelyJ and in some instances tno limit vas $b,uU0 in T.Bonds. All things considered, the Reans perform as veil as if not better tnan Y-Donas in tnis type of basic test- Table D shovs 0 comparison between tne Soybeans and T.Bonds and includes tne trade direction and percent profitability. My purpose nore vas to determine If tneee patterns nad a similar oias in eacn market. Tf so, it vould seem to suggest some consistency in price behavior rrom commodity to commodity. Eight out of thirty-two nad different biases, suggesting some difierence in behavior betveen the tvo markets. Tnere were some marked dirferences in numbers jt 4, is and it with opposite trade direction on reasonably high percent profitability.
You will notice in Table С that the biggest winning trades wore sales (Patterns l, iu, , iO). There vere no buys over J10,0UU Gross Profit- The percentage winner seemed to favor tne upside but Gross Profits definitely favored the snort sales- T-Donas displayed a better balance in the cateqory of Gross Profits.
As mentioned above. Pattern jo vas a sale on Day l s open. The generic pattern produced 94 trades in vnich fifty-nine percent vere profitable. Winners averaged 4U J and losers averaged $iU- Gross Fronts vere f 13/VtJV- This pattern actually tested better than most of the patterns especially in tne Average f>ollar Win/Loss area. An attempt to improvo on this pattern produced some interesting results- Xl you aad an expanding daily range on Day relative to Day 3, it accounts for sixty-three oi the ninety-four trades of the generic patterns (Diagram 2). Profitability dropped to f\fty-one percent but profits were relatively stable at and a
$421 to 5170 Win/Loss ratio. Percentage profitability increases dramatically when you change Day l to a narrowing range rather tnan an expanding range (Diagram 3). This accounts for 30 trades vith Ь7% profitability. Оголв Profits vere $b,5Ul and it displayed a %44W to \$S2 Kin/Loss ratio. This is adequate but not nearly as rewarding as T,Bonds. Tn general X have so far iound it very difficult to come up vith nigh profitability, high Win/Loss ratio short-tern trades m the Soybeans. Tne more involved patterns that usually vork very veil in tne Bonds and potter in tne S+t* do not nave tne same consistency in the Soybeans. This suggests that intraday volatility is necessary for short-term
2 3 4 ... 24
© 2003 GARUN.RU.
Копирование материалов запрещено.