T h e idea is to t r y to m a x i m i z e t h e efficiency of e a c h trade as m u c h as possible. T h i s is a c o n c e p t p o p u l a r i z e d by R I N A s y s t e m s . W i t h the highly improved systems per-f o r m a n c e s u m m a r y in TradeStation 2 0 0 0 i , you c a n n o w get a per-feel per-for both the entry a n d exit efficiencies, as well as t h e overall efficiency of y o u r trades as R I N A sys-t e m s a n d O m e g a R e s e a r c h define sys-t h e m .
As an initial m e a s u r e a n d starting point, TradeStation's efficiency calculations w o r k v e r y well, b u t if y o u w o u l d like to e x p a n d y o u r research and try to o p t i m i z e t h e possibilities these t e c h n i q u e s offer, the TradeStation p e r f o r m a n c e s u m m a r y will n o l o n g e r b e e n o u g h . TradeStation again m a k e s the m i s t a k e o f calculating all m o v e s in dollar t e r m s , rather than in p e r c e n t a g e s . In the c a s e of efficiency analysis, this is OK w h e n l o o k i n g at the n u m b e r s , b u t not if you w o u l d like to do s o m e t h i n g about t h e m . Let us take a look at h o w TradeStation does it.
For Long Trades
Total efficiency = (Exit price — Entry price) / (Highest price - Lowest price) Entry efficiency = (Highest price - Entry price) / (Highest price - Lowest price)
Exit efficiency = (Exit price - Lowest price) / (Highest price - Lowest price)
For Short Trades
Total efficiency = (Entry price — Exit price) / (Highest price - Lowest price) Entry efficiency = (Entry price - Lowest price) / (Highest price - Lowest price)
119
Exit efficiency = (Highest price — Exit price) / (Highest price — Lowest price)
For All Trades
Average total efficiency = Sum of total efficiencies / Number of trades Average entry efficiency = Sum of entry efficiencies / Number of trades
Average exit efficiency = Sum of exit efficiencies / Number of trades
Figure 8.1 s h o w s an individual long trade. In this case we enter the market at 1,350, reach a low at 1,330, a n d then a high at 1,390, before we exit at 1,380.
Inserting these n u m b e r s in the above formula, we get an entry efficiency of 6 7 % , an exit efficiency of 8 3 % , a n d a total efficiency of 5 0 % . H a d this instead b e e n a short trade, the entry efficiency w o u l d have b e e n 3 3 % , the exit efficiency 17%, and t h e total efficiency —50%.
N o t e , however, that in the case of a short trade, we w o u l d have started out w i t h a profit, then m o v e d into a substantial loss, before we recaptured s o m e of that o p e n position loss. W h a t really h a p p e n e d then, w a s that b o t h the entry a n d the exit w e r e quite good. It w a s the m i d d l e part of the trade that failed. T h i s , however, does n o t s h o w up in the p e r f o r m a n c e s u m m a r y and, therefore illustrates the importance of dividing t h e trade a n d especially the d r a w d o w n into several subcategories, as
F I G U R E 8 . 1 An Individual long trade.
CHAPTER 8 Efficient Trades 121
to a n e w total equity h i g h at, let's say, $ 9 , 0 0 0 . B u t before y o u m a n a g e d to exit the position, it gave b a c k $ 1 , 0 0 0 of its o p e n equity.
A c c o r d i n g t o m o s t analysis p a c k a g e s , this m e a n s that even t h o u g h y o u man-a g e d to man-a d d $ 3 , 0 0 0 to your closed out equity, you n o w man-are $ 1 , 0 0 0 in the h o l e man-as c o m p a r e d to y o u r latest total e q u i t y high. F u r t h e r m o r e , if your next trade turns out to be a loser that i m m e d i a t e l y h a d y o u stopped out with a $4,500 loss before, your closed o u t e q u i t y n o w is $ 4 , 5 0 0 , a n d y o u r d r a w d o w n is $ 5 , 5 0 0 . T h e n you have a
$ 5 , 0 0 0 w i n n e r w h i c h , before it t o o k off, started out g o i n g the w r o n g way with an additional $ 1 , 0 0 0 , leaving y o u $ 6 , 5 0 0 in the h o l e . T h e n it gave b a c k $ 1 , 5 0 0 of the o p e n profit, a n d your total equity d r a w d o w n w o u l d n o w be a $ 1 , 5 0 0 . Figure 8.2 s h o w s w h a t this s e q u e n c e of trades c o u l d have looked like.
However, if you look closely at these n u m b e r s , they consist of three different types of d r a w d o w n s . At point 3, we are dealing w i t h the end trade drawdown ( E T D ) that tells us h o w m u c h of t h e o p e n profit we h a d to give b a c k before we w e r e allowed to exit a specific trade. At p o i n t 4, we are l o o k i n g at t h e closed trade d r a w d o w n ( C T D ) that m e a s u r e s the distance b e t w e e n the entry a n d exit points without t a k i n g into consideration w h a t is g o i n g on within the trade. At p o i n t 5, we are dealing w i t h the start trade d r a w d o w n ( S T D ) that m e a s u r e s h o w m u c h the trade went against us after the e n t r y a n d before it started to go our way. A n d at point 7, finally, we again h a v e to do w i t h t h e E T D .
F I G U R E 8 . 2
A look at different types of drawdown.
CHAPTER 8 Efficient Trades 123
To be sure, in recent years a few system developers and market analysts have addressed this issue in various ways, but as far as I k n o w n o b o d y has really nailed it d o w n w h e n it c o m e s to how one must go about examining the markets and the sys-t e m s in an appropriasys-te and sciensys-tific fashion. O n e of sys-these analyssys-ts is John Sweeney, technical analysis editor at Technical Analysis of Stocks and Commodities, w h o , in his two books Campaign Trading (Wiley Finance Editions, 1996) and Maximum Adverse Excursion (Wiley Trader's Advantage Series, 1997) came up with the
con-cepts of maximum adverse excursion ( M A E ) and maximum favorable excursion ( M F E ) . M o r e recently, David Stendahl at R I N A Systems has taken this a little S T D charted on the horizontal x-axis and the result of t h e trade on the vertical y-axis. F r o m this chart we can see that only one trade w i t h an S T D / M A E of m o r e
m a r k e t , from w h i c h , after this level h a s b e e n tested a n d y o u ' v e e n t e r e d t h e trade, the m a r k e t u s u a l l y m a k e s yet a n o t h e r c o r r e c t i o n before t h e final p e n e t r a t i o n and follow-through t a k e s us away ( w e h o p e ) in a major t r e n d i n g m o v e . Yet, other s y s t e m s m i g h t c o m p l e t e l y lack a n y specific levels a r o u n d w h i c h to p l a c e any natural o r d e r s . In our directional slope system, for instance ( w h i c h b a s e d its entries on t h e slope of t h e m o v i n g average rather than t h e crossover), g e t t i n g a feel for t h e S T D is crucial, b e c a u s e this s y s t e m , as it is c o n s t r u c t e d so far, oper-ates not only c o m p l e t e l y w i t h o u t a n y actual levels for w h e n a trade gets triggered, b u t a l s o c o m p l e t e l y w i t h o u t any exit rules as we d e f i n e t h e m . Figure 8.4 shows that t h e r e is no c o n n e c t i o n b e t w e e n the level of the p r i c e or the m o v i n g average, a n d the r e a s o n w h y t h e trade is triggered. ( N o t e that in this section we w o r k with a n 18-bar E n t r y M A a n d a 12-bar E x i t M A . )
To start l o o k i n g for the S T D / M A E a n d E T D / M F E levels that should work well on several different m a r k e t s a n d t i m e p e r i o d s , we m u s t do all our research with the R A D contract. To export t h e n e c e s s a r y data for the directional slope sys-t e m insys-to a spreadsheesys-t p r o g r a m , u s e sys-the following TradeSsys-tasys-tion c o d e :
Input: EntryMA(18), ExitMA(12);
Vars: EntryAvg(O), ExitAvg(O), LongEntry(O), ShortEntry(O), LongExit(O), ShortExit(O), LongEntryDate(O), ShortEntryDate(O), LongExitDate(O),
F I G U R E 8 . 3
The final profit In relation to the STD/MAE for the intraday Meander system.
CHAPTER 8 Efficient Trades 125
F I G U R E 8 . 4
The directional slope system applied to the Japanese yen.
ShortExitDate(O), LongEntryBar(O), ShortEntryBar(O), LongExitBar(O),
If Conditionl = True and Condition2 = True and MarketPosition = 0 Then Buy at Close;
If Conditionl = False and Condition2 = False and MarketPosition = 0 Then Sell at Close;
If Condition2 = False Then ExitLong at Close;
If Condition2 = True Then
CHAPTER 8 Efficient Trades 127
MAE = 0;
CHAPTER 8 Efficient Trades 1 2 9
T h e result is that TradeStation enters a n d exits e a c h t r a d e on t h e v e r y same intra-bar on a breakout. You will not be able to exit on the close, or intraintra-bar of that same bar, no matter h o w m u c h the market m o v e s against you, as suggested by the closing price of that s a m e bar. (That is w h y I have chosen to do all the research for the intra-day version of the M e a n d e r system within Excel, rather than in TradeStation.)
To m e , this is t h e equivalent of Excel's p u r p o r t i n g to calculate everything in
CHAPTER 8 Efficient Trades 131
F I G U R E 8 . 5
The final profit in relation to STD/MAE for the directional slope system.
Figure 8.5 s h o w s t h e S T D / M A E for t h e directional slope s y s t e m traded on t h e J a p a n e s e yen. T h i s chart s h o w s that o f all trades with a n S T D / M A E o f m o r e t h a n o r equal t o 4 % only two did n o t e n d u p a s losers, while m o s t o f t h e trades w i t h a n S T D / M A E o f 2 % o r less e n d e d u p a s w i n n e r s .
In t h e case of t h e M e a n d e r system, after y o u have m a d e sure that y o u h a v e o p t i m i z e d y o u r exit t e c h n i q u e s a n d taken every other step necessary, p e r h a p s all that m u s t be d o n e to t h e entry is to wait a c o u p l e of ticks before y o u enter, or at least not enter i m m e d i a t e l y at the o p e n w h e n t h e volatility can be high. In t h e case of t h e directional slope system, o n e solution c o u l d be to wait for a p u l l b a c k a n d then enter on a b r e a k t h r o u g h the p r i c e level that causes t h e E n t r y MA to c h a n g e its direction.
Just a s t h e S T D i s not the s a m e a s t h e M A E , the e n d trade d r a w d o w n ( E T D ) is n o t the s a m e as the m a x i m u m favorable excursion ( M F E ) , although we m i g h t be using essentially t h e s a m e t e c h n i q u e s to derive the necessary information. M F E is the m a x i m u m o p e n profit of the position. E T D is the difference between the M F E a n d t h e exit point, a n d the a m o u n t you are giving b a c k to the m a r k e t before the syst e m signals systhe syst i m e systo exisyst. T h e M F E and systhe final profisyst should systherefore be m a x -imized w i t h the a m o u n t invested a n d t h e n u m b e r s of contracts traded. T h e E T D , on t h e other hand, should be m a n a g e d with different types of exit techniques a n d stops,
such as a trailing stop, time-based stop, or profit target stop. T h e only way to com-pletely avoid any E T D is to exit all trades with a limit order that is triggered at the top of the equity curve. Figure 8.6 shows the relationship between the E T D and the final profit for the simplified M e a n d e r system, while Figure 8.7 s h o w s the rela-tionship b e t w e e n the M F E a n d the final profit for the directional slope system.
From Figure 8.6, we can see that an E T D of m o r e than 0 . 5 % very often resulted in a loss greater than 1%. This happens w h e n the trade goes immediately against us, and the S T D a n d E T D b e c o m e s the same. With a stop loss or trailing stop at an appropriate level (to address E T D ) , or a m o r e conservative entry technique (to address S T D ) m o s t of these losses could have been reduced considerably.
Figure 8.7 illustrates the final profit in relation to the M F E for the directional slope system traded on the Japanese yen. It shows that not until we get an M F E above approximately 5% does the system m a n a g e to signal an exit before we have given away all the open equity. Looking at Figures 8.6 and 8.7, it is not too difficult to see that it is easier to handle the information from Figure 8.6. Sure, we realize that something needs to be done somewhere around the 5% level for the directional slope system, but what? To get a better feel for this, we must look at the E T D in relation to the MFE instead of in relation to the final profit. This has been done in Figures 8.8 and 8.9.
B e c a u s e this m o d i f i e d version of the M e a n d e r s y s t e m o p e r a t e s with a d y n a m i c profit target that triggers an exit with a limit order, the h i g h e r the M F E , the lower the E T D . Figure 8.8 s h o w s that for all trades w i t h an M F E above 0.5%,
F I G U R E 8 . 6
The final profit in relation to ETD for the Meander system.
CHAPTER 8 Efficient Trades 133
F I G U R E 8 . 7
The final profit in relation to the MFE for the directional slope system.
F I G U R E 8 . 8
The ETD in relation to MFE for the Meander system.
however, there are still several that have a larger E T D t h a n M F E . With a profit p r o -tection or breakeven s t o p , m a n y of these losses can be avoided.
As is t h e case with so m a n y other l o n g - t e r m trend-following s y s t e m s , the directional slope s y s t e m s e e m s to give away a substantial part of the profit before it allows us to exit. For instance, Figure 8.9 s h o w s that b o t h trades w i t h an M F E i m m e d i a t e l y above 2 0 % did not allow us to exit until the m a r k e t had taken back about a third of our o p e n profit. N o t e that in this case the E T D is d e n o t e d with negative values.
T h e closed trade d r a w d o w n ( C T D ) is t h e d r a w d o w n you experience on your account from one trade to another for all y o u r losing trades. It also is important to distinguish t h e C T D from the T E D , as t h e differences c a n be quite substantial, especially for a l o n g e r - t e r m trend-following s y s t e m that is p r o n e to give b a c k a substantial part of the open profit for a trade before it allows you to exit. Because m a n y t i m e s the C T D also i s the s a m e a s the S T D , you m u s t c o m e t o t e r m s with t h e C T D by limiting t h e n u m b e r of trades with an overriding filter, w h i c h lets t h r o u g h only those setups a n d l o n g - t e r m m a r k e t situations that s h o w the highest likelihood of p r o d u c i n g a w i n n i n g trade, given the entry trigger you use.
Especially for l o n g - t e r m trend-following s y s t e m s , the C T D also can be a function of the E T D , w h i c h m e a n s it also should be m a t c h e d with an appropriate exit
tech-F I G U R E 8 . 9
The ETD in relation to MFE for the directional slope system.
CHAPTER 8 Efficient Trades 135
n i q u e , s u c h as a trailing stop. To get a feel for t h e C T D , y o u c a n chart the final profit, separated into w i n n i n g a n d losing t r a d e s , against b o t h the S T D / M A E , a n d against t h e M F E for all losing t r a d e s . F i g u r e 8.10 s h o w s that if a trade within t h e directional slope s y s t e m t u r n e d o u t to be a loser, only on a few o c c a s i o n s w a s the final p r o f i t / C T D significantly different from t h e S T D / M A E . T h i s suggests that w h e n a trade g o e s bad, m o s t of t h e t i m e it d o e s so right off t h e bat. Figure 8.11 s h o w s that for t h e w i n n i n g t r a d e s , t h e relationship b e t w e e n the final profit a n d S T D / M A E isn't as clear-cut as for t h e losers, w h i c h indicates that to c o m e to grips w i t h the C T D , it is i m p o r t a n t to analyze it s e p a r a t e d from w i n n i n g trades, a n d n o t clutter t h e analysis w i t h u n n e c e s s a r y information.
From Figures 8.10 a n d 8.11, we also can see that of all w i n n i n g trades, only six h a d an S T D of m o r e than 2 % , of which only three m a n a g e d to p r o d u c e a prof-it of m o r e than 1 0 % . At the other e n d of the s p e c t r u m there w e r e 16 losing trades w i t h an S T D of m o r e than 2 % , w h i c h in m o s t instances also p r o d u c e d a losing trade of m o r e than 2 % . In fact, if the trade turned out to be a loser, m o r e often than not, it also m a n a g e d to be closed out very close t o , or at, the lowest low of the trade.
T h e s a m e p h e n o m e n o n also can b e seen i n Figures 8.12 a n d 8.13, w h i c h s h o w t h e C T D respectively, t h e final profit for t h e M e a n d e r system, in relation to the S T D / M A E . As y o u c a n see from Figure 8.12, if t h e t r a d e t u r n e d out to be a loser, only on a few occasions w a s t h e final p r o f i t / C T D significantly different
F I G U R E 8 . 1 0
The final profit for all losing trades in relation to STD/MAE for the directional slope system.
F I G U R E 8 . 1 1
The final profit for all winning trades In relation to STD/MAE for the directional slope system.
from the S T D / M A E . T h i s suggests that w h e n a trade goes bad, m o s t of the time it does so right off the bat. F r o m Figure 8.13 y o u again c a n see that the similarity between the l o n g - t e r m directional slope s y s t e m a n d the s h o r t - t e r m M e a n d e r sys-tem is striking. T h e relationship b e t w e e n the w i n n i n g trades a n d their S T D / M A E a r e n ' t as clear-cut as for the losers, w h i c h c o n f i r m s that to c o m e to grips with the C T D it is i m p o r t a n t to analyze it separated from w i n n i n g t r a d e s , with no unneces-sary information.
F r o m Figures 8.12 and 8.13 we also can see that very few trades w i t h an STD of m o r e than 2% m a n a g e d to p r o d u c e a profit above 0 . 5 % , a n d only one came close to 2 % . On the other hand, t h e r e are p l e n t y of trades with an S T D of more than 1% that also p r o d u c e d a final loss greater than 1%. A n d j u s t as w a s the case for the directional slope system, m a n y of these trades s e e m to have been closed out at or very near their lowest o p e n equity. T h e question is, w h a t w a s the current mar-ket situation like to p r o d u c e these b a d trades, and could this have b e e n dealt with before entering the trade in the first p l a c e ?
W h e n we are c o m p a r i n g the final profit for t h e trades within t h e directional slope s y s t e m w i t h the M F E we can see that there are a few trades that s e e m not to be s t o p p e d out until there has b e e n an adverse m o v e in the n e i g h b o r h o o d of 5% or m o r e . Obviously, this is t o o m u c h to give back, w h i c h m a k e s it a g o o d idea to
CHAPTER 8 Efficient Trades 137
F I G U R E 8 . 1 3
The final profit for all winning trades in relation to STD/MAE for the Meander system.
F I G U R E 8 . 1 2
The CTD in relation to STD/MAE for the Meander system.
e x a m i n e the possibility of a d d i n g s o m e form of a trailing stop. Figure 8.14 also s h o w s that a total of 17 out of 24 losing trades for the directional slope s y s t e m only m a n a g e d to p r o d u c e an M F E of 2% or less. With a g o o d w o r k i n g filter m a n y of these trades could have b e e n w e e d e d out as well. A g a i n , this indicates that most losers go b a d from the p o i n t of inception or very soon thereafter. T h i s indicates that w h e n a trend gets hold of the m a r k e t , it is likely to be persistent a n d lead to a profitable trade. A l s o , the larger the M F E , t h e s m a l l e r the final loss.
T h e s a m e holds true for the M e a n d e r s y s t e m in Figure 8.15, w h i c h has quite a few trades that start out w i t h a profit but then allow for an adverse m o v e of more than 1.5% or m o r e before w e ' r e allowed to exit on the close of the bar. T h e r e also are p l e n t y of trades for w h i c h m o s t of the losers failed to p r o d u c e an M F E of more than 0.2%). In the former case, we c o u l d have avoided s o m e of the losses with a stop loss of s o m e sort. In t h e latter c a s e , m a n y trades could probably have been avoided c o m p l e t e l y with s o m e sort of a filter. A g a i n , t h e larger t h e M F E , the smaller the final loss. A l s o , w h e t h e r w e ' r e trading long t e r m or intraday, w h e n a t r e n d gets h o l d of the market it is likely to be persistent a n d lead to a profitable trade, no m a t t e r in w h a t t i m e perspective w e ' r e trading.
A n o t h e r w a y to s u m m a r i z e t h e e n t i r e life s p a n of a t r a d e is to chart the S T D / M A E , t h e M F E , a n d t h e E T D / f i n a l profit i n o n e s e q u e n c e , a s t h e y are a s s u m e d t o h a v e t a k e n p l a c e . F i g u r e 8.16 a t t e m p t s t o s h o w t h e s u m m a r i z e d life
F I G U R E 8. 1 4
The CTD in relation to MFE for the directional slope system.
ER 8 Efficient Trades 139
for a t r a d e in t h e d i r e c t i o n a l s l o p e s y s t e m . T h e a s s u m p t i o n we m a k e is that t h e further w e get from t h e p o i n t o f o r i g i n , t h e m o r e d i v e r g e n t the p a t h s the t r a d e c o u l d h a v e t a k e n , w h i c h l e a d s to a h i g h e r s t a n d a r d d e v i a t i o n of the o u t c o m e s . T h e r e f o r e , it is r e a s o n a b l e to a s s u m e that, in t h e c a s e of t h e d i r e c t i o n a l slope s y s -t e m , m o s -t M F E s h a p p e n e d af-ter -t h e S T D s . T h i s i s n o -t en-tirely c o r r e c -t b e c a u s e n o t all t r a d e s register t h e i r M F E s after t h e i r M A E s . B u t a s s u m i n g this i s the c a s e , we c a n see from F i g u r e 8.16 that the a v e r a g e trade for t h e d i r e c t i o n a l slope s y s t e m o n t h e J a p a n e s e y e n first starts out b y g o i n g a g a i n s t u s w i t h a b o u t 2 % (the m i d d l e line), b e f o r e it t a k e s off in t h e r i g h t d i r e c t i o n to register an M F E of a b o u t 7 . 5 % . F r o m t h e n o n , however, i t i s d o w n h i l l a g a i n , before w e are s t o p p e d o u t for a final profit of a p p r o x i m a t e l y 3 % , after we have given b a c k m o r e t h a n 5 0 % o f t h e o p e n p r o f i t .
T h e s a m e p h e n o m e n o n c a n be seen again in Figure 8.17, which shows the s u m m a r i z e d life of a trade in the M e a n d e r system. It is reasonable to a s s u m e that the further we get from the point of origin, the m o r e divergent paths the trade could have taken, which leads to a higher standard deviation of the o u t c o m e s . This s e e m s not to be the case for the M e a n d e r system, which m e a n s that we have to be a little m o r e careful interpreting our S T D a n d E T D findings. Figure 8.17 s h o w s that the average trade starts out with a w r o n g m o v e of approximately 0 . 6 % before things start to go our way, a n d we can register an M F E of about 0 . 6 % and a final profit of about 0 . 1 % . A g a i n , however, this a s s u m e s that the course of events is as outlined.
F I G U R E 8 . 1 5
The CTD in relation to MFE for the Meander system.
F I G U R E 8 . 1 6
The summarized life for a trade in the directional slope system.
F I G U R E 8 . 1 7
The summarized life for a trade in the Meander system.
CHAPTER 8 Efficient Trades 141
T h a t this is not always the case is noticeable in the standard deviation interval (the u p p e r a n d lower lines), w h i c h are likely to m o v e further apart from each other the further away from the entry point we get; this does not h a p p e n in this chart.
N o n e t h e l e s s , individually e x a m i n i n g each point, with its standard deviation interval, also can give us s o m e clues to w h e r e and h o w to place our stops. For instance, in the case of the directional slope s y s t e m we can expect that 6 8 % of all trades should p r o d u c e a n S T D / M A E s o m e w h e r e between - 3 . 7 t o 0 % , a n M F E i n the interval -2 to 1 7 % , a n d a final profit b e t w e e n - 5 . 5 a n d 1 1 . 5 % . Trades that do not m a t c h these criteria could then be assumed to be better or worse than accept-able. R e m e m b e r from Part 2 that a trade that is too g o o d to be true is n o t necessar-ily a g o o d thing. Different m e t h o d s for h a n d l i n g these trades, as well as better ways to depict the life of a trade, are dealt with in the next chapter, w h e r e we take a clos-er look at Sweeney's M A E a n d M F E . A l t h o u g h the old a d a g e " y o u r worst drawd o w n is still to c o m e " is likely to c o m e true sooner or later, it drawdoes not have to h a p -p e n first thing tomorrow, -provided you have d o n e your h o m e w o r k correctly.