back start next


[start] [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] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71]


26

chapter five

Youve Made Some Money - How to Keep It

• Use of Logical Cut-Loss Criteria

• Extension to Trailing "Sell" Signals

• How to Construct Selling Analogs

• How to Make and Use Non-Real Time Envelopes

• Selling Short

• Selling Rules to Remember

Lets assume youve made a buy using any of the techniques of Chapter Four. You have done so knowing that the odds are very good that your timing was correct, and that you stand to make a profit on your transaction. But some small odds still favor an incorrect decision. How do you protect yourself against these small chances of transaction failure? You could pick an arbitrary amount of loss you are willmg to risk in return for the long-odds profit potential you have set up for yourself. This is the usual route to go, and may be implemented either by a stop-loss order, or by a self-imposed level at which you will sell out by calling your broker. This approach is generally unsatisfactory, however. You always know in the background of your mind that the level you have chosen is arbitrary. With no particularly logical reason behind the choice, human psychology is such that it is a very strong temptation to "wnit just a little longer" in the hopes that your buy decision will be vindicated. Tiiis usually results in the process known as "letting your losses run"!

USE OF LOGICAL CUT-LOSS CRITERIA

There is a much better way, based solidly on the price-motion model. Once again, what is desired is a decision made in advance that if the price of the stock behaves in a certain way in the future, an automatic signal to cut losses short is given. This signal must not be arbitrary in nature; in fact, it should be so firmly based on cyclic precepts that minimum temptation will exist to ignore it!



Lets go back to our valid downtrend line concept. This is a straightened-out upper bound of a curvilinear channel enclosing the shortest duration cyclic component observable on the daily chart. But, by definition, a channel always has two bounds-upper and lower. Locate and sketch in the corresponding lower bound (it does not need to be accurately done). Now note the lowest low of the cyclic component enclosed by this channel just prior to the time when the valid downtrend line was broken. Draw a horizontal line on your chart at this price level, and either enter a stop-loss order at this value at the time of your purchase or vow vehemently to sell instantly on its violation!

Lets talk about the rationale behind this procedure. First of all you didnt mike the purchase at all until the time was cyclically right and the stock price told you that the cyclic lows anticipated had been passed. Under these circumstances, you expect the stock to move no way but up after your purchase, buoyed in this direction by the summed rates of price motion of several cyclic fluctuations. In no circumstance do you expect the price to get lower than the bottom of the channel enclosing the shortest duration fluctuation present. If it should do so, the channel would still be on the downside, indicating that the sum of all longer duration components is srill downside. This would mean that the expected multiplicity of lows has not yet occurred-and a low probability transaction timing error has caught up with you. Either this is true, or an unforeseen change has taken place. In either event-you want nothing but out until the situation resolves itself

Set up in this manner, you will find that the resulting loss is seldom more than twice in-out commission costs. Best of all, you now have a solidly rational reason for cutting your losses short, with an attendant increase in likelihood that you will do so! Your reasons for selling at this point are at all times as good as the reasons for the purchase in the first place. If you have developed confidence enough to buy on cychc criteria, you should also have enough confidence to cut losses short on the same criteria.

Lets try this procedure on for size in the case of the Gruen Industry buy signals. Turning to Figure V-1, we see the now familiar edge- and mid-band buy points for Gruen. The first buy level is at about / when the edge-band valid downtrend line was violated. The lower channel bound (of which this trend line was an upper bound) is shown dotted. The breakthrough told us that this tiny section of channel had curved sharply upward. This could only occur if the sum of all components of duration longer than the one enclosed by the channel has turned upward. In this case, the intra-channel component has a duration of about one-half week, and the components which have turned up sharply are the one-week, two-week and four-week ones. We mark the low from which prices pulled away on the breakout as TLL 1 (Trailing Loss Level No. 1).

If our analysis has been correct, and if cyclic action is not overridden by fundamentals, the stock price should not see this level again until our profit objectives are reached. If it does, something is wrong, and the stock should be sold without delay.

It is noticed that the stock did not violate this level, but continued to behave per expectation. But suppose that it had. What would have been the nature of our loss? We would have bought at and sold at 7 1/8. The percentage loss due to decreased stock price is 5% and we are charged another 4% in in-out commissions. Tliis is just about



A Triangle Resolution "Hold" Signal

1-1-!-1-1-

GRUCN INDUSTKCS

£ -1

4 . I

TU. 4

-TLL3

TLL 2

twice commission costs-a loss of 9% which we have risked against a precalculated potential for a 55% gain. Coupled with the fact that we have prestacked the deck heavily in favor of the large gain (as opposed to the small loss if we are wrong), the prospects for tiiis transaction are too good to pass up! It will be found in practice that this method of setting up stop-loss levels, predetermined at the time of the buy, will average about 10% as the stop-transaction lews factor. Best of all, you have a soUd reason for believing that you should take the loss-and a correspondingly better chance of doing so rather than let losses ran because of emotion and pride. At this point you should reason through for yourself the use of the price level marked TLL 3 in Figure V-1 as the initial stop-loss level for the mid-band buy signal.

EXTENSION TO TRAIUNG "SELL" SIGNALS

A fairly common trading practice is to use so-called trailing stop-loss orders (or price levels) to protect profits on a transaction, the theory being that as paper profits mount, you keep raising the level at which you will sell out if prices start to drop -thus always assuring retention of some portion of the maximum possible profits on a transaction. This scheme is good, but the same difficulties are encountered as in the setting of an arbitrary cut-loss level. If you trail prices too closely, a small dip may sell you out causing you to miss a much larger rise. If you trail too far behind prices, you salvage less and less of your paper profits if a dip turns out to be a major tum-around.

The concept for cutting losses short based upon the price-motion model points up an excellent way of getting around this difficulty. The initial cut-loss level, of



[start] [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] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71]