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] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [ 172 ] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205]


172

Rid; Preference

The Bernoulli theory of utility is general, but each investor has their own unique objectives and attitudes toward risk. Some of the participants in futures martlets would like to keep risks low and retums steadj-; others would like to risk all of their cspital for a chance at the "big move." This trait is called the investors ridq)reference. The risk preference or utility of an investor for a specific venture (in this case a trade) can be found by adding the expected value of the investors utilities or preferences for the various outcomes of that event,

P ~ Wipx + Mzp2 + + Wp, where X = 1

where there are n possible outcomes. The weightmg factors may be the results of personal bias or may be the calculated probabilities of each outcome. For exanple, a gold frade has a likely profit of $4,000, with a risk of $1,500 Adjust the reward values by dividing by 1,000. If the probability of success is 60° o, the total utility of the trade is:

P(trade) = 0.60 x 4 + 0.40 x (-1.5) = 1.8

If the probability of success were increased, the utility P would increase linearly. But investors do not feel the same about different rewards. Given a scale of 0 to 100 (negative to positive reaction), an investor may rank the 60°o chance of a $4,000 profit and a 40°o chance of a $1,500 loss as a 65. If the reward is increased to $8,000 while the risk remains at $1,500, the investor might raise the preference of the trade to 80, although the utility would be 4.2, more than twice as large.

The various pattems of a curve drawn through the computed utilities represent the rid; preference of the individual. Figure 23-2 shows the curve formations progressing from exfreme risk aversion to exfreme risk seeking. As the rid; increases in (1), the frader is less likely to participate in the trade; in (3), there is equal chance of taking the trade at all rid; levels; and in (5), the trader is more likely to enter a trade that has higher risk.

FIGURE 23-2 Investor utility curves Concave downward Linear Concave upward

(3) Neutral (4) Modereie (5) Extreme to risk risk seeker risk seeker

Source: R.J.Teweles. C.V. Harlow, and H.L. Stone. The Commodity Futures Gome. Who Wins. Who Loscs.V/hy? (McGraw-HiH. NewYork. 197-4. p. 133).© 197-4. Reprinted by permission ofThe McGraw-Hill Companies.

Common Sense

Rid; contrcj is as much an issue of common sense as it is complex riUes and mathematics. Most of the chapter thai follows shows how various measurements, diversification, and leveraging techniques can reduce risk however, successful traders have spplied common sense, without complex formulas, for a long time. Some of these principles are:

1. Dont risk very much of your total capital on any one trade. The total risked should allow you to comfortably survive more losses than you would normally expect.

2. With whatever method you use, determine the maximum loss for the current trade in advance. Many floor fraders believe that their survival is the result of being the first to exit a frade. ("I didnt pay to get in, but where do 1 pay to get out?")

3. Exiting a trade is more difficult than entering because the timing is not alwajs yours. Exiting quickly is often



safer than being clever.

4. Stay with your trading philosophy. If you are a trend follower, then keep losses small and let profits run. You cant be a trend follower by taking many small losses and a few small profits.

5. Every sjstem has a trading profile. Be sure that profile is agreeable to you

6. Plan for contingencies. Not everything goes as planned, and you must be prepared for infrequent, but importani exceptions.

LIQUIDITY

The realities of trading are directly related to the liquidity of the market. Even though liquidity can have more than one meaning, lack of liquidity has only one trading result: poor execution. A liquid maiket does not necessarily mean good fills, but an illiquid maiket assures bad ones. Two tjpes of illiquidity are most often encountered in the maikets:

1. Fast-moving markets, in which there are mostly buyers or sellers, and few traders are willing to take the other side of the position.

2. Inactive maikets, usually in the deferred months in which there is less interest in either hedging or speculating.

The fast maiket is the result of a supply-demand imbalance, or perceived imbalance, where everyone is in agreement on the direction the market should move. The few hedgers and traders willing to take a position do little to offset the vast number of orders that flood the floor, pushing the price in the direction of the trend. In the inactive maiket, a premium must be paid to attract another trader to take the opposite trade. This will only succeed in

RISK CONTROL

an inactive market when the bid or offered price clearly appears to give the other trader a guaranteed short-term profit.

The illiquidity of a market produces the most important execution cost, often greater than a sizable commission and sometimes unreasonably large. Ginter and Richie2 have described this as a function of the order size, volume, and the speed of the maiket, in a single formula:

where = cost of execution due to liquidity Q = size of the order entered V = volatility of the maiket L, = volume of the maiket (all deliveries) L~ = volume of the specific option (delivery month) = constant factor

The volatility V might be the daily range (high minus low), or a factor of the standard deviation of the range. This would provide a measure of how much volume would move the maiket a specified number of points. The total volume (liquidity) L, is important because it implies liquidity due to interdelivery spreading Volume also serves as a measurement of the general interest in the product. More active trading in other contract months opens the possibility for trading in all months given the right circumstances.

The constant factor will vary according to the nature of the order being placed and the current direction of the price with regard to that order. If prices are moving higher and a buy order is placed, will be large; if prices are moving higher and a sell order is placed, will be smaller. The investor can see that trading oveifiead, including both commissions and other execution costs, will have considerably greater impact on sjstems that trade more often, have smaller average profits, and trade in the direction of the trend. Given a choice between sjstems of equal gross profits, the investor should favor the method of fewest trades in the most liquid markets.

CAPITAL

Success does not depend on having enough capital, but in using it property.



Every tjpe of sjstem has its own profit-and-loss cycle, depending on what price pattems or behavior it is measuring. A simple movmg average will be profitable in a trending maiket and will lose in a nontrending one., a sjstem that operates within a tradmg range will profit fran nontrending situations. It has been estimated that prices spend 70°o to 80°oof their total time in a nontrending movement; therefore, when using a trend-following sjstem, it is likely that trading will begin in a nontrending period. Because the rid; of the sjstem during a sustained trendless maiket may be unknown, it is best to start small. Find the minimum amount that can be used to follow the sjstem in a representative manner. During the start-up period, it is best to profit on 10°o or 20°o of capital than lose on 100°o Alwajs begin slowly-actual trading performance is rarely as good as expected.

Establish the maximum maigin based on the experiences of nontrending maikets- if uncertain, keep the total maigin well under 50% of availaUe cspital. The following sections review wajs of increasing maigin and compounding profits.

Diversification reduces both rid; and reward; done property, ihe rid; is reduced more than the reward. Using a well-dishibuted portfolio for a sjstem whose risks are known, a

G inter and J Eicbe, "DataErr.*.-andPnce DisLrtons," in Tecimical Analysis m Ojumodities, Peirv J Eaufman (ed) (j-lm Wiley & ?,s. New Y.dc 1 1.

laiger portion of the total cspital can be allocated to maigin. Advantages of statistics become spparent in the netting out of profits and losses daily; the majority of trades will move in a favorable direction a majority of the time. The volatility of a portfolio will be dampened by this netting effect; a trade with a slow upward trend and a distant stop-loss will be offset by a slow downward trending position in another commodity.

An undercapitalized account cannot be traded safely by selecting a sjstem with a favorable rid;-reward profile. Because there is very litUe dajing power, the trader must accept the fact that there is a small chance of success. There are two choices, based on the personality of the trader:

1 . Try for the maximum profit possible on a single trade with the capital available, if it loses, trading is halted, if it profits, there may be enough capital for a diversified spproach.

2. Take a low-rid; approach, which allows the most losses for the given cspital. This increases the time in the maiket, with comparably reduced profits. It is necessaij to have an unusually long run of small profits to reach a point of success.

There are two important rules in managing an account: (1) Dont meet maigin calls, which implies that a maigin call represents an objective identification of a bad trade, or a sjdem that is not meeting expectations. A maigin call is a time to review trading performance. (2) Liquidate your worst position when lightening up. The profitable trades have proved that they are trending or performing property; the losing ones have proved they are not. Stay with the good positions and liquidate the worst.

The management of capital is especially important at the beginning stages of trading, although it continually affects results. At some point, every trader wants to compound profits-a sjdem that is profitable can be more profitable. This philosophy most often leads to disasfrous results. The next sections present methods for measuring rid;, followed by a traditional and a conservative proach to compounding and increasing leverage.

MEASURING RISK

An important part of both sjdem development and money management is the recognition, measurement, and ultimately, the limiting of risk. Broadly speaking, money management is the art of limiting the risk of a portfolio while maximizing its return. To perform these tad;s, it is first necessary to select, or develop, profitable trading strategies to be used for each commodity, it is the results of these strategies, both simulated and real, that will be used to stmcture a rid;-controlled trading program (see the discussion of "Sjstem Trade Offs" in Chapter 22).

Selecting a Trading Model

A trading model is a set of riUes and formulas that, when applied to a price series, yields trading signals. These signals might be as simple as buy and sell Maiket on Close or complex contingent orders involving multiple delivery months and markets. To simulate results, the model must be well defined; that is, all the rules must be logically dated and programmable. Interpretive methods, such as many charting techniques, cannot be simulated; they depend on pattems that cannot be defined in advance. The use of actual trading results for later analjsis is alwajs preferaUe and is



[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] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [ 172 ] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205]