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]


174

The selection of a tjpe of trading spproach implies the tjpe of risk you are willing to take on each trade. The difference between trend and countertrend sjstems was discussed in the section Sjstem Trade-Offs in Chapter 22 ("Practical Considerations"). In general, a trend sjstem has many smaller losses and fewer laiger profits; a countertrend sjstem has the opposite, many smaller profits and fewer laiger losses This is similar to the choices availaUe with gambling sfrategies.

Even among trending sjstems there are enormous differences in trading rid;. A tjpical moving average sjstem, which qualifies under conservation of capital, keeps individual frade losses small compared with individual profits, but may have a series of losses that add up to one very large continuous loss. In addition, trend sjstems fail when they encounter a sidewajs period, and frequently reverse from long to short and back again until prices finally pick a direction. Tests show that you cannot reduce the number of trend frades below a threshold level, even if you me the trend period very long, it alwajs gets caught in a sidewajs price pattem that generates a series of small losses.

A breakout sjstem, the other very popular frending technique, exchanges a much laiger rid; for much greater reliability than the moving average or exponential smoothing. When a new long position is entered based on a new high price for a specific time period, the rid; is defined as the price needed to make a new low for the same period. Therefore, the frade risk at the time of entrj is the difference between the high and low for the period,

trade-rid; = (aliighest(close, period) - (alowest(close, period)

If the highest high and lowest low are used, the risk can be much greater.

The advantage of the movmg average or smoothmg method is that individual losses are small. Usmg the breakout spproach, the advantage is that the method does not demand as much of price movement; that is, prices may be very erratic, and take as long as needed to move to a new level, as long as they do not make a new low. It seems illogical that, if prices had given a buy signal by moving to new high levels because something fundamental was affecting the maiket, prices could tum around and make a new low. That would indicate that the reason behind this move has disappeared.

Rid;-Adjusted Rehims

The decision whether to choose one trading approach with higher retums and higher risk, or another with smaller retums and lower rid; can be resolved by converting performance to an annualized, rid;-adjusted return. Using any method for measuring rid;-a standard deviation of monthly equity changes, average maximum refracement, or even laigest drawdown-divide the annualized rate of rdum by the risk to get the risk-adjusted rate of return. The highest value is the better choice.

Time to Recovery

A frading program that recovers quickly fran losses is more desiraUe than a program that lingers at its lowest level of loss. The fact that the first program recovers quickly does not change its level of risk, only its practical sppeal.

How Much Bigger Could the Rid; Really Be?

Many of the rid; measurements described in ihis section attempt to express the probability of a fiiture loss Within reason, many of them succeed. An investor must distinguish between the calculated probability and the effects of a real loss. If there was only a Po chance of a 50°o loss in a single month, then by the 50th month there would have been a 50So chance of a 50°o loss. There is also a large error factor in most of these calculations because the amount of data used to find the statistics was too small. If 100 cases were used

there is still a 10°o chance of error in the results. The value of the measurement is also dependent upon the period in the maiket that was used to create these values; the risk in crude oil was exceptionally high during the Gulf War. If that period is not included, the risk of trading crude oil might seem very low

LEA-ERAGE

The consequences of leverage are readily seen in a rid;/reward analjsis, but it also plajs a crucial role in the trading strategj itself Futures maikets offer exceptionally high leverage opportunity, and most traders and analjds acl



as though they are obligated to take advantage of the maximum allowable. Without leverage, futures prices show no more risk than stocks.

Consider the case of an investor with the substantial sum of $50,000 allocated to futures trading. If the price of soybeans is $6.00 per bushel and silver is $7.00 per ounce, a 5,000bushel contract of soybeans is worth $30,000 and a 5,000-ounce contract of silver is valued at $35,000. With no leverage, the investor could only purdiase one contract with the certainty of being to hold that contract as long as necessary to achieve a profit. Let us assume that, because of fundamental reasons, the investor believes that soybean prices wig move to $8 per bushel and silver to $11 per ounce within the next 6 months. The nonannualized gross retum on investment will be for soybeans and 57°o for silver, equivalent to a 20°o and 40°o retum on $50,000, respectively. With no leverage and not enough money to invest in both, the choice must be silver, which has the highest retum; in neither case could their be any rid; of ruin, in both cases there would he surplus cspital

But I00°o capitalization is hardly necessary. Even the most conservative investor would agree that the price of either commodity would not drop to 30°o of the current value; therefore, there is no significant risk in cspitalizing 70° o. rather than 100%, of the contract values. By reducing the individual cspitalizations, the investor can then purdiase both silver and soybeans, thus adding diversification and a much higher retum on the available investment. Table 23-1 shows that 30° leverage would increase the retum on investment From 40° to 60° o, in addition to the way in which the investment could be allocated, given different levels of cspitalization.

As the capitalization is lowered and the leverage increased, both risk and retum get rapidly larger. In Table 23-1, there are certain levels which are more significant than others:

TABLE 23-1 Varjing the Leverage of an Inveshnent

Sofbeat)!

Silver

TotolKetum

Leverage

Needed

/

Needed

/

/

Needed

30,000

35.000

20,000

20,000

isfloo

27,000

31.500

20 00

31.500

31.500

24.000

28,000

20,000

28,000

28,000

21.000

10.000

24300

20,000

45.500

ie.000

10.000

21 00

20,000

30,000

39 00

15.000

10.000

17.500

40,000

S0.OOO

50.000

12.000

10,000

14.000

40.000

S0.OOO

40,000

9 00

20,000

10300

60,000

80.000

49.500

6.000

30,000

7 00

80,000

110,000

46.000

00

70,000

3.500

160.000

230.000

49.000

1300

150

I.7S0

300,000

450.000

487S0

30% leverage, the point at which there is added diversification, an increase in retum, and a negligible increase in ri.

90°o leverage, equivalent to exchange minimum margin levels. It allows wellcapitalized traders to have margins averaging about 10° of the contract value.

95 leverage, equivalent to the margin level given to hedgers and spreaders, based on owning the physical produd as offsetting rid;.

At 90°o leverage, the investor is required to use $49,000 for margin. The positions may be maintained without an additional invedment as long as the account does not show a loss in excess of $12,2507 (with a total of 30 contracts, i.e., equivalent to a loss of $408 per contract, or 80 per ounce of silver and 8c per bushel of soybeans). Because the daily trading range of both commodities exceeds that value, it is very possible to have a margin call on the first day the position is edablished. It is not possible for an investor to hold this position for long.

Reserves

Most professional money managers use reserves to reduce the leverage and rid;, thereby increasing their staying power. The size of the reserves usually averages about 60°o of the capital; however, ranges of I0°o to 80°o have been known. The use of a 50°oreserve effedively halves the retums and the rid; (Figure 23-5).

Looking again at Table 23-1, there is now only $25,000 of capital availaUe for margin. If exchange minimum



maigin rates are used, there is 90% leverage available; this means each contract of soybeans requires a $3,000 deposit and each contract of silver $3,500. The 80°o leverage line, with one-half the maigin shown, would require a total maigin commitment of $23,000, slightly under our limits. The net effect of 90°o leverage on one-half the total investment is the same as 45>o leverage. Retums are-

6 futures tra.bng. initial mi in is the mininiuni am-u Uainteiance marein is the level f c.italization tiiat m level) i tiie initial ma.gin level reijiire the invest to re

lerosit irfien the position is entered idintiieaccunt Losses that re.luce e,jiit7 to beL.w 75 itiie i citalization to the full initial level

FIGURE23-5 Vaijing reserves, assuming a 50°o rehim on $50,000 (with standard reserves= t). I

DR-ERSIFICATION

3 contracts of soybeans $ I o.ooo profit $30,000

4 contracts of silver x $20,000 profit = $60,000 Total profit $90,000 Net return on investment 180%

Diversification means spreading ride it is the well-established way of lowering sjstematic ride and every investor is well advised to diversify. For the pmposes of risk reduction, it is necessaij to distinguish between sjdematic ride which can be reduced by diversification, and maiket risk, which cannot he eliminated. The benefits of diversification are greatest when the maikets traded are very different and the methods of making decisions are also unrelated. Tjpical investment portfolios contain a variety of fixed income, equities, real estate, art, as well as different invedment philosophies, all intended to provide different rates of retum with different patterns and not suffer losses all at the same time. Unfortunately, this is not alwajs the case. Maiket ride including catasfrophic risk, is not predictable, and can surprise even the best investors. Avoiding maiket risk is the subject of this and other sections in this chter.

Practical diversification can be implemented using a broad selection of maikets within a single tactical sjdem, a selection of different trading drategies for one product, or a combination of multiple sjdems and products. The objectives of diversification are:

1 . Lower daily risk due to the offsetting of losses in some markets and sjstems "o-.-ith profits in others.

2. Insure participation in major moves by continuously trading those markets in various groups that are likely to reflect fundamental changes. This avoids the need to identifj which market will perform the best.

3. Offset unespected losses caused by a sjdem failing to perform in the current maiket, or a single trade thai generates a large loss. It may be only had luck that one Sjdem gave a short signal in coffee the day before the freeze that moved prices limit-up for 21 dajs, but another sjstem might have given an offsetting buy signal

4. Reduce cynosure in any one maiket or related group by, havmg funds didributed over many groups. If there



[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]