freezing temperature of 0 degrees to a boiling temperature of 100 degrees. The latter goes from a freezing temperature of 32 degrees to a boiling temperature of 212 degrees. The formula for this temperature transformation is

F = 32 + Cx9/5

where is the temperature in degrees Centigrade, and F is the temperature in degrees Fahrenheit.

In the system Intermarket One, shown in Figure 14.9, we first transform T-bond prices to the equivalent S&P prices. This is done at each bar, so we are not

FIGURE 14.9 SYSTEM: INTERMARKET ONE.

Page 1

Type : System

Name : Intermarket One

Notes

Last Update : 01/20/97 05:33pm Printed on : 02/07/97 10:19am Verified YES

{Datal is S&P. Data2 is T-Bonds. This system mathematically transforms T-Bond prices to S&P prices. It generates a trading signal when the Transformed T-Bond line crosses the S&P line. It is a simple system with no inputs, and is shown for development purposes }

Var: Y(0), b(0), begSP(O), begTR(O), g(0), j(0), m(0), n(0);

If CurrentBar = 1 then begin

begSP = Close of Datal;

begTR = Close of Data2; End,

If CurrentBar > 1 then begin m = AbsValue(Close of Datal n = AbsValue(Close of Data2 If n<>0 then b = m /n, g = b * begTR; j = begSP - g;

Y = j + b * Close[l] of Data2, End;

If Y crosses above Close of Datal then Buy {the S&P} on Open;

If Y crosses below Close of Datal then Sell {the S&P}on Open;

Source: Printed using TradeStation PowerEditor by Omega Research Version 4.02.15-Jul < 09 1996.

- begSP); begTR);

{This is the transformed T-Bond line.}

relying on unknown information. Then, we utilize a crossover of the transformed T-bond line over the S&P line to signal a buy or a sell.

The results, at first glance, are poor. As shown in Figure 14.10, net profit is $155,075. Although 52 percent of the trades are profitable, the system developed a maximum intraday drawdown of $65,450.

The poor performance stems in part from the fact that Intermarket One is too simple a system. Stated another way, the formula for the indicator is a very crude interpretation of the data series. More complexity is required to better fit the data.

Moreover, the system generates a large number of trades, and transaction costs weigh heavily on the results. Improvement can be obtained by adding a filter.

improving the system with a Meaningful Filter, or setup

Most simple systems are improved by adding a filter to the system, to reduce the number of trades and improve accuracy. This filter can be thought of as a setup, a condition precedent to executing a trade. It would be analogous to the cocking of a gun to get it ready, before you can pull the trigger to bring about the action.

Momentum is traditionally used in technical analysis as an oscillator, or overbought and oversold indicator. If momentum gets too high, prices have risen too fast, and it is time to sell. If momentum gets too low, a reversal to the upside is indicated.

Momentum, which is the change in price over time, is calculated by subtracting the price N bars ago from the current price, and plotting the point at the current point in time. If momentum is positive, prices are increasing; if momentum is negative, prices

Figure 14.10 Performance summary: Intermarket one.

Intermarket one

S&P 500 Index - CME-Daily Performance Summary:

Total net profit Gross profit

Total # of trades Number winning trades

Largest winning trade Average winning trade Ratio avg win/avg loss

Max consec. winners Avg # bars in winners

Max intraday drawdown Profit factor Account size required

04/21/82 - 12/19/96 All Trades

$ 155075.00 $1376500.00

1776 919

16425.00 1497.82 1.05

-65450.00 1. 13 65450.00

Open position P/L Gross loss

Percent profitable Number losing trades

Largest losing trade Average losing trade Avg trede(win & loss)

Max consec. losers Avg bars in losers

Max # contracts held Return on account

$ 14225.00 $-1221425.00

52% 857

$ -30725.00 $ -1425.23 $ 87.32

13 2

237%

Source: Printed using TradeStation by Omega Research Version 4.02.15-Jul < 09 1996.

are decreasing. Used in this way, we can construct a fdter, such that we will only consider buying the S&P if the momentum of T-bonds is positive, and selling the S&P if the momentum of T-bonds is negative. System: Intermarket 2 is shown in Figure 14.11.

Intermarket Two adds momentum to Intermarket One. Lenl represents the number of bars we go back for the momentum calculation. Figure 14.12 shows system performance across numerous values of Lenl. For most values of Lenl, there is a significant improvement in performance results with the addition of the filter.

Note that for Lenl = 3, net profit is $306,350. However, for the values of Lenl immediately preceding and following, performance falls off drastically. Furthermore, there is no stability in the drawdown figures. The system is not robust because neighboring input values give poor results. A slight change in future market conditions would result in a large decrease in system performance.

Figure 14.1 i system: intermarket two.

Page 1

Type System

Name : Intermarket Two

Notes

Last Update : 02/07/97 10:38am Printed on : 02/07/97 10:40am Verified YES

{Datal is S&P. Data2 is T-Bonds This system adds a momentum filter to Intermarket One.)

Var: Y(0), b(0), begSP(0), begTR(O), g(0), j(0), m(0), n(0); Input: Lenl(3);

If CurrentBar = 1 then begin

begSP = Close of Datal,

begTR = Close of Data2; End;

If CurrentBar > 1 then begin

m = AbsValue(Close of Datal - begSP); n = AbsValue(Close of Data2 - begTR), IfnoO thenb = m/n; g = b * begTR, j = begSP - g;

Y = j + * Close[l] of Data2, {This is the transformed T-Bond line.) End,

If Y crosses above Close of Datal and Momentum(Close, Lenl) of Data2 > 0 then Buy on Open,

If Y crosses below Close of Datal and Momentum(Close, Lenl) of Data2 < 0 then Sell on Open,

Source: Printed using TradeStation PowerEditor by Omega Research Version 4.02.15-Jul < 09 1996.