Wiser Trader Stocks and Options Newsletter

 

 

 

 

1.0 Precision Money Management

 

      This report describes a natural model that relates trading system performance, stop loss settings and profit goals.  Simple algebraic equations are used to make specific trading adjustments and set trailing stops to meet varying profit goals.  A few examples are provided to show the role played by stop loss settings in determining profits.  While this relationship has been sought for several years, it has not appeared in print before now.  I hope it is not too tedious.  I wanted to give you first glance at it before publishing an article.

 

      Most of us think of a trailing stop loss when the term money management is mentioned.  The book, “How to Make Money in Stocks” used a value from 7% to 8%.  Many stock advisories including Stansberry and Associates, Outstanding Investments and the Oxford Club typically use a 15% to 25% trailing stop loss.  Option advisories use still higher values in the 35% range, as is done by Michael Lombardi, and as high as 50%, as used by Dr. Stephen Cooper. It is seldom explained how trailing stop losses are arrived at, leaving the impression that they can be arbitrarily chosen.  However, this is not the case.  Too narrow a stop loss setting can eat into profits by exiting volatile trades too early.  Too wide a stop loss setting can eat into profits by consuming too much capital.  A systematic way of choosing an optimum stop loss setting is needed to achieve a precise level of money management.

 

      Intuitively, the higher one’s success rate in correctly choosing the direction of trade and the higher one’s average gain per trade, the looser one can afford to set his stop loss.  However, when one sets a specific earnings goal, this relationship needs to be more precise.  Fortunately, the ability to gather trading system performance statistics allows the use of an engineering approach.  Consistent use of a trading system allows one to define a very precise relationship between the stop loss value one should use, the average return for a series of trades and the percentage of correct choices one has made in the direction of a trade. 

 

      The methodology introduced here for precision money management is based on average values of historical trading system performance and is only applicable when a trading system is consistently followed.  The model should probably not be applied to unstructured trading across a variety of instruments requiring varying trading techniques.  Each trading system or technique generates a unique set of statistics for which this methodology can be applied on an individual basis.

 

 

 

      We need to define several fractional averages based on historical trading data.  First, for N number of trades, F_P is the average fractional profit equal to the sum of the fractional gains and losses for all trades divided by the total number of trades. 

 

      F_P = (sum of fractional gains + sum of fractional loses) / N

 

      For example if there were 3 trades resulting in +25%, -15% and +30% gains, the average fractional profit would equal 0.133 = (0.25 - 0.15 + 0.30)/3.  This assumes that each trade uses an average amount of capital which we will call C. 

 

      Since the “sum of fractional gains” is equal to the number of gains N_G times the average fractional gain F_G, and the “sum of fractional loses” is equal to the number of loses N_L times the average fractional loss F_L, the above definition can be expressed as,

 

      F_P = (N_G F_G + N_L F_L)/ N

 

      It is understood that N_G + N_L = N.  The value of N_G divided by N is the fraction of trades chosen in the correct direction F_C.  N_L divided by N is the fraction of trades chosen in the wrong direction (1 – F_C).  So N divided into N_G and N_L leaves the following form.

 

      F_P = F_C F_G + (1 – F_C) F_L                                                                                              (1)

 

Where,

      F_P is the average fractional profit for N trades that each uses an average amount of capital C

      F_C is the fraction of trades chosen in the correct direction

      F_G is the average fractional gain for N_G winning trades

      F_L is the average fractional loss for N_L losing trades

 

      Each fractional quantity can be expressed as a percentage but decimal fractional forms should be used in the equation.  What we are now have is a relationship between the stop loss value one must use F_L, the average return for a series of trades F_G and the percentage of times that the correct direction of a trade must be chosen F_C in order to achieve a given average fractional profit F_P.  This is exactly the relationship we are after. 

 

       The average fractional profit F_P, the fraction of correct choices in trade direction F_C and the average fractional gain F_G should come from historical information based on a number of trades that is representative of one’s trading system.  In order to clearly see how to use equation (1), we need to establish a profit goal over a definite period of time.

 

      The dollar profit per trade needed to meet a specific dollar goal in a given amount of time depends on the number of promising trades likely to be identified by one’s trading system within that time period.  The number of promising trades that become available within a given time period must be chosen judiciously because the last thing we want to do is force a trade under less than ideal conditions.  In other words, we want to remain true to whatever system we are using.

 

      For N trades each valued at an average capital amount C, the average fractional profit can also be defined by the total dollar profit goal D_G divided by the dollar sum of all N trades D_S,

 

      F_P = D_G / D_S

 

      The dollar sum of all N trades D_S can be represented by C times N, giving the following form,

 

      F_P = D_G / (C N)                                                                                                                 (2)

 

      D_G, D_S and C are expressed in dollars, whereas N is a whole number and F_P is a fractional value, as before. 

 

      This completes the model that consists of equations (1) and (2).

 

 

 

Example 1:

 

      Let us suppose that we have done a sufficient number of trades using our system to determine that the average fractional profit is 10%, the average gain per trade has been 29% and the fraction of times we chose the correct trading direction was 70%.  Further let us choose to earn $3,000 over a 1 month period.  By our estimate, we figure that we can safely enter an average of 3 trades a week and remain within trading system guidelines.  This equates to 3 trades per week times 4.33 weeks per month or an average of 13 trades per month. 

 

Variables:         F_P = 0.1

                        N = 13

                        D_G = $3,000

                        F_C = 0.7

                        F_G = 0.29

 

      Using equation (2), and solving for C,

 

      C = D_G / (F_P N)

 

      The average dollar amount of each trade needs to be, C = $2307.69. 

 

      Rearranging equation (1), the average stop loss setting F_L must be no wider than,

 

      F_L = (F_P - F_C F_G) / (1 - F_C) = - 0.3433 or -34.33%

 

 

Example 2:

 

      Using essentially the same conditions, we can look at what the effect of certain improvements in trading would have on the profits.  Say we habitually exit winning trades too early and could possibly increase the average fractional gain F_G from 29% to 36%.  The resulting stop loss setting F_L from the rearrangement of equation (1) could then be widened to -50.66%. 

 

Example 3:

 

      Let’s suppose for a series of potentially high yielding trades we know that an extra wide stop loss setting of -60% is needed and we want to know what the effect will be.

 

      First we might want to look at the effect of a wider stop loss setting on profits with everything else remaining constant.  We do this by equating the right sides of equations (1) and (2) and solving for D_G,

 

      D_G = (C N) [F_C F_G + (1 – F_C) F_L]                                                                                      (3)

           = ($2307.69) (13) [(0.7) (0.29) + (1 – 0.7) (-0.6)] = $689.99

 

      Clearly, our original profit goal of $3,000 can not be met without some additional changes, such as an increase in the number of trades from 13 to 57 over the month period.  But this is not feasible since it was already established that the maximum number of trades identified by the trading system was estimated at only 13 over the given time period.

 

 

Example 4:

 

      Next, since the trades in example 3 are believed to be potentially high yielding trades, we look at the increase in F_G needed to justify the wider stop loss setting of -60% and still meet the original profit goal.  By rearranging equation (1),

 

      F_G = [F_P - (1 – F_C) F_L] / F_C

           = [0.1 – (1 – 0.7) (-0.6)] / 0.7 = 0.4 or 40%

 

      So the average fractional gain for winning trades F_G would need to increase from 29% to 40% to justify a widening of the stop loss from -34.33% to -60%.

 

 

      The foregoing examples give insight into how trading system characteristics give rise to various stop loss settings.  Narrow stop loss settings imply a smaller fraction of trades chosen in the correct direction or a smaller fractional gain for winning trades.  Wider settings imply the opposite.  Stop loss settings should not be arbitrarily set independently of trading goals and trading system performance.  Stop loss levels more or less define future profits for a given set of trading rules, whether the user realizes it or not.  While it is laudable that traders are encouraged by their advisors to adopt money management, advisory recommendations of fixed stop loss values without knowing a client’s profit goals can be misleading.   This model of trading system performance, stop loss settings and profit goals enables more precise money management.

 

      If you have a need for a formula related to investing, savings rates, compound interest and other complex profit related issues, send me an email and I will work on it for you at no cost.

 

 

1.1 Option Alerts 4 and 5

 

      This section reviews option trades for the past week.  Progress on the down-market filter experiment has been moved to the results section.

 

      Two alerts were issued last week, one for KBH July 115 Calls (alert no. 4) and one for EBAY July 75.00 Puts (alert no. 5).  Early Monday morning the market was up, meeting conditions to buy the KBH, JUL 115.00 Call, KBH GC.  Later the same morning, the markets were down meeting conditions to buy the EBAY, JULY 75.00 PUT, XBA SO.  On Tuesday, EBAY was up more than 3% so an email was sent out to inform everyone that the trade would be aborted if EBAY rose above resistance at 78.00.  It did.  In fact EBAY is hovering around 81.00.  The loss was -17.6%.  I did not wait for the stipulated -35% loss before exiting because confidence is low in shorting stocks selected by the original stock candidate filter.  More will be said about the experiment below.

 

      Home construction stocks were down 4% in Friday morning’s pre-market with the news of Dell Computer’s soft outlook going forward.   Judging by the sluggish response of home builders to a rise in the major averages later in the morning, investors were focusing on something not apparent in the general news.  This is in contrast with the  previous day where the percentage change in KBH was in lock step with the NASDAQ to within 0.05%.  One item of note is a widely circulated newsletter that hit my mail box the same morning.  The newsletter in question is highly respected, generally conservative and consistently negative about investment prospects for real estate.  The author, Martin Weiss,  is a time honored bear who remembers when a fall in real estate values lead the way for major market declines some time ago.  The fact that consumer credit, including mortgages, was 50% below consensus earlier in the week reinforced the negative newsletter tone.  If, indeed, the newsletter was the cause of this sudden home construction index decline, it will take about a week for the index to bottom, if major averages hold up.  If it’s more than that, I will exit the trade if KBH falls below its support in the $110.00 range.

 

      This week is a plunge back to earth from the stellar beginning for option alerts the previous week.  Average profit stands at 3.5%. No new alerts will be sent until after Monday, if any, due to numerous open positions.

 

1.2 Other Trades

 

      In other trading, positions were taken in Monsanto (MON) Cameco Corp. (CCJ) and two other home builders Toll Brothers (TOL) and Beazer Homes (BZH). 

 

      CCJ, a uranium explorer and processor with interests in gold mining, was traded based on the TTC-A template.  The company reported a 27% increase in revenues and a 34% increase in profits on January 27^th.   The stock was momentarily up 9% on Thursday with the September $35.00 call netting a 54% profit. 

 

      MON, TOL and BZH call option positions remain open.

 

 

2.0 Market Analysis

 

      The industry leaders list in Table 1 contains the top 10 industries for four periods consisting of 1 week, 1 month, 2 months and 3 months.  Results are ranked highest to lowest based on the average percentage gain per week within any period times the number of appearances of an industry in any of the four top 10 lists. 

 

            Table 1

 Market Summary

 Week ending 02/12/05:

 Indices for the Week:

 Dow Jones     +0.7%

 NASDAQ        -0.5%

 S&P500 Index  +0.2%

 Russell 2000  -0.4%

 Industry Leaders:

 Coal

 Exploration & Production

 Oil Equipment & Services

 Heavy Construction

 Full Line Insurance

 Home Construction

 Tobacco

 Pipelines

 Oil & Gas

 Toys

 Marine Transportation

 Oil Equipment, Services & Dis.

 Health Care Providers

 Steel

 RE Holding and Develop

 Oil & Gas Producers

 Integrated Oil & Gas

 Semiconductors

 Non-life Insurance

 Insurance Brokers

 Gold Mining

 Insurance

 Platinum & Precious Metals

 Gold for past 30 days:

 USD    -1.09%

 CAD    +0.88%

 CHF    +1.11%

 GBP    -0.68%

 EUR    +0.82%

 JPY    +1.24%

      The continuing bounce in the broad market is in part a classic relief rally such as often occurs as earnings season winds down and fears of poor reports fade.  Next week will be far more active with reports on retail sales, industrial production, housing starts, and PPI all on the calendar.  The 5 day RSI for the DOW is overbought.  For the S&P500 and NASDAQ it is neutral.   The 5 week RSI is neutral for all 3 major indices.  Other sentiment indicators are given in Table 2.

 

 

Table 2

Sentiment Indicators 02/12/05

 

 

3.0 Procedure

 

      The following stock screens were generated with tools from AAII.  The short term trading filter used for Table 3a looks for optionable stocks whose percentage relative strength over the past 6 months is greater than 90%, EPS Growth over the past 12 months is greater than 80% and are within 5% of their 52 week  high  with  a  minimum  price  of  $50.   When a trigger has been received, call or put options are supplied depending on the direction of trend as reflected by the respective positive or negative slope of the 30 day SMA. 

 

 

Table 3a

Short Term Options (Original Stock Candidate Filter) as of   02/12/05

 

 

 

      House keeping resulted in the removal of ARLP, BG, BZH, EBAY, KBH, MTH, NUE, PHS and TOL because either their price action, earnings growth or relative strength fell too far out of specifications for the stock candidate filter.  This was done even though open trades exist for KBH, BZH and TOL.

 

 

Key

 

 

 

      The stock filter resulting in Table 3b is for the potential to write covered call options.  The only difference between filters for Tables 3a and 3b is that in 3b the maximum stock price is $20.  EPAY and MDR were removed as their prices fell out outside of filter specifications.

 

 

 

Table 3b

Short Term Covered Call Writing Screen as of   02/12/05

 

 

 

 

 

 

 

 

 

      For the screen in Table 3c, the number of selections is reduced by eliminating stocks having P/E’s greater than 30.  Stocks removed for hose cleaning were CEDC, FBP, MSA, MOH, ODFL, PHM, SSD, TCBI, TS and WIBC.  These stocks fell too far outside specifications for the filter.

 

Table 3c

Growth Momentum (Intermediate Term) Screen as of   02/12/05

 

 

 

      For the Peter Lynch screen in Table 3d, again the number of selections for this screen is reduced by eliminating stocks having P/E’s greater than 30.   CSLMF missed earnings and TMIC fell more than 10% in the past month.  Both were eliminated from the list.

 

 

Table 3d

Peter Lynch Value (Intermediate Term) Screen as of   02/12/05