04/03/07 -36.7% Half of the loss was bad trade management half was hope... Postive reward would have been 11.1%, but Dow was too strong today. I am now below the level of 04/03/07 (lost more than 7 days). To get back to the top I need about 58% gain.
04/05/07 +10.5% The probability to loose when opening the postions was 17% (implied risk). In case of a loss 43.0% of the account would have been effected. Statistics on that trade ----------------------------- Implied chance 83% (at opening) Implied risk 17% (at opening) Account risk 43% Account reward 10.5% (Realized) Risk/Reward 4.1 Risking 40% of the account needs an edge of 70% (70:30). The chance to win has been 83% when I established the position. Therefore even an allocation of 66% (83 - 17) would have been correct when applying Kelly criterium. During the day the implied risk increased to 31%. In that situation an allocation of 38% (69 - 31) would have been correct. Therefore I was temporarly 5% over optimal Kelly and should have been selling that excess risk of 11.6% (5/43) on the market immediately by closing positions. Please let me know what you think!
04/16/07 -23.6% Loss too big because of emotional trading after expected loss. I did hesitate to close a loosing position early. Hope...
04/18/07 +12.2% Big losses always trigger a lot of analytic work. I like to use spreadsheets. Applied statistical data. Traded with the trend. Made some good decisions. There is a long way to go. But I am confident.
The Dow has a specific profile of moves from close to close on a daily basis. I have attached a graphics showing how much the INDU fluctuated in the last 65 days (absolute close to close comparison). Below table gives you an overview about occurences of events (second row) for specific conditions (first row). 25 50 75 100 125 150 175 42 29 17 10 5 4 2 In 42 from 65 cases the INDU made 25 or more points either up or down. In 29 from 65 cases the INDU made 50 or more points... To get a positve permanent expectancy is the target. For example if 42/65 means 42 wins and 23 losses (65-42) one need the win/loss amount ratio to be greater than 44.7% (23/42). If one wins $650 and looses $1,000 in that case the calculation would be as follows (historical data): 42 * $650 = $27,300 23 * -$1000 = -$23,000 Profit $4,300 Avg. Profit is $66.15 per trade (sample size). Worst possible outcome (before slippage, bid-ask spreads and commisions) is 65 * -$1000 = -$65,000 for a linear money management. According the Kelly criterion the position size would be adapted to the account size. The bigger the absolute account size the bigger the total amount risked on each trade is (and vice versa). An edge (excess winning expectancy to 50:50) of 10% (60:40) means that 20% of the payroll is risked any time. In our example we have 10% to be risked (applying full Kelly) while having 42/65 - 26/65 * $1000/$650 f = p - q * risk/gain
Plugging in full Kelly k=1 and for example d=0.1 one can see that the probability of losing 90% (=(1-0.1)*100%) of the current bankroll is whole 10%! The numbers drop very quickly with lower Kelly fractions. For example, for half Kelly the probability of losing 90% of the bankroll is just 0.1%, and it is negligible 0.0000001% for more reasonable fifth Kelly bettor. A reasonable utility function starts probably somewhere around fifth or eight of Kelly, maybe even twentieth of Kelly. For this it is important to note that the ideal definition of bankroll is anybody's investment capital (not necessary currently available in cash). The Kelly fraction should ideally be related to one's whole investment capital, not just some artificial bankroll. It is not completely correct to split your investment capital into more bankrolls for different purposes, and then use higher Kelly fraction. However, most people do this, probably from psychological reasons. Source: http://www.bjmath.com/bjmath/ror/ce.htm
04/19/07 -16.3% I am not good at trading at the moment. Fighting trends and hope. Can't work. No need to max out or force something. Why the pressure?