Quote from Trend Fader:
Assuming we use a %5 fixed stop loss on ABC .. which would put the stop at $95. Assuming the max i allow for trade $ loss is 3% of $50k... which is $1500... position size would be 300 shares @ 100 which equals $30k which is 60% of account. With my proposed method.. the most we can risk would be 125 shares= 1.25% of account at $625.. which means 25% account allocation to ABC trade.
If I understand your methods correctly, you allow for a 12% risk on your total account (4 overnight holds at 3% risk per trade). Although you diversify your risk across 4 different equities, you expose your entire portfolio to significant risk - significantly greater than the methods I described. While 40% of available capital resides in a single stock (TASR), my total risk remains at 2%. Now, should the number of tradeable signals increase beyond the current one per week average, I would definitely divide the available capital and reapportion my risk parameters.
Again using your parameters, a 15% gap DOWN occurring on your ABC example equity would result in a loss of $15.00 USD per share or $1875 USD in total losses ($15.00 x 125 shares = $1875.00 USD). Whereas, the same 15% gap down on TASR (based on a current price of $11.00 USD) results in a total loss of $1363.00 USD on our 2900 shares (See above post for the math). Because we have already experienced significant unrealized profits with TASR, our "envelope of error" has significantly widened. As a result, the methods you posted fail to improve one's risk parameters until you get beyond the 18% to 20% gap DOWN in price. Again, not impossible, but do the odds of such an occurrence outweigh the potential costs associated with smaller position size?
As smtrader pointed out, failing to take the maximum position size may not compensate for the loss of expected capital appreciation when taking larger positions. In other words, even if an event were to occur causing a 50% gap DOWN in price once every 10 trades, are we better off by taking the maximum size each trade, or do we (as you suggest) plan for such occurrences and only apportion 25% maximum capital?
Let's look at two mythical accounts.
Account One has $50,000 initial capital, uses a 2% total account risk parameter and purchases the maximum shares allowed.
Account Two has $50,000 initial capital, uses a 25% equity allocation formula and purchases the maximum shares allowed.
Both accounts trade Stock ABC @ $10.00 USD per share. Both accounts use a 10% price appreciation target ($11.00 USD). In order to make the math easier for everyone, both accounts purchase the maximum shares based on the initial capital and do not increase position size as profits enter the account. Each account makes ten trades:
Account One:
Using a 2% risk parameter on Total Account Equity, Account One buys 2000 shares (5% stop loss and 2% equal risk) of ABC at $10.00 USD per share. Account A holds for a period of 4 days and experiences a $1.00 per share profit. Account A repeats this procedure for a total of 9 trades. Profits are as follows: 2000 shares x $1.00 x 9 trades = $18,000 USD gross profit.
Account Two:
Using 25% of Total Account Equity, Account Two purchases the maximum shares ($50,000 x .25 / $10.00 per share) for 1250 shares @ $10.00 USD per share. Account Two holds for a period of 4 days and experiences a $1.00 USD profit per share. Account Two repeats this procedure for a total of 9 trades. Profits are as follows: 1250 shares x 9 trades x $1.00 profit = $11,250 USD gross profit.
Both Account One and Account Two experience a a gap open loss of 50% of the purchase price of stock ABC during their 10th trade.
Account One: losses total 2000 shares x $5.00 USD per share = $10,000 USD.
Account Two: losses total 1250 shares x $5.00 USD per share = $6250 USD.
After ten trades (9 profitable and 1 catastrophic loss):
Account One:
$18,000 - $10,000 = $8000 USD in profits
Account Two:
$11,250 - $6250 = $5000 USD in profits
Now admittedly, I have oversimplified the position sizing in order to keep everyone on the same page. To compensate, I created an artificially high rate of "Black Swan" occurrence resulting in a 50% loss every ten trades. Even if the catastrophic event caused the stock price to hit zero, the resulting consequences would have a net effect of a Total Account loss of $2000 USD on Account One and a $1250 Total Account Loss for Account Two.
The above example does not prove one method's superiority over another. Rather, the example shows that based on varying events, one method shows superior results. Based on another set of circumstances a second method yields superior results.
The point I have tried to make is that one needs to effectively quantify the type and amount of risk one wishes to protect against and plan accordingly. The correct use of money management does indeed prevent risk of ruin. However, improper use of money management and risk assessment retard profit growth. A multitude of methods exist and each trader should examine all to determine for themselves which methods provide the best results for their own style of trading.
I recommend the following reading:
http://members.aon.at/tips/moneyMan10.htm
I hope you find the above information useful.
- Spydertrader
that one just seems to go in unexpected directions for me... 
I hope it treats you better!