| 
 PureBytes Links 
Trading Reference Links 
 | 
List: 
 
I've been thinking recently about the subject of risk and stop losses.  We 
always hear that we shouldn't be risking more than 1-2% of our accounts on 
any one trade (risk defined by your stop loss' distance from your entry 
price). 
 
Recently I read William Gallacher's revised edition of  "Winner Take All" 
(excellent book, BTW).  He discusses at length an idea I have heard echoed 
by some other great traders over the years: the idea that, rather than 
stops being placed according to a 2% risk, that they really should be 
placed "correctly" for your system, which is to say they should be at a 
point where it is "a violation of the reason why [you] wanted to get into 
the trade in the first place" (Tom Basso).  For many trend following 
stop-and-reverse type systems I would imagine that such an exit point would 
be the mirror image of your entry point, rather than 2% from your 
entry.  This point seems reinforced by the fact that when I introduce fixed 
percentage stops (volatilty based as I am using a position sizing model) 
into my own trend following system, the accury (% winning trades) goes down 
substantially.  This tells me that the stops are placed 
"incorrectly."  Gallacher states: 
 
...the severity of an equity drawdown to an account is almost exclusively a 
function of the size of the position habitually put on in that account. 
Some commentators downplay the exposure issue, suggesting that a trader 
protect himself from "overtrading" by limiting the amount he risks on each 
trade to a fixed percentage of his equity....I intend to show that the 
question of risk percentages is academic and that only exposure matters in 
the long run. 
Let's say...a trader with $20K true risk capital anticipates capturing an 
80 cent move in soybeans ($4000/contract). The trader has been reading a 
book that cautions him to risk no more than 5% of his equity on any one 
trade. Accordingly, he risks $1000 on one contract, hoping to make $4000: a 
trade with a risk to reward ratio of 1 to 4. 
Now, imagine this same trader reading elsewhere that he should risk no more 
than 2 percent of equity on any one trade. If he takes this advice, the 
trader may still trade one contract, but may risk only $400. He is still 
hoping to capture the same $4000...move and is therefore working on a risk 
to reward ratio of 1 to 10 -- or so he may think. 
Let us compare the two approaches to equity preservation....Both are 
equally exposed to a sudden sharp price change, since they are both holding 
the same position size....The 2 approaches differ only in the amount 
initially risked, and if both pursue the same consistent strategy of going 
for that 80 cent move in soybeans, the strategy risking $400 can expect to 
be stopped out 2 1/2 times as often as the strategy risking $1000; it can 
therefore expect to experience the same magnitude of drawdown over the same 
period of time.....Risk to trading equity cannot be reduced by reducing the 
amount risked on each trade. You can drive from Toronto to Miami in one day 
or spread the driving over three days; it still takes the same amount of 
gas to get there....it is very much a question of paying now or paying 
later....[With stop and reverse systems] it is the market, not you, that 
decides how much you are risking. This...truth is perhaps clearer when you 
consider the operation of reversal systems [ie systems that are "always in 
the market"]....In a reversal system, market action dictates absolutely the 
amounts that are risked on each new position. Such systems cannot possibly 
risk fixed amounts, and it would be ludicrous to try and constrain them to 
do so.  (William Gallacher, "Winner Take All") 
 
 
Please keep in mind that I am assuming you are using some sort of position 
sizing model, and not that you are trading overly large positions for your 
account size and thus forcing your "2% stops" to be arbitrary and 
artificially tight.  For myself, I would be using the model: 
 
(.01 x Account Size) / (N * BigPointValue) 
 
where N is the 20 day ATR. 
 
So 2% stops would be equal to 2N for each market traded.  So I am not 
talking about trading 10 bond contracts in a 30K account and then risking 
2% on such a position. 
 
I welcome any opinions, or comments. 
 
Thanks. 
 
David 
 
 
 
 
--- 
 |