You can't have losses and still do well...

One of the things that I really try to limit is losses (duh).  But you may be more aggressive about doing the same when you see the following.  First, is a great quote From Ken Fisher, Forbes, 1989:

Here’s the paradox: The odds are overwhelming I will end up richer by aiming for a good return rather than a brilliant return -- and sleep better en route. Folks who seek a killing usually get killed. Gunslingers get shot, and often in the foot, with their own guns. While there is always some guy around on a red-hot streak, his main function is to tempt the rest of us into becoming fools and paupers. A return of 15% to 20% annually is a lot more than most folks realize, or need. If a 30-year-old with $10,000 in an IRA gets 15% annually, he’ll be a millionaire before normal retirement. That’s the power of compound interest. If that same 30-year-old were to sock away another $2,000 per year at 15%, he would end up as a 65-year old $3 million fat cat. At 20%, it’s an incredible $13 million. That’s a lot, but it’s not too much to ask. The two most definitive studies ever on long-term returns, the Ibbotson/Sinquefield and Fisher/Lorie studies, both point to average annual returns for stocks of 9% plus per year going back to the mid-1920s. So 15% to 20% per year is really 66% to 100% better than the market as a whole. That’s tough but doable. Consistency is the key. It is close to impossible to get a good, long-term, rate of return if you suffer serious negative numbers en route. It’s the math. A single year that is down 30% means you have to get 30% per year positive returns for the next four years to get back on track for a 15% annual average. Or, if you score 20% annually for four years, and then suffer a 30% decline, your five-year average return is only 7%.

(This is also quoted fairly regularly by Jeff Saut - over here and here at Minyanville)

A big quote I know, but important.  If you have a large loss, it really is almost impossible to get back to were you would be if you had a more consistent yet lower return. 

If you have a method that can return 10% annually, first off you are doing pretty well.  But if you're shooting for the stars, looking to make the "quick buck", you may want to re-evaluate.  Looking at the returns required over the next 4 years after a loss that must be averaged in order to get back to that 10% annual return.  A 20% loss in the first year requires 4 years of almost 20% returns.  That is going to be pretty tough, especially without taking on even more risk.

The moral of the story:  Protect capital, even at the expense of higher returns, because it is very difficult to recover.