Concerning risk per trade sizing
|April 4th, 2012, 08:48 AM||#151 (permalink)|
Futures Experience: Advanced
Platform: Ninjatrader, Tradestation & TWS
Broker/Data: Optimus, Tradestation & IB
Favorite Futures: ES, EURUSD
Posts: 9 since Jun 2010
Thanks: 2 given, 131 received
Thanks for taking the time to read my post and respond with comments. I will also check out the thread on Risk of Ruin. I'm glad you highlighted this point as I probably didn't expand on it enough due to the length of the post and wanting to get across in a succinct way the importance of changing one's mindset to that of probabilities and focusing on the process not the outcome of each individual trade.
You are correct in your thinking that you can always DEFINE your risk without knowing your objectives. However, knowing HOW MUCH to risk is different. Position sizing has 2 key elements - 1) number of ticks risked 2) $ amount risked. Let's say that after conducting your analysis, you have determined that the most efficient way to place protective stops is based on the recent price action, for example, 1 tick beyond the signal bar's extreme, so if entering a long position, you place your protective stop 1 tick below the signal bar's low. Your risk in terms of number of ticks in then known and can be easily defined for each trade before entering a position. The next decision you have to make is how many contracts / lots / shares to trade and this is where the $ amount and knowing your objectives comes into play. Let me explain why...
Using exactly the same system, with pre-determined stops + targets, altering the position sizing alone results in huge variations in the equity curve and will ultimately determine if you reach your objectives or not. Every trader has psychological biases that determine how they respond in different situations and one of the secrets to success in trading is understanding how these biases affect you and using this information to turn yourself into an effective trader. If you try to project what you read or learn from others outside of yourself onto the market, you will have little chance of reaching your objectives. I won't detail all the biases here as the post will be far too long and will lose its impact, but Van Tharp has written the best book in my opinion on position sizing, called "Definitive Guide to Position Sizing". A simple example that most will be able to relate to is the need to be right - this is stronger in some traders than others, and those who are affected by it the most, will tend to favour a methodology where they have a higher win % while sacrificing their average win:loss per trade (R multiple). This is generally achieved through more of a short term scalping methodology and/or scaling out of position to lock in profits as the trade progresses.
So once you have determined the expectancy of your system, what is the optimal position size to give you the highest probability of meeting your objectives? This is probably the most important question you can ask yourself as a trader. Many newer traders believe that it is the one that will you the largest return. Slightly more experienced traders will say it is the largest risk % without going bankrupt. However what hasn't been defined in these scenarios are, "What is your goal? and "What is disaster for you?". Only then can you start to design a position sizing algorithm that is optimal for YOU. First you need to appreciate the exponential impact and difficulty of recovering from large drawdowns as explained in the post. Next you need to ask yourself questions such as - what is your definition of failure (at what point would you quit)? What is your definition of success (what equity increase would delight you)? What probability of your maximum drawdown occurring would you be willing to tolerate? You will then be thinking in terms of risk and what you can psychologically handle while at the same time embracing risk, rather than fearing it when you have an inevitable drawdown.
I will give you a personal example, I define my ruin point where I stop trading as 20% drawdown of my total starting equity - at which point I have to stop live trading for the remainder of the YEAR and move to SIM. This is in my own best interest and is there as a worst case scenario to ensure I remain disciplined in my approach staying sharp, focused and patient enough to wait for only my high probability setups. Trading is my life and my full time business, so of course, I want to do everything I can to ensure that the odds of me reaching this point are very small, ideally less than 1%, so how do I go about this? I break this down into monthly, weekly and daily loss limits. So my monthly loss limit is 6%, my weekly loss limit is 3% and my daily loss limit is 1%. Obviously I need to start off with a risk amount that enables me to have enough losers to still reach my objective before hitting my loss limits. This is where there is a slight difference in discretionary vs mechanical approaches. In discretionary trading, you have an additional component to consider, which has a big impact - emotions. Do you notice any recurring patterns in your equity curve after a string of winners / losers? How are you results affected by the length of time you trade? Do your results suggest you are you better just trading part of the session? How will this affect your frequency? In mechanical trading, for example, the worst thing you can normally do is reduce position size after a drawdown as it will skew the probabilities, however in discretionary trading the reason for the drawdown may not in fact be the system but the trader! All these factors have to be considered.
For me personally, I have determined that I am detailed oriented and extremely focused individual who does best just trading the 1st 2 hours of the session and trying to capture the best 1-2 swings that occur during this time. I experience more frustration and pain from missing the moves than I do from being wrong. I also know that if I have 2 consecutive losers, that my read is generally off and I am better to sit out the remainder of the session and protect my capital for another day. Why is this important? Because now I know the frequency of my setups and their relative statistics I can realistically expect (win %, avg win:loss), I can determine what is a realistic risk % to set in order to give myself the best probability of reaching my objective before hitting my daily loss limit. Therefore I know that 0.5% is an optimal bet size for me per trade, as if I have 2 consecutive losers I hit my daily loss limit and I am done for the day. I also know that statistically I can have around 45 losers in a row before reaching my point of ruin and odds of my system reaching that point are less than 1%. All this has to be considered before I can know how much to risk!
As an additional component for those interested, I will expand upon how I manage trades. This will hopefully further explain why you have to understand your objectives before you know how much to risk. As my objective is to ensure that I capture the main swings from the start of the session in the most efficient way, I often go into a trade small to ensure I am positioned if the market doesn't pullback and the breakout is strong, with the intention of adding to it if I get strong follow through or on the pullback. I therefore will often go in risking less than 0.5%, to give me the ability to add to the position, without ever exceeding my max risk of 0.5% per trade should I get stopped out on the entire position. Under most circumstances, my scale ins will be above my original entry and allow me to move up my protective stop, thereby allowing me to reduce risk anyway, however, under some circumstances the pullback may test my original entry and I need to add at a similar price to my original entry, which wouldn't be possible without increasing my overall position risk if I went in full size when I initiated the position. A little bit deep possibly for the topic being discussed, but I thought it may give some food for thought and question the traditional approach to trade management. As I explained in both this post and the previous one - only YOU know what is optimal for you and you find that out through experience and understanding your psychology and making it your edge. Believe nothing, question everything and test it for yourself, only then will you have confidence in your own approach and ability knowing it has a statistical edge and is in alignment with your beliefs about the market.
I could elaborate a lot further on different risk models, but I suggest traders read the work from Van Tharp if they would like to understand it more indepth. The only thing I would add is that altering risk % based on your equity curve is an advanced topic and should only be done by those with a consistent track record otherwise you may be increasing risk after an extended winning streak only to accelerate the drawdown that follows. Most traders have their biggest drawdowns shortly after their biggest winners, so this can be counter-productive. Instead you have to understand and see how you handle the drawdown that follows before considering increasing your bet size. Tracking your equity curve, and altering your bet size based on how you would trade the market can be a useful exercise for advanced traders to further enhance their profitability. An example, would be increasing your bet size after a strong BO and PB, while the equity curve is in an uptrend making HHs and HLs. Again something else to consider for those traders at this level. Same can be done to the downside to reduce risk when your trading may be going through a difficult stretch.
Trade well and feel free to ask questions
Last edited by jamiej83; April 4th, 2012 at 09:08 AM.