Position Sizing

I agree with one of the prior posts-
Position sizing is just a smart business decision if you are entering the trading arena.
Most novice traders want to get rich quickly and don't care for the restraint a position sizing approach would seem so limiting....LOL
If you want to survive not only the trend up, but the pullback down, position sizing is your best friend.

iT doesn't need to be complicated- as the prior poster indicated- a simple 10%/1% allocation max risk works well to start-

As your account improves gradually, you get to Risk More.
If your approach hits a hard spot, you get to Risk less.
In the process, you neither get rich quick , nor do you go bust-
You simply get the opportunity to learn and build from your experience.
Every business adventure has it's unexpected moments- Trading can be filled with them- Position sizing will be your new best ally in learning a new business-
It doesn't matter who conceived it- Van Tharp does a good job of articulating it-
Just use it and good luck in your future success-SD
 
it is difficult to know the odds even after setting target and stop. so it is hard to apply kelly before a trade. here is my solution.

trade small to get comfortable. after a while, if your performance stabilizes, you can calculate daily mean and daily standard deviation. your correct kelly leverage relative to your current level is the ratio of mean to variance. say your daily mean is 1% and your daily standard deviation is 5%, then 1% / (5%)^2 = 4. So you can go 4 times bigger in terms of position for optimal kelly. That will lead to a new mean of 4%, but a standard deviation of 20% daily. If the mean variance ratio is 0.5, you need to cut current sizing by half. You never want to bet more than Kelly, but you can bet less to fit your risk tolerance.

for me personally, my mean is about 0.2-0.25% and my standard deviation is 1% daily. So while theoretically I should leverage up 20-25 times more relative to my current sizing, I am limited by margin requirement set by the broker and exchange and I am also quite comfortable with my dd.
 
Quote from njrookie1:

say your daily mean is 1% and your daily standard deviation is 5%, then 1% / (5%)^2 = 4.
More...

Daily mean of what? of returns? A daily mean return of 1%? I must be doing something wrong if you mean that.:)
 
I mean daily return.

I am just giving you an example. A 0.2 mean/std ratio, or sharpe ratio, on a daily basis is pretty common I believe among traders.
 
Quote from njrookie1:

it is difficult to know the odds even after setting target and stop. so it is hard to apply kelly before a trade. here is my solution.

trade small to get comfortable. You never want to bet more than Kelly, but you can bet less to fit your risk tolerance.

So while theoretically I should leverage up 20-25 times more relative to my current sizing, I am limited by margin requirement set by the broker and exchange and I am also quite comfortable with my dd.
More...

Before I started trading oil futures live I asked an experienced trader about position sizing and he suggested I start with minimum size (1 lot) and he suggested a monthly profit goal easily attainable if day trading this instrument. If I was able to reach this goal, add a contract the following month with the profit goal being twice the 1-lot goal. If I was unable to reach the 2-lot profit goal, back down to 1 lot.

Despite reaching the profit goals, I hit a psychological wall with position size. I found that I cut my winners quite short with larger size on because the profit I saw was larger. Then I had a tendency to stop trading or become less aggressive after a certain gain was realized, which makes no sense. It was totally irrational, but I struggled with the problem until somewhat recently.

The solution for me was to remove my P/L from view and focus on excellent execution of my trading plan regardless of size. Without the unrealized and realized P/L numbers in front of me, there was nothing to influence my trade management when trading larger size.

I keep a spiral bound notebook in front of me, jot down entry time/avg price and exit time/avg price and at the close of each trade I note the # of ticks profit or loss. These are very small numbers (i.e., +21, -14, +30, -7, +1, +1) and so it has no irrational influence on my trading. This way I can ascertain at-a-glance when the trading is getting choppy, when a trending move is showing signs of reversing, and when the time of day is resulting in diminishing opportunity, just by seeing the tick totals in isolation from actual P/L.

I'm sure a lot of traders have no problem with this, but it was a great solution for me :)
 
Quote from NoDoji:

The solution for me was to remove my P/L from view and focus on excellent execution of my trading plan regardless of size. Without the unrealized and realized P/L numbers in front of me, there was nothing to influence my trade management when trading larger size.
More...

Then how to you execute your stops if you don't know the P/L? In advance you place them?

Can you list the 20 or 30 most recent actual trade P/L figures so we can get a better idea of what you are after? You sound like a random-walk trader to me.

I may be wrong but I get the impression you do not actually trade. But I apologize in advance if I get the wrong impression.
 
Quote from plsnobull:

Then how to you execute your stops if you don't know the P/L? In advance you place them?

Can you list the 20 or 30 most recent actual trade P/L figures so we can get a better idea of what you are after? You sound like a random-walk trader to me.

I may be wrong but I get the impression you do not actually trade. But I apologize in advance if I get the wrong impression.
More...

I only execute limit orders, the market executes stops :p

I place my orders based on technical price patterns on a chart. I don't trade directly off the chart (though have been considering this for some time now), I use the DOM and I have the price ladder so I know the price at which to place my orders without having to see a P/L column. Seeing the P/L was not of benefit to me; it had a negative influence. Many experienced traders on ET have said to remove your P/L from the screen, and I always thought that was insane until I realized that the MOST important part of trading is following one's trading plan. Execution of the plan is EVERYTHING; The P/L is irrelevant except at the end of the day/week/month/year (depending on your trading time frame).

I don't know what a random walk trader is. I'm a trend-follower who will only take counter-trend trades if a reversal signal appears and there's enough airspace between my entry and a logical price pullback level.

I shoot for a minimum profit of 20 ticks, but sometimes in ranges or choppy conditions I end up with less. I will not place a stop loss greater than 19 ticks, with my average stop being around 13 ticks.

An example of some consecutive trade P/L's:

-7, +13, -0, +20, +16, +35

+1, +20, +10, +50, +2

-0, +20, -0, -0

During a period of time when a strong trending move is underway (all trades same direction):

+21, +21, +4, +50, +20, +25, +55, +27, +2

That run explains why I'm now a trend follower (I used to be a 100% counter-trend trader).

Consecutive trades during a period of major choppiness:

-9, -0, -0, +11, +20, +11, -0, -0, +16, +21, -14, -0

With regard to any impressions you get about me or my trading, that's solely your business and totally out of my control :)
 
for me it's pretty simple. I add to winners and I add to losers and I deal in units. A starting postition is always 4 units. A unit is equal to the account value

so 10k account 4 units would be 40k

50k account 4 units equals 200k

so for example 1 ES is worth about 72k, so with a 72k account I would put on 4 contracts to start with and then add from there.

If I had less than 72k I wouldn't trade ES
 
You must log in or register to reply here.
Back
Top