How to Really Set and Accomplish Your Financial Goals
In the last article I described the detailed process of goal setting and used the example of weight targeting; now we are going to relate the same exact process to financial targeting via stock trading.
In this case, specific goals will not be “inevitable” over a fixed period of time like they essentially are with the weight targeting if we follow the procedures. There is too much uncertainty, variability and chaos involved in trading. However, what you can do it plan for an “exact probability” of achieving a particular goal or more given certain assumptions with the right tool, or an approximate probability given that the assumptions are also approximately accurate.
First, set goal and break it down. You want to return 100% with lower risk. That might translate into a consistent and reliable 6% per month compounded. You might look at trading system expectancy and use 1% of risk per trade and determine how many trades you need to complete. For example, if you earn an increase of 1 times your risk per trade you need 100 trades with 1% risk over the course of the year.
Realistically though, there are some issues that we didn’t have with the weight-loss example. The above measurements are flawed. If you had a 50% drawdown and you need 100% gains to get back to even so if you have a 50% drawdown and 100% loss, the “average” result is 25% gain but the net result is still no change. So you need a tool that turns this into more accurate and actionable information. Additionally, success is not inevitable and therefore, you only have a particular probability of obtaining your result in a year, given your strategy.
I created a tool that would help me simulate results given certain expectations and assumptions and compare results at different levels of risk.
If the system above produced 200 trades a year I could simply look at the simulated results of thousands of simulations and determine estimated probabilities to measure expectations and risk level.
Based upon the above assumptions, I was able to run 5000 simulations and determine:
1) The number of times out of 5000 (probability) that my portfolio is down at all after 200 simulated trades.
2) The number of times out of 5000 (probability) that my portfolio is up 100% or more after 200 simulated trades. (probability my goal is reached)
3) The number of times out of 5000 (probability) that my portfolio “kill switch” is hit at all after these 200 simulated trades with kill switches of:
a)15% b)20% c)25% d)30% e)35%
4) The probability that goal is met with kill switches of
a)15% b)20% c)25% d)30% e)35%
Using these I am able to monitor actual goals based upon what I consider a realistic set of expectations to balance my personal desire for reward with my desire to avoid risk and avoid drawdowns. There is always a tradeoff. From there, I am able to adjust position sizing to determine what position size most accurately reflects my goals.
Chance you will be up after 200 trades given position sizing of:
1%:87.5% chance (fees eat into results increasing change of a decline)
Chance you will produce a 100% return after 200 trades given position sizing of:
You may be looking at this like “oh, sweet, I guess I should risk larger position sizes”. You’d probably be wrong.
The greatest danger is in declaring the results “aren’t good” they “don’t work” or “they don’t represent reality”. It is balancing the odds of their being “something wrong” with the system as large drawdowns occur, with your own personal ability to continue to trade the system as well as your own personal ability to withstand large changes in account size. This is why determining the probability that drawdowns occur is valuable.
Let’s look at the probability of a 15% drawdown from highs occurring with this system at different position sizes:
Now let’s relate this to the probability you reach your 100% goal WITH a 15% drawdown in place:
Not letting the system work comes at a cost. But are you going to just let the system work even when your account continues to decline? That also comes with a risk that the initial assumptions about the system were wrong.
But let’s say you have faith/conviction/high confidence in the “system” and understand what you are using has a high degree of leverage. So you are emotionally calm even in the face of a 25% drawdown, what then?
What are your chances of a 25% drawdown over 200 trades given position size, given the assumptions are correct?
As a result, what are your odds of a 100% goal being met over 200 trades if you quit after a 25% drawdown?
A 3% position size may still increase your chances of accomplishing your goals at the cost of stress,
If you have the peace of mind of a Zen master, or don’t mind the stress induced ulcers, vomiting, heart attack, etc… Then you may be able to endure 35% drawdowns or greater.
What are the chances of a 35% drawdown occurring with given risk size?
As a result, what are your odds of a 100% goal being met over 200 trades given that you quit after a 35% drawdown?
Let’s say you create an intelligent robot so that emotions aren’t an issue and for health issues you intentionally are not even going to be aware of what is happening in the system until 1 year from now. Only a 50% drawdown will automatically cause the robot to stop trading. What then?
Chances of 50% drawdown at given risk size:
1%: less than 1%
2%: less than 1%
100% goal met over 200 trades as a result
The previous image is what the relationship between risk and volatility of account is assuming:
1) Unlimited time. (With limited time the results either ARE better than expected or worse. The probability of worse than expected results increases as you increase risk when time is limited. )
2) Certainty in the set of expectations. (The reality is uncertain and therefore it rewards risking less.)
3) That you can weather HUGE volatile swings. Note specifically that the cost in terms of volatility skyrockets as the benefit flattens out and then declines.
4) That fees are never an issue. Fees can make both very aggressive strategies due to volatility and very conservative strategies unprofitable.
5) That your edge remains consistent and results are normally distributed over time. If they are not, it favors more conservative risk.
6) That there is zero correlation between trades as correlation increases, the benefit for additional trades begins to decrease.
7) Probably some other things I haven’t considered.
With a finite time, increased risk skews results so that very few outliers of phenomenal results skew the “mean” results so the mean is greater than the median. The Distribution of Results at 1% is relatively normally distributed but with 3% risk, 5000 simulations of results are distributed like this:
The simulation allows you to better understand the relationship of returns given that these assumptions are different by running simulations with different numbers. You can manage your expectations IF the results vary in either direction to better manage uncertainty.
So how you make financial goals should really contain in it not only a very deep understanding of the type of drawdown you want to endure, the awareness of the cost, but also contain the probability of a particular result with an understanding of the downside and so on. For example, your “goal” should really be considered after running some simulations and broken down to:
1) Probability of achieving goal (you may even consider probability of exceeding it by a particular amount)
2) Probability of breaking even
3) Probability of avoiding certain drawdowns at various levels
4) The Average Expectation of results
For example you might set the following goal:
1) 100% return or more
2) 90% chance of breaking even or better
3) Probability of a drawdown greater than 35% less than 10%, probability of drawdown greater than 20% less than 25%.
4) Probability of obtaining goal greater than 40% with a kill switch of 40%.
Given that goal, and GIVEN the expectancy of the system described, you can determine what position size meets those goals and if not what the tradeoff is (from which you might adjust your goal). You will then find that a 3% position size is too much, a position size of 1% is too small. So you need something between 1 and 3% 2% works but a 1.8% or 2.4% might also work. It depends on if you’re more willing to take a better average and greater probability at expense of more volatility and greater probability of drawdowns, or prefer increasing the chances of not having a bad result and decrease the volatility and drawdowns. Alternatively, you might actually wish to define a range and be a bit discretionary with position size as long as it falls in the range in exchange for less awareness of what your odds of attaining your goals. Let’s say 1.8% is the smallest position size that will meet all of your goals and 2.4% is the largest position size. You might set a RANGE of position size, and then aim for 1.8% when your confidence is low and/or R/R and/or expectancy is lower than the average for the system, and aim for closer to 2.4% when it is higher. Additionally, you might also decide that there are certain instances of the market where you get a good oversold signal and market breadth has just flipped after a serious decline and you have a larger than usual edge across the board. You may want to position conservatively in all other instances so that you have the capital and ability to take on new aggressive positions when this moment comes aiming for maybe 20% of the year to make up 80% of your gains. There are many ways to break it down, but ultimately the combination of your system, your tolerance to drawdowns and volatility, your desired distribution of results, your need for awareness of expectancy and position size will help you to monitor and achieve your goals with the help of a position sizing simulator.