Yesterday we defined our subjective risk thresholds. Today we’ll look at the metrics – the objective approach to defining them.
There are certain key numbers that I collect from my equity journal. Remember that while I will be speaking in terms of ‘dollars’, I also collect the numbers for ‘% of initiating price’ (see The Art Of Position Sizing):
- Average Dollar Win
- Win Rate
- Average Dollar Loss
- Loss Rate
- Maximum Drawdown
- Maximum Positive Return
- Maximum Consecutive Wins
- Maximum Consecutive Losses
- Average Number of Consecutive Wins
- Average Number of Consecutive Losses
- Standard Deviation of Monthly Returns.
- Average annual return
The information provides the objective bedrock for estimating my normal position size. Let’s illustrate my approach by way of example let’s say subjectively I accept a 4% as my maximum risk. My maximum number of consecutive losses is 3 and the average loss is 1%. I can now estimate that the objective maximum percentage loss.
I multiply my maximum number of consecutive losses by 3. So now the number is 9. So my possible worst case drawdown on average would be a 9 x 1% = 9%. I then multiply the standard deviation of monthly returns by 3. Let’s say that comes out at 27%. I now have the boundaries for my worst case scenario: 9% to 27%.
I can now estimate how long it would take me to recover from a worst case scenario since I know my average annual return. If it’s around 25%, then a 27% loss would take me a year.
The idea is to play with the numbers so you know the most comfortable normal risk for you. This number is partially subjective and partially objective. Once you have this number, you can then adopt a position-sizing algorithm. There are as many algorithms as there are successful trading methodologies. Our job is to find a comfortable one for us.
We need one more factor before using a position sizing algorithm: a measurement of the volatility of the market e.g. the Average True Range (ATR) of ‘x’ days. With this we can have a look at some position sizing formulas.
One of the simplest is the Turtle formula.
(% Capital to Risk x Capital)/$value of Average True Range (ATR)
The ATR is the volatility component that the market brings to the equation, we bring the rest. Notice that too often the ‘% Capital to Risk’ is a figure plucked out of the air e.g. 2%. But if you don’t do the work, you will not know whether that figure is appropriate to you and your style of trading. The purpose of position sizing is balance ‘maximization of profitability’ with ‘minimization of risk of ruin’. If we adopt a random figure we’ll never know whether we have struck the appropriate balance for us. Moreover by doing the numbers we have a ‘feel’ for our run of losses and wins. In this way, we prepare for the drawdowns (and the accompanying anxiety) and exuberant profits (and the accompanying euphoria).
One other point. The ‘Capital’ is the cash you have at the end of your measuring period (weekly, monthly, quarterly) +/- open profits as measured from the stop loss. For example: your cash is $100k and you have an open position profit of +$5k. However, your current stop is at -$3k. Your ‘Capital’ is $100 – $3k, not $100k + $5k.
I recommend you set a measuring period rather than calculate the ‘Capital’ on a trade by trade basis. Accompanying the measuring period, you set a high threshold and a low one which, if hit. you would re-calculate the “Capital’. I use a monthly measure and a threshold of 50% of my average annual return as the upper threshold and 25% of ‘three times the standard deviation of monthly returns’ as my lower threshold. This means I reduce normal size more quickly than I increase it.
So now you have a normal size position. But this is not the end of the story. I believe that the position size should be varied depending on the context of the trade. To do this, I use the standards:
- Probability of Success
- Where I am in the Ebb & Flow
Of the two, the Ebb & Flow is the more important factor.
The probability of success is easy enough to understand. You can calculate the probability of success for your rule (setup) if your equity trading journals record the Rule number for a trade (and yours should). This historical probability will be tempered or enhanced by the current context.
So lets’ say that your ‘313 Outside’ setup has a 68% probability of success (normal position size). But on this occasion all the timeframes line up and you feel that this raises the probability to 85%. You may want to raise the position size to 1.5 times normal.
The Ebb & Flow is based on my view of the markets and trading plans. I see our plans as an inlet onto which the waves (the market) wash. Sometimes they only partially fill the inlet. This is the norm. It means we win some, lose some. In this situation, we use normal position sizing. Sometimes the waves cover all the inlet. At those times, we can do no wrong; so we’ll look to increase our size. Sometimes the waves have all but totally receded. At those times, we can do no right; so we’ll look to reduce position size.
We identify the Ebb & Flow by our trading results. If we start to see losses above the norm, especially if we have increased size after a prolonged exuberant profitable run, we’ve moved from flow to ebb. It’s important to understand that, like selling the top or buying the bottom, we’ll always be slightly late in the identification. In other words, we won’t know until after the first few ‘environment changing’ trades that a transitional stage is happening. That’s OK; it’s better than not catching it at all.
In my next blog, I’ll look at the USDJPY and include a position sizing example.