I have often said that if you want to make money trading, then reduce your size. And this is true, as most of us tend to over-leverage. But simply reducing your size isn’t the most effective strategy…how much you reduce it, and the way you size your positions can have significant impact to your returns as well.
So, here I’m going to examine the effect of different trade size strategies that you can employ in your trading that may make your road to riches a bit more enjoyable.
Most books tell you your position size should depend on the amount of risk you want to take, in other words, size your position based on how much you can tolerate losing. And in my experience, that’s probably going to be the end result…losing that amount of money. If we would just step off that band wagon of managing risk, and start thinking more in terms of managing our winners, we would be far better off.
There are lot’s of ways to control the size of you position, I’m going to examine four, and they are:
The trading strategy we will use should be inconsequential, so let’s make it super simplistic, we’ll use a counter trend model employing the Relative Strength Index (RSI) to detect overbought and oversold conditions. This is perhaps the most popular indicator in every chartists heads, and the one most often referred to by the talking heads on CNBC. Of course that’s no ringing endorsement, but we’ll go with it!
In our strategy we’ll only look at long trades, and use the daily timeframe of the SPDR S&P 500 ETF (SPY) going back to October 25th, 1995, that should provide enough data points to make this exercise relevant. We will start with $100k of initial capital to establish a baseline for comparison. I have also included slippage and commission so that the results more closely reflect real life conditions.
Here’s the basic strategy:
|Buy Rule||RSI(2)||Crosses below||30|
|Sell Rule||RSI(2)||Crosses above||70|
Fixed Share Amount
This is a non-risk based method that buys 500 shares of SPY for every trading signal given by the strategy.
Fixed Dollar Amount
This is the second non-risk based method that invests $100,000 of SPY for every trading signal given by the strategy.
The fixed-fractional method is one of two formulas that we examined that incorporates a risk component into the denominator of the trade-size calculation. In this formula, the risk component is decided as a point amount that you would be comfortable risking on a trade. In this test we risked 2 percent of the capital in each trade. Some call this method compounding.
The percent-volatility method was our second risk-based trade-size formula. The risk component in this formula was once again in the divisor, but this time it is computed from the security’s (SPYs) average true range. So as the volatility of the security increases, we take a smaller trade size and as volatility decreases we take a larger trade size. We risked 2 percent of the capital in each trade and divided it by the average true range, times an ATR multiple.
Analysis of Different Strategies
We focused on two types of trade-size methods in this paper. The fixed-fractional and percent-volatility methods each incorporate an element of risk into the divisor. As we mentioned earlier, as the risk element grows (shrinks) in size, the position would be smaller (larger). The other trade-size approaches we analyzed were two non-risk-based techniques: the fixed-share amount and fixed-dollar amount. In the fixed-share approach, we invested 500 shares per trade, and in the fixed-dollar amount, we invested $25,000 per trade. These values did not change over the time period tested.
The main purpose of this paper was to compare trade-size methods that incorporate risk to methods that do not. Risk is a vital concept and can be addressed through different modes of portfolio management. With respect to position size, risk can be a personal dollar-amount preference, as was the case with fixed-fractional formula, or it can be a characteristic (average true range) of the security, as was the case with the percent-volatility formula. One factor that distinguishes the fixed-fractional formula from the percent-volatility formula is that the percent-volatility formula is a dynamic type of position sizing; the trade size adjusts as the volatility of the security increases or decreases. The fixed fractional is dependent on your personal risk preference in terms of dollars per trade. Both the fixed-share and fixed-dollar methods use a static trade size and do not consider any type of risk in the calculations.
The performance impact these trade-size methods had on the RSI strategy positively favored fixed fractional and percent volatility over the fixed-share and fixed-dollar trade sizes.
- The fixed-fractional and percent-volatility methods had the highest annual returns of 5.51% and 6.05% when compared to the fixed-share and fixed-dollar methods, which only had annual returns of 3.30% and 2.06%.
- The net profit values of the fixed-fractional formula (Net Profit: $174,414) and percent-volatility formula (Net Profit: $204,017) were also greater than the fixed-share method (Net Profit: $83,600) and fixed-dollar method (Net Profit: $42,217).
- The fixed-fractional and percent-volatility methods also had the best risk-adjusted returns, with Sharpe ratios of .13 and .17. The fixed-share and fixed-dollar amount methods had negative Sharpe ratios of .09 and .01, although their total return and annual returns were positive for the period tested.
The fixed-fractional and percent-volatility methods only scrape the surface in terms of what is possible in developing trade-size formulas. For example, we could have easily modified the fixed-fractional method to have a dynamic risk amount in the denominator of the formula that adjusts to levels of percentage drawdown that the strategy experiences. Good money management is acknowledged as an important part of strategy trading. Through creative thinking and an understanding of how a risk component can affect the size of a position and, in effect, the performance of the strategy, one can experiment with this area of money management to help improve the risk-and-reward statistics of a trading strategy.