In a recent Forum post Ernie Stokely raised the subject of position sizing. Ernie also hinted that I might not want to get involved in the Forum discussion. He is right – this is not a simple and straightforward subject and can get very mathematical/technical and beyond the desires or interests of most Platinum members reading this site or, indeed of most investors. It is also the tip of a very large iceberg. However, I would be remiss not to say something on the subject for those members that may have a deeper interest – but I suspect that we will not be going into a lot of detail on this subject on this site.
In response to Ernie’s initial post Lowell provided a link (http://stansberryresearch.com/investor-education/position-sizing/) to an article that introduces the concept of R-Multiples. This is a useful, but very simplistic, introduction to the subject of position sizing (using R-Multiples) that has been described in detail by Van Tharp. For anyone interested I would strongly recommend Van Tharp’s book “Trade Your Way to Financial Freedom” (http://www.amazon.com/Trade-Your-Way-Financial-Freedom/dp/007147871X/ref=sr_1_1?ie=UTF8&qid=1445779849&sr=8-1&keywords=van+Tharp). This is an excellent book and covers a lot more information related to building a trading/investment system than just explaining R-Multiples. However, it covers the subject of R-Multiples in much more detail than in the Stansberry Research article.
One of the difficulties comes in understanding how the R-Multiple concepts might be applied to “Investing” rather than “Trading”. As Ernie suggests, and I agree, this is just a matter of the time scale in which we think and in which we plan to trade/invest. Most books and articles on this subject tend to focus on “trading” and this generally means trading individual stocks on a short term time scale. When we focus on ETFs we are (usually) investing in a basket of stocks (as a single asset) that will represent far less risk than a single stock holding and we will plan to hold for longer periods of time (generally longer than 1 month) – therefore we need to keep this in mind. It does not mean that we cannot think in terms of R-Multiples but we need to think of them in a different way. For example, using the examples given in the Stansberry Research article we would probably never invest 50% of our portfolio in a single asset since we would have to place a stop loss order at 4% below our purchase price to keep our trade loss below 2% of our portfolio value (assuming that 2% was our maximum discretionary risk limit). This is a “tight” stop loss for an “investor”.
A lot of justification for position sizing comes from game theory and the concept of “Risk of Ruin”. This can be “googled” for more information. However, this leads us to concepts such as the Kelly Criterion and more complex allocation calculations, such as Optimal-f, introduced into trading by Ralph Vince (http://www.amazon.com/New-Money-Management-Framework-Allocation/dp/0471043079/ref=sr_1_6?ie=UTF8&qid=1445783273&sr=8-6&keywords=Ralph+Vince). This is all very mathematical and complex and, for most of us, far too difficult to implement in a useful practical manner. Furthermore, practical applications suggest that “recommended” allocations may, in practice, be too generous.
I mentioned above that we need to think rather differently when we are considering an investment “system” to trade ETFs. In this case we need to know and understand how our “system” performs i.e. win/loss ratio, average winning trade, average losing trade, system expectancy (again, see Van Tharp for more detail – also my post at https://itawealth.com/2015/02/20/expectancy/), Maximum Adverse Excursion (MAE), Maximum Favorable Excursion (MFE), drawdowns etc… This will lead us back to R-Multiples in line with the fact that we are “trading” multi-asset vehicles (ETFs) in our account.
Howard Bandy has written a number of good books on system development but his latest book, “Quantitative Technical Analysis” (http://www.amazon.com/Quantitative-Technical-Analysis-integrated-development/dp/0979183855/ref=sr_1_1?ie=UTF8&qid=1445785076&sr=8-1&keywords=Howard+Bandy) is an excellent up-to-date review of many “best practices” in system development, including the use of machine learning, and offers insights into a new algorithm for position sizing that he calls safe-f. Unfortunately, Bandy uses AmiBroker and Python to demonstrate the techniques covered in the book – so this requires some work to convert code/ideas to other platforms – however, it does not detract from the ideas/methods described in the book.
I think you might be able to see why I am reluctant to get into this subject on this site – I feel it is too complex for most investors managing their own portfolios. However, this does not mean that I think it is an unimportant subject – indeed I feel that all investors should at least be aware of the subject and have some form of risk management (and position sizing?) in their investment plan.
In the past, at ITA Wealth, we have covered a number of allocation plans, including Strategic Asset Allocation (SAA) Plans for “Passive” management or more active allocation options that might be applied to a momentum based strategy, but have not directly addressed the question of position sizing except, perhaps, in the use of risk parity allocations to align position size to asset risk. I have always promoted the “if in doubt, equal weight” recommendation – and this may still be the safest plan for most of us, even if not optimal – if used in conjunction with a risk management exit plan. Furthermore, we have not really considered “Cash” as a position/asset other than in the case where we have no assets qualifying for inclusion in our portfolio (i.e. ranking higher than SHY in a momentum strategy). Thus it becomes a default position rather than a chosen/planned position. However, there are systems that might use “Cash” directly, as an asset, to control risk – these are described in Bandy’s book.
Bottom line, be aware of the subject of position sizing and decide, according to your comfort/interest level, how far you are prepared to go with it – it is not a trivial subject with a simple recommended action. Van Tharp is fairly easy to read (I have all the books he has written) and, even if you decide not to follow this path, I would strongly recommend his books for insights into the psychology of trading/investment, system development and particularly in the importance of Risk Management, including position sizing. The other books (particularly Ralph Vince) are very technical and probably only of interest to the truly dedicated investor with sufficient interest and time to get into the details (and, even then, some of it requires a very strong mathematical background).