Risk parity- something you should really consider


In an earlier article I wrote, I advocate on the importance of allocating and balancing your portfolio to achieve optimal returns based on the risk level you are willing to take. One of the important concepts illustrated in that particular particle is Tangency Portfolio. For the uninitiated, the Tangency Portfolio refers to the point where the Capital Allocation Line (CAL) intersects with the Efficient Frontier. (You might want to re-read that article to refresh your memory). It is an important point as it represents the most optimal combination of the various asset classes to provide you the maximized expected return for a given risk level, with the risk-free rate of return accounted for.


While this might sound good in theory, it is practically very difficult to achieve a tangency portfolio in real life. As you might have notice, the tangency portfolio tends to be constructed based on hindsight. Usually, it’s already too late for you to adjust your portfolio to model after the tangency portfolio when you know what the tangency portfolio is like. While hindsight might be 20/20, it’s not exactly useful for an average mortal like me/you who do not have the ability to look into the future. We would need to be aware of other concepts which could potentially bring us to a result as close to an optimal portfolio as possible in real practical terms. And hence, the topic to be addressed today- risk parity.


Risk parity refers to a kind of portfolio allocation strategy which uses risks to determine the allocations across various asset classes/components of your investment portfolio with the goal of ensuring that you are exposed to the same amount of risk for your investment in each of these asset classes/components based on the correct weightings. Ultimately, risk parity allows investors to be able to target specifically a certain level of risks in your portfolio and hence divide the risk across various asset classes/components of your investment portfolio to achieve optimal results. In essence, the risk level of the various asset classes/components of your investment portfolio will be the determining factor behind how much weighting each of these asset classes/components should have in your portfolio.


If there is a particular asset class/component in your portfolio that is particularly high-risk and exceeds the optimal risk target, the weighting for their particular asset class/component should have a smaller weighting as compared to the other asset class/components in the investment portfolio. Unfortunately, this was not always considered in most traditional portfolio strategies. For example, I'm pretty sure most of us have heard about the 60/40 portfolio whereby 60% of the portfolio is allocated to stocks and the remaining 40% is allocated to bonds. In this strategy, the differences in risk levels based on the weightings of stocks and bonds is rather extreme with 97% of the risk attributed to the 60% weighting of the portfolio in stocks. Hence, you might be exposing yourself to too much risk in this case. Of course, all is well when stocks consistently outperforms bonds. But could you be sure that this will always happen? If you don't, then it will be important to consider risk parity.

At this point of time, you might be wondering what risk parity has to do with achieving a portfolio mix as close to the optimal portfolio as possible in real practical terms. Well, here is the answer. Incorporating risk parity in your balancing or allocation of portfolio might potentially bring you to a result that is close to the optimal portfolio in real life. And simulations were actually done to prove that. In some studies done by banks (eg. JP Morgan's research here) , they tried to compare the performances of two types of portfolios. One is a portfolio which has risk parities considered while another is a portfolio which follows a "textbook" template. This "textbook" template refers to an approach where you simply follow the "tangency portfolio" provided by historical estimates of the various asset classes/components in your portfolio. It's not that there's anything wrong with this "textbook" template. It is just that ta "tangency portfolio" built on historical estimates will always have a certain degree of uncertainty which cannot be accounted for in real world performances. And that is also the reason why a "tangency portfolio" is always good, and only good in hindsight. In those simulations, it was found that the portfolio with risk parities considered seems to outperform the "textbook" strategy 80% of the time.

Of course, similar to all strategies, there must also be certain downsides and the risk parity strategy is no exception. For example, the risk parity strategy usually works well when the sharpe ratio of the various asset classes/components in the portfolio are close to each other (eg. within a 0.1 range) but not so well when the sharpe ratio of the various asset classes/components are too far apart from each other. Divergence of the sharpe ratios have an impact on the effectiveness of the use of risk parity in your portfolios. And hence, the risk parity strategy is not one which works well in all scenarios. But generally speaking, the closer the share ratios of the various asset classes/components in your portfolio are, the more sense it is to adopt the risk parity strategy.


If you are thinking where to get started to explore the use of risk parity in your portfolio, you might like to take a look at Portfolio Visualizer. They have a lot of great tools which are free to be used by the general public like you and me. One of their tools is Portfolio Optimization which easily allows average joes like us to be able to experiment the use of risk parity in our asset allocation/balancing. I have been experimenting with it of late and find it really useful to understand better how to optimise my own portfolio, and personally strongly recommend you to check it out.


Till the next time.


Last but not least.. please scroll down and subscribe for regular content if you like what you read!

262 views
  • Facebook

©2019 by datascienceinvestor