Gate of Medrese (Athens, Greece)
This is a follow-up to the HedgeHunter Risk Parity (RP) post. This blog post will give readers an overview of how RP is calculated. This not for the mathematically squeamish, but the concept is not difficult if one can get past one step in the logic. I agree with HedgeHunter as to flaws in this logic.* I think RP falls under the Schwert Rule, but that can wait for another discussion. Let me explain the RP calculation by first beginning with a reference to a paper (Risk Parity Allocation by Edward Qian) that brought this idea to my attention.
A standard portfolio breakdown between stocks and bonds is 60%/40%. Most investors will argue this is a well-diversified and somewhat conservative portfolio. It is certainly more conservative than most of the ITA Wealth Management portfolios. But is the 60/40 mix a conservative portfolio? I just checked my data sources and found that the standard deviation (SD) of stocks over the past five years is +/-16% and the SD for bonds is +/-3%. However, when one takes variance into account, we must square each number so the 16% has a variance of 256 and for bonds it is 3 x 3 or 9. In terms of variance, stocks are 28 (256/9 = 28.4 or 28 when rounded) times riskier than bonds. That is a huge difference. Even if stocks carried a standard deviation of 15% and bonds 5%, in terms of variance, stocks are nine times riskier than bonds.
To borrow an analogy from Edward Qian, if we go back to the 60/40 stock to bond split, we have six stock eggs and four bond eggs. To calculate the true risk to stocks we find we have 172 (28 x 6 + 4) eggs in total. One hundred sixty eight (168) out of 172 is approximately 98%. Very close to 100% of the 60/40 portfolio of risk is carried by the stock portion of the portfolio.
I need to begin rethinking how a portfolio is put together and Platinum readers would do well to pay attention. The 70/30 or even 80/20 mix is fine when stocks are in the ascendency. But let another bear market strike, and the high stock to bond portfolios are in for another two to three sigma shakedown. That is exactly what we want to avoid. How to pull together the ideas of Risk-Parity and the ITA Risk Reduction model is a worthy project.
* Risk Parity has had a wonderful run due to the long decline in interest rates over the past 30 years. Results look particularly attractive when TLT is one of the “bond” options.
I read this a few days ago. I think Lowell may have referenced the article.
“Asset allocation requires swimming against the tide.
Investors, regardless of whether they are institutional or individual, tend to follow trends
when determining asset allocation decisions. Yet, history has shown that swimming against the tide and investing in unpopular asset classes has been the route to true diversification and investment success. ”
from: Richard Bernstein – Worried about the downside?
http://www.rbadvisors.com/templates/dmd/images/pdfs/worried_about_the_downside.pdf
Rick,
I’m not sure if I referenced the article, but it is worth reading. One of my favorite Richard Bernstein quotes (not to be confused with William Bernstein) is the following.
“Diversification isn’t based on the number of asset classes, it’s based on the correlation of returns among the asset classes.”
This is what the Cluster spreadsheets are all about – finding low correlated ETFs.
Lowell