Use access key #2 to skip to page content.

TMFPostOfTheDay (< 20)

Backtesting and Risk Parity



July 02, 2013 – Comments (0)

Board: Macro Economics

Author: yodaorange

Backtesting of investment strategies has become very mainstream. Here at the Fool, the Mechanical Investing board focuses on many strategies proven out using backtesting. The board has one or two backtesting programs/data/websites that allow users to see how stock picking strategies would have performed in the past. There are also several commercially available backtesting websites sources like and With a little bit of learning, backtesting is within reach of most investors.

Invent a “new” strategy, prove it out using a backtester and get ready to reap the rewards! Sounds too good to be true and it often is. I do not closely follow the Mechanical Investing board, but have seen Mungofitch caution many newbees not to expect future results to be as impressive as the backtest results.

Most individual investors use backetesters for stock picking only. Their databases might only have 20 years of data for US stock. Institutions also have backtesting capability. In many cases, their backtesters have more data, particularly with the addition of asset classes other than stocks. Different types of bonds, currencies, and foreign stocks for example that are not as readily available to the individual investor.

A “new” investment strategy called Risk Parity aka RP came to light in the last few years. A lot of top notch financial researchers, including Cliff Asness/AQR backtested and researched RP. [1] Ray Dalio runs the massively large hedge fund, Bridgewater Associates which pioneered the All Weather Strategy [2] which was the pre-cursor to RP. Ray developed the All Weather Strategy in 1996 and has successfully implemented it. Bridgewater’s success prompted other researchers to backtest similar strategies, like AQR did. RP is also related to the “Permanent Portfolio” introduced by Harry Browne many years ago.

Briefly, the strategy behind RP was that each asset class should be scaled to achieve “Risk Parity.” Backtesting showed that stocks were about 3X “riskier” than bonds. RP solved this by using leverage to buy 3X the bonds, compared to stocks. This makes the stocks have the same risk as the bonds. Ditto for the other asset classes.

Several large funds read this research and decided to introduce RP funds. Their timing was PERFECT! With the recent run-up in long term interest rates, bond prices fell. You can imagine what happened when you had 3X leverage on your bond holdings. The losses were magnified by the same 3X. It was NOT pretty.

For some reason, last week was “Lets Beat up Risk Parity Funds” week for many publications. Everywhere you looked was a story about how poorly RP funds had performed recently and how traditional 60/40 balanced funds had performed better. The story even made it to USA Today [3] in addition to the Wall Street Journal. [4] Graphs of four RP funds are here. [5]

BOTTOM LINE: is that backtesting is NOT a panacea. Maybe RP funds will outperform over the long term, but in the short term it is ugly. RP joins a long list of other “strategy of the day” approaches that worked well in backtests . . . only to disappoint when implemented.

BTW: I have a lot more to say about backtesting in a future post.



[1] Cliff Asness, et al: Leverage Aversion and Risk Parity>/i>

[2] Bridgewater: All Weather Strategy

[3] USA Today: Wall Street’s newest fad: Risk-Parity funds

[4] Wall Street Journal: Fashionable 'Risk Parity' Funds Hit Hard

[5] Risk Disparity

0 Comments – Post Your Own

Featured Broker Partners