Trouble For 'Bridgewater Risk Parity' Model

by: Chris Ridder, CFA

Last week, I came across an excellent article explaining the basics of the "All Weather" portfolio, which is attempting to replicate an asset allocation strategy, used by the hedge fund Bridgewater Associates, with mutual funds and ETFs open to all investors.

The "risk parity" portfolio put together had excellent risk adjusted return over 10 Years vs the benchmarks. However, as this article explains, the trouble is with the expected returns in the next 10 Years. Just as winning two Superbowls in a row is hard, a 7.3% average annual return will be very difficult to repeat!

Portfolio 10 Year Return Sharpe Ratio
Bridgewater Risk Parity


VBINX (SP-500) 7.7% .53
VFINX (60% Stock 40% Bond) 8.1% .33

Source: Above mentioned article

The problem stems from the weights, and therefore returns, used in the asset allocation. Based on a 10 Year optimization (with constraints - see article) over 42% of the portfolio would now be in "Nominal Bonds" (VBMFX mutal fund symbol or BND symbol for ETF).

(Click to enlarge)

Source: MyPlanIQ

I went over to Vanguard's website to take a look at the components of the fund VBMFX / BND . This fund attempts to represent the total U.S. bond market. The information, under the tab "Portfolio & Management" shows that "Total Bond Mkt Index" has an average maturity of 7.2 Years, Coupon of 3.7%, and Yield of 1.9%. Over the last 10 years the fund returned 5.07% before taxes, as of Dec. 31, 2012. For investors who use exchange traded funds, instead of mutual funds, here is a table showing matching, or similar, funds.

Mutual Fund Exchange Traded Fund
PEBIX ALD, ELD, EMB, EMCB, PCY (Substitutes- none from Pimco)
Gold (London Fix) GLD
VWEHX HYG, JNK,(Substitutes- none from Vanguard) Link to other alternatives

I then performed a thought experiment to "price" this whole fund as a "bond" using an excel spreadsheet, and then find out what would happen to the returns if interest rates moved up or down. Currently the price for a 7.2 Year, 3.7% Coupon Bond with a Yield of 1.9% is $112.01. In a year if everything stood exactly the same, the total return of this bond would be $3.70 (coupon) / $112.01 (Bond Price) = 3.30%.

This is 1.77% below the 10 Year average. But, the Fed will ride to the rescue again bond bulls argue. I thought, let's see what happens if the FED "pins" this bond, with a current yield of 1.9% (3.7% Coupon - a 1.8% coupon spread), to the zero rate lower bound. The price would become $126.65 for a total return in one year of 16.37%. Of course this could only be a "one off", since now the "bond" yields zero percent, and for the next 9 years we can only clip our coupons.

For simplicity I assume that this "bond" would continuously have the same term and yield spread, 1.8%, over the thought experiment. Unrealistic, but a 6.2 Year bond with a 1.8% coupon, at 0% yield, prices at $111.65. Coupons are only reinvested once a year instead of semi-annually. Below is a chart of a "7.2 Year, 3.7% Coupon Bond" at various yields:

(Click to enlarge)

What is the return in 10 years if the Fed pinned the "bond" to zero (remember there is no redemption of a bond at par when in a bond fund)? Just 1.49%. Year 1 would start off great with a total return of 16.32%, but then the "bond" resets to a 7.2 Year 1.8% Coupon bond. The price of this bond, at 0% yield, would be $112.97. Over 10 years, with yields fixed at 0%, and the "bond" having a 1.8% coupon (the original coupon spread), this 7.2 Year "bond" would provide a total return of $129.86 vs. an initial investment of $112.01.

Below is a table I made of this "bond" for changes in yield and holding period. For instance, if rates moved to a yield of 1%, and stayed there until the end of the 10 year time horizon, this would mean a return of 9.20% in the first year but falling to an average, annual total return of 2.17% in Year 10. If rates rose to 3.75% then the first year return would be -7.75%; and if they stayed there the next 10 years then the average, annual total return would be 3.87%; as the 1.8% coupon spread makes up for the initial rise in rates over time. (In this model if yields rose to 3.75% the coupon received would be 3.75% + 1.8% = 5.55%)

(Click to enlarge)

The table, above, shows the average total return for the various movements in yield. I listed this for holding periods, years 1-10. I then showed the average for these 10 years.

Where is the problem here? It is with the assumed average annual returns over a 10 year holding period. Whether 1.49% or 4.57% this is much lower than 5.07%. This of course affects the returns which the other assets in the allocation must achieve.

Inflation protected bonds (VIPSX or VTIP ETF) achieved a 10 year average annual return of 6.39%, which is 1.32% above the nominal bond return. However, I rounded this up to 1.50% in my asset allocation return table.

One can see that, in the next 10 years, if nominal bonds return 3.07% and inflation protected bonds 4.57%, then the remaining 35% of the portfolio would have to return 14.2%. That is some heavy lifting for asset classes.

What if somehow the bond returns were better? Say 3.75% nominal and 5.75% inflation protected. Then one gets these results:

This is an improvement, but still a return over 12.5% is needed.

What has VTSMX, equities, returned the last 10 years ending Dec. 31, 2012? 7.83% before taxes ( and 8.45% since 1996)

What has PEBIX, emerging market bonds, returned the last 10 years? 11.8% (11.67% since 1997)

Gold has returned 17.61% (without fees), however it has the smallest allocation of 4.84%.

VWEHX a "high" yield (i.e. junk) bond fund returned 8.09% the last 10 years, (Okay, maybe it is not so "junky") and 8.94% since 1978.

If one uses the highest possible returns stated, that is, 8.45% for VTSMX instead of 7.83% and 3.75% for the nominal "bond" fund, then one would expect a portfolio that returns almost -.5% below the bench mark of 7.3%.

Now what happens if we go with the lower returns and gold cut in half? Which, is not an outrageous assumption after a 10 year bull run and the possibility of diversification into other commodities whose returns are not so high.

Now, the portfolio returns -1.59% below the benchmark over a 10 year time horizon.

Another danger in the model is the equity risk premium. Consider the previous risk premiums received over the last 10 years:

What happens to the portfolio return if these same risk premiums occur?

Now the portfolio is -1.84% less than the historical target of 7.3%. Translated into dollars, the shortfall under the 7.3% target means $321,327 less, for a starting portfolio of $1,000,000, at the end of 10 years. Ouch!

The above shows the difficulty investors, and asset managers will have, in the coming years, with asset allocation, because of low yields/high bond prices. The last 30 years of declining yields gave investment portfolios a tail wind; however, now there is not much more bond price appreciation from falling yields to boost 10 year average annual returns. Bond investors will just be clipping coupons and that does not sound like a winning strategy for another investing championship.

(Click to enlarge)


Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article. I have no positions in any stocks mentioned, but may initiate a Long or Short position in GLD over the next 72 hours.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here