Risk Parity + Trend Following = ?

Bridgewater & Associates is commonly known as the world’s largest hedge fund, with more than $140 Billion in assets under management and close to 1,500 employees. Its founder Ray Dalio is about as eccentric as you would expect from the founder of the world’s largest hedge fund – penning his own manifesto, and installing a work culture where employees watch recordings of each other’s meetings in an effort to have “radical transparency”.

But behind all these billions and quirky management techniques is actually a pretty simple investment concept known as risk parity. A strategy a 2015 Financial Times article estimated has upwards of $400 Billion of assets using it. So what is Risk Parity?

The basic principal is that you should risk the same amount in different asset classes to get true diversification, and adjust your portfolio weighting to reflect current “risk”, as measured in the volatility of each asset class.  The main concept is constructing a diversified portfolio which buys and holds multiple assets based on risk alone, not on expected returns, AND reweights that portfolio consistently.

The practical example of this is that if stocks are, say, twice as volatile as bonds – you should have 2 times more bonds in your portfolio than stocks. With just those two in a risk parity portfolio, you would have 66% in bonds, and 33% in stocks, almost a 180 degree flip from the standard 60/40 stock bond portfolio.  Add in some hedge fund leverage to get the bond returns up in the range you want them to be in (7, 10, 15% or what have you), and that’s risk parity in a nutshell.

So, what’s the catch?
We sat down with Kevin Doyle, who cut his teeth with two of the best in the business in terms of derivatives and systematic approaches – Monroe Trout and Toby Crabel. Now at Emil Van Essen and running a program he refers to as ‘Tactical Parity’, Kevin had some good insight into the issues with the classic risk parity approach:

Well, risk parity can be good for investors with a long time horizon who are looking for diversified long asset exposure in a balanced fashion. The strategy has performed very well in-itself but, like any, does have its weaknesses.  By its very nature a risk parity program won’t go short any asset classes, but stays long, albeit in a risk adjusted manner, through thick and thin. This can lead to some issues when there is a general melt-down across all or a majority of the asset classes at the same time.

Like we saw in the latter half of 2018, as witnessed by robo-advisor Wealthfront, who had a bit of an ill-timed launch of their new risk parity mutual fund. Here’s how Wealthfront explained it to investors:

Since the Fund’s launch in January 29, 2018, all but one of the asset classes represented in the Risk Parity portfolio…are down… Portfolios constructed from these asset classes have similarly experienced declines, regardless of the asset allocation methodology used to construct them

To summarize, if everything in the portfolio is going down – no amount of Risk Parity or videotaping meetings is going to make the portfolio be up.

Trend Following Shorts
The obvious solution to the problem of all assets being down in the portfolio is to allow the portfolio to go short those asset classes, enabling the portfolio to make money on assets that are down and ones that are up. Indeed, that’s the whole concept of trend following and global macro approaches which go both long and short global asset classes. [as an aside, they do that in a systematic way which allocates the same amount of risk on each trade, so sort of a risk parity with shorts approach].

But there’s a few issues with the short trades in a trend following approach, according to Doyle:

Trend following short positions may look good for short-term insurance, but our research has shown they tend to lose money over the long term holding period. Shorts in financials fight the natural futures carry (and the further they fall the more that carry increases). This means you are fighting a headwind so long as you own the short. What’s more, short positions tend to be entered in higher volatility regimes, which can lower the return value even more due to the position being sized based on that current volatility (and thus smaller) as the market moves to a lower point.

Here’s an example of research showing how short positions in the S&P and US 10 Year Note futures have performed as opposed to long positions.

It shows a portfolio of equally sized dollar risk positions in both the E-Mini S&P and US 10 Year Note futures which are volatility adjusted daily. The portfolio is long the individual contact position when it closes above the 200 day moving average or short the individual contract position when it closes below the 200 day moving average.

 

Best of Both Worlds
So, if risk parity is pretty good, except for missing shorts, and trend following shorts are really only good when there is a crisis, and not that great over the long term; what about combining the two in some way?

Enter the Global Tactical Allocation Program (GTAP) from Kevin Doyle at Emil Van Essen, in which he’s modeled a program which does just that, using trend following components as an overlay on a risk parity core in an attempt to grab the bulk of the benefits of both approaches without the well-known drawbacks.

GTAP is described as:

The Global Tactical Allocation Program (GTAP) is a long-asset, systematic managed futures program. It utilizes a dynamic risk parity approach to a core model portfolio together with a multi-strategy tactical overlay to control individual asset positioning and overall portfolio risk exposure. Our term for this is Tactical Parity.

Here’s Doyle explaining:

It acts in the same way that a risk parity program does in the sense that the long exposure is balanced across five different asset classes, and the exposure to the individual assets, when they are active, is balanced through volatility. We also take other specific nuances of the individual assets into account when looking to control this overall portfolio risk. For the tactical overlay, we’ve found success using multiple trend-following like models. These models layer in and out of positions taking the portfolio from full exposure (as defined by the risk parity core) to flat when longer-term expectations are negative (as defined by the model overlay). This tactical reduction in portfolio exposure is the major difference between GTAP and risk-parity programs with the goal of keeping it out of trouble when there is a major asset melt-down.

Basically, remove the drawbacks of risk parity by going flat different assets when they are showing a down trending profile, but stop short (pun intended) of going short those asset classes so as to avoid the negative carry – thus removing the trend following shorts drawbacks.  Not too shabby.

In analyzing the fact sheet, the portfolio takes positions in liquid exchange traded futures contracts covering Asia, Europe, and the US, with exposure in stock indices, bonds, interest rates, currencies, energies, and metals. It has a minimum account size of $2 million ($100k for its privately offered fund), and charges low fees of either 1/10 or 0/25.

Here’s the performance:

*Past performance is not indicative of future results*

To download the fact sheet and more analysis, click here.