Sunday, January 29, 2023

Selectively Adding Complexity Into A Diversified Portfolio

From Twitter on Sunday;


I've been referring to bonds with duration as being sources of unreliable volatility. A common reason to have fixed income exposure, often 40% of a portfolio, is to help manage equity market volatility. Longer bonds were down just as much as the S&P 500 in 2022, the longest of bonds were down more. This year, they appear to be up the same. One month is obviously a very short period of time but for now, investors are getting gains from holding duration but not diversification. 

For many months now, we've been looking at ways to diversify away equity market volatility using tools that differ from intermediate or longer term bonds. I've been comfortable with short term maturities in most fixed income sectors, alternatives that look like horizontal lines that tilt upward and alternatives that have the potential to go up a lot when stocks drop a lot. 

Part of our conversation has looked at return stacking with this paper from ReSolve Asset Management spurring several blog posts. The ReSolve guys put out a new paper (maybe it's an update to the other one) revisiting some of their ideas which should be relevant and worth learning about for anyone concerned about the diversification potential from longer dated fixed income like I am.

The simple definition of return stacking is using some amount of leverage such that you potentially increase the annualized rate of return a little bit while reducing the risk a little bit. The following might be the most benign possible application of the concept.

 

A 67% allocation to NTSX should equal 100% to VBAIX leaving 33% leftover in Portfolio 2 to put into something else. That something else could be very conservative or aggressive or a mix of things. Putting the 33% into a cash proxy should mean you're just adding a few basis points of yield.

Another form of return stacking that I don't think the ReSolve guys have ever talked about but we have is to nudge up the equities exposure and offsetting that with something(s) likely to do a better job offsetting equity volatility than bonds like certain long/short strategies maybe.  

 

Portfolio 2 allows for much more equity exposure compared to VBAIX leading to a much higher CAGR but with a somewhat lower standard deviation. It's a form of leverage but not in the traditional sense. The risk of this would be that in some random downturn, client and personal holding BTAL doesn't "work." It has worked more often than not but nothing is infallible and it might not work in some future market event. 

The ReSolve guys usually talk about creating the effect of the SPY/BTAL example with far more moving parts and leveraging up with funds that are capitally efficient. Look at the first link from them and you'll see the following;

 

It's a mix of funds, obviously doing different multiple strategies with a total notional (leveraged up) exposure of 161%. If you read the second paper they talk a couple of times about "professionally managed leverage." I wouldn't get too excited about that. Long Term Capital used professionally managed leverage and it blew up. The various multiple strategies by virtue of different correlation attributes should blend together to avoid disaster. Should. Yes, the odds of a disaster are low but not zero. 

This is not something to get carried away with. I do believe it is something to learn about and can influence portfolio construction. I think some of the ideas here have been in portfolios I managed since long before the terms capital efficiency and return stacking started to become popular.

I use one fund that comes up in all the conversations about return stacking but return stackng is a strategy like any others, some people will end up using it in a way that is not effective. Where I have come to describe what I do as investing in simplicity hedged with a little complexity, the heavy return stacking goes much further down the complexity road than I think is prudent. 

 

The chart compares the Rational ReSolve Adaptive Asset Allocation Fund (RDMIX) compared to VBAIX. RDMIX had a 20 month, lights out run including a run of going up when stocks and bonds went down a lot that garnered it a 5 Star rating for a time. You can see in the table above it allocates to five of the six asset classes/strategies listed. The research they do is fascinating to read but it is not clear that the execution of their ideas offers a compelling, all weather approach. 

I think the better approach is to build it yourself, use simplicity (plain vanilla) and then use a couple of complex-ish single strategies to hedge your simplicity. Over the long term, it will be the simplicity, the plain vanilla equity exposure that goes up the most. The context is how to offset some of the volatility of our equity exposure if bonds are less effective than they used to be. 

BTAL and BLNDX are long-time personal and client holdings. I have no interest or intention of using any of the other funds mentioned in this post personally or for clients.

Opening Day!

And the Sweet Sixteen! What a great day for sports fans. The Padres game started when the stock market closed and finished in time to start ...