How To Build A Risk Budget

Director of Texas' state retirement plan explains how he uses ETFs in his fixed-income portfolio.

LaraCrigger_200x200.png
|
Reviewed by: Lara Crigger
,
Edited by: Lara Crigger

Leighton ShantzFor institutional and individual investors alike, it isn't enough anymore to allocate your entire fixed-income portfolio to the AGG (Bloomberg Barclays U.S. Aggregate Bond Index) and call it a day. So says Leighton Shantz, director of fixed income for the $29 billion Employees Retirement System of Texas (ERS).

When Shantz came on board in 2012, he transformed the plan's fixed-income allocation, shifting away from its traditional core-plus mandate and into discrete "Rates" and "Credit" portfolios. He'd implemented a similar successful move at his previous position as managing director for the Lockheed Martin Investment Management Co. The shift in strategy, which also required the use of ETFs, helped ERS win ETF.com's Best Institutional ETF User award in 2015.  

Shantz is a featured speaker at the upcoming Inside Fixed Income & Dividend Investing Summit 2018, where he will be discussing the institutional approach to fixed-income investing, among other topics. We caught up with Shantz in advance of the event to get some of the highlights of his intended remarks.

ETF.com: Institutional investors often approach their portfolios in terms of risk and risk management. But what does this mean in practical terms? 

Leighton Shantz: Most institutional boards have an asset allocation framework that says, for example, "We're 60% equities, 40% fixed income." From there, [portfolio managers can invest in] almost anything. 

They overlay what's called a risk budget. A risk budget says, "You have to have a maximum tracking error," where tracking error is nothing more than the standard deviation of the excess return. For instance, if you're to be [invested] 15% in credit, then you might have a tracking error limit of 300 basis points (3%), with a target of 200 basis points (2%).

That gives the portfolio manager pretty clear insight as to the magnitude of the overall risk they're supposed to take. A risk budget quantifies pretty clearly for everybody involved, "This is what we expect to happen."

Now, [a risk budget] doesn't always work, because in market sell-offs, volatilities go up. Your risk budget is based on the volatility and correlations of all these underlying assets, but in a sell-off, volatilities spike and correlations drive to 1.

ETF.com: It fails exactly when you need it most. So what do you do to prevent disaster when a crisis hits?

Shantz: Well, there are three separate correlation matrices—three different states for the underlying correlations between different asset classes. There's the normal state, which is maybe 80% of the time, where interest rates are negatively correlated with U.S. stocks. Then there's a deflationary market, where both growth and inflation are falling. And there's the stagflationary correlation matrix, where growth is falling but inflation is rising.

Each has a distinctly different [asset] correlation matrix, but I don't know which environment we're going to have; I have to bet appropriately. So I invest the portfolio to the assumption that 80% of the time it will be a normalized correlation matrix. But I'm not going to go up to my targeted risk limit, because if a volatility spike happens, then I'll be over my limit.

ETF.com: What does this mean in real-world terms?

Shantz: We run high yield credit—we just call it "credit,” but it's all below investment grade. We're told to target a 200 basis point (2%) tracking error. But the way we run the internal portfolio, which is roughly 80% of the money, is that it runs best at about 100 basis points (1%) of tracking error. We don't reach for the really risky stuff, because we don't have the time and manpower to take a deeper dive.

For the 20% of the portfolio that's external, I ballpark about 500 basis points (5%) of tracking error. This is our more satellite strategy; it's higher risk, higher octane stuff. That comes to 20% of the book at a 5% tracking error—another 100 basis points (1%).

Together, those two are about 200 basis points (2%). They're not perfectly correlated, though, which means it doesn't actually show up as 200 basis points, even if I'm right on with those numbers.

ETF.com: What changes did you make to ERS' risk budget when you came on board, and what motivated you to make the changes you did?

Shantz: When I got here, the [chief investment officer] said, "We know it's not working right, but we don't know why." What I told them was: It's because you're doing the wrong thing. You're investing in things that you have no expertise in. Or you have them in your benchmark, but you don't actually have the ability to do them. And that adds tracking error right there.

Every asset class should have a reason. Their reason for having fixed income was to provide liquidity in sell-offs, because we're what's called a "mature" plan, which is a very polite way of saying "shrinking."

In a normalized growth environment, that doesn't really matter. But in a generalized sell-off, when risk assets are mispriced, you should be buying them, not forcing yourself to sell them to make benefit payments. Because then, you're never going to recover, right?

So, instead of doing this core stuff, which is really just rates, where Treasuries drive everything—we split it up: Have a Treasury portfolio and have a high-yield portfolio. The Treasury portfolio can preserve value, and you can sell those to make benefit payments or to rebalance. The high-yield portfolio can generate your needed returns. Those pay down whether you sell them or not; they have maturity dates.

ETF.com: That must have meant selling off hundreds of thousands of individual securities.

Shantz: There was the effort, right? What do you do with all these investment-grade corporates?

Well, we wanted to get rid of them, but in 2013, street liquidity was horrible, and I was terrified of the idea of showing up every day with a bid wanted in comp. [A "bid wanted in competition" is where an institutional investor submits a list of bonds to several dealers, who can then make bids. The highest bidder wins the contract.]

The street will figure that out very quickly. You just don't get good execution. There are those things that no one puts a bid on, so you end up with this portfolio of all these odd ducks.

I had started researching ETFs, for high yield, actually. And I went into it thinking I was going to find a way to make money from the ETFs; that they were fundamentally broken instruments and that there was a way to arbitrage and profit from them.

But I walked away thinking, yes, these are probably pretty useful tools. We approached an ETF sponsor, saying, "Here's our list of bonds. Tell us how we can get those in ETFs, and we'll use the ETFs to sell."

ETF.com: What changed your mind about ETFs?

Shantz: When I got here, there was no internal capability to do high yield, so I had just started the effort of finding people and building up that capability.

In that due diligence process, we met with at least a dozen high-yield managers, looking for ideas. Every single one of them was openly critical of the ETF—which makes sense from their business model; it's a direct competitor hurting their margins.

But they were so hostile to them that I just decided, well, obviously there's money to be made from them; if ETFs are this dumb, let's figure out how we're going to do that.

But through that due diligence, eventually I flipped. I was wrong; they aren't dumb. They're not perfect, but they’re actually pretty well-thought-out and constructed tools, at least in most cases.

Contact Lara Crigger at [email protected] 

Lara Crigger is a former staff writer for etf.com and ETF Report.