We use bank loans in many of our income-focused and total portfolio strategies. When we search for high income asset classes, we look for two characteristics.
First, we want high risk-adjusted income. Our income products seek to provide increasing annual income. Growing cash flow requires both searching for attractive yield opportunities and prioritizing risk management. With these goals in mind, we cannot consider yield in a bubble without adjusting for risk.
Second, we want exposures that diversify the other asset classes in our universe. When we talk about diversification, we don't mean just low correlations. After all, correlations between asset classes are notoriously difficult to estimate. We care more about finding asset classes with diversifying structural features. Diversifying structural features give us confidence that the asset class may perform well in a set of economic environments where other assets classes in the universe may struggle.
Bank loans have two such features. First, while they are often rated below investment grade, they typically have seniority and security. Seniority means that bank loans sit at or near the top of the company's capital structure. Seniority puts bank loan holders at the front of the line to collect payment in the event of a bankruptcy. Security means that the loans are collateralized by specific company assets. Security also improves the prospects of bank loan holders during bankruptcy.
Second, bank loans have floating interest rates. Oftentimes, the interest paid will be equal to LIBOR (London Interbank Offered Rate) plus a spread. Because of this feature, bank loans have almost no interest rate sensitivity. As a result, we can expect bank loans to perform well when rates rise during times with positive credit conditions.
We run across few people that disagree with there being a place for bank loans in most portfolios. A more controversial proposition is accessing bank loans through an ETF. This article from Barron's does a good job of summarizing the debate. We also wrote about the issue previously on our blog.
Critics of bank loan ETFs tend to attack on one (or both) of the following two paths. Some - like Blackrock's Larry Fink and DoubleLine's Jeff Gundlach - argue that pairing a relatively illiquid asset class with the "fast money" of an ETF can create problems if a majority of investors all try to liquidate the ETF position at the same time.
Others argue that differing liquidity profiles between the ETF and the underlying loans can create tracking error between the ETF and the underlying index.
We disagree strongly with both of these arguments. And we are not alone. PowerShares, State Street, and Eaton Vance have also pointed out problems with both arguments.
The first argument at least makes sense in theory. However, the same argument can be applied to bank loan mutual funds, which have survived many prior crises just fine. In addition, bank loan ETFs have a number of tools to handle sudden redemptions. The Barron's article linked above gets into more detail on this rebuttal.
We believe the tracking error argument is completely nonsensical. It wrongly assumes that people buy an ETF to access an index. People buy ETFs to access the asset class or strategy. The index is simply a representation of the performance of that asset class or strategy. Penalizing tracking error assumes that index performance is somehow the true performance and that anything different is bad. And indeed sometimes tracking error is bad. For example, I would be very concerned if an ETF tracking the S&P 500 had a lot of tracking error. But what if the index is stale and the ETF provides a better indication of current value. We would argue that in this case tracking error is not only not bad, but preferred.
Luckily for us, the last five weeks or so have been a perfect case study for why these arguments against bank loans are wrong.
You might be thinking I am off my rocker. One of the largest bank loan ETFs, the PowerShares Senior Loan Portfolio (ticker: BKLN), is only down 48bps in the last five weeks. Not exactly a stressful time to test our theory.
But I'm not talking about bank loans at all. I'm talking about Greek equities. This is an extreme example of a disconnect between ETF and underlying security liquidity. The Greek equity market was closed until Monday. So there was quite literally no liquidity for Greek equities. The GlobalX Greece ETF (ticker: GREK) traded almost 60 million shares of this same period.
And guess what? Investors in GREK survived just fine. If you use tracking error to pick ETFs, you might be screaming at me right now. GREK had an astronomical 121% of annualized tracking error over the last five weeks!
And you know what, this is exactly the type of tracking error that we want as investors. GREK became the default proxy for the Greek equity markets. It was the only tool that investors in the U.S. could use to trade Greek equities. Had these investors accessed the Greek companies directly, they would have been stuck with no way out.
The tracking error only existed because the index was stale. There were no trades in Greek equities for five weeks. And so there was no way for the index to update.
Lo and behold, when the Greek equity markets opened on Monday, the index and the ETF converged immediately. This confirms that the ETF was operating exactly as expected. The tracking error was due to issues with the index and so in no way should reflect poorly on GREK.
While this case study will in no way close the debate on bank loan ETFs, we believe that it illustrated in real time why the anti-bank loan ETF arguments are flawed.