If you haven’t already been looking at fixed-income ETFS, chances are you will be in the near future. Stocks remain turbulent and we’re seeing all sorts of rhetoric extolling the benefits of diversifying with fixed income. And with ETFs so easy to trade and track, this seems to be turning into the preferred way to get it done. Robo-advisers are taking that to heart: Unless your answers to their brief electronic questionnaires portray you as a bold gunslinger, you’re likely to have substantial exposure to fixed-income ETFs. But this may be an area in which ease kills.

Why Fixed Income

This is easy to understand. Fixed income entails lower risk than stocks and is, therefore, less volatile. So it’s served investors well for many years as a way to diversify portfolios.

Why Fixed-Income ETFs

They’re easy to trade and track because they feel and act in this regard just like stocks. With ETFs, we don’t have to hunt for bonds nor do we have to deal with the un-stock like mechanics of contributing to or redeeming interests in open-end mutual funds. Indeed, fixed-income ETFs have becomes staples among the robo-advisors, which pride themselves in convenience.

Why You Can Get Burned in Fixed Income

Do I really have to go into detail on how low interest rates are, and how (aside from trivial downward squiggles now and then) we’re looking at stagnation as a best-case scenario or more likely, eventual rate gains, which would hurt fixed income. Bank of England Chief Economist Andy Haldane puts this into perspective by pointing out that rates now are lower even than they were at any time since 3000 B.C.E. That means all those graphs and charts showing you how wonderful fixed-income diversification are now of value primarily for what they can fetch on EBay as collectables, right along side rotary-dial telephones.

Why You Can Get Especially Burned in Fixed Income ETFs

Consider the iShares 20+ Year Treasury Bond ETF (TLT). Its weighted average maturity is 26.7 years and its weighted average duration is 18.6 years. If you don’t know bond math, consider it the fixed income equivalent of equity beta above 2.00. In other words, it’s not merely volatile. It’s VOLATILE!

That alone should scare you given the prospect of rising interests. Saying they aren’t rising today doesn’t really help much. Even if rates start rising two years from now (most seem to think it won’t take nearly that long), the holder of TLT would still be akin to someone guaranteed to be burned at the stake two years from now. Yeah it doesn’t hurt today. But is the “present value” of assured agony two years hence really such a good thing? Why not simply avoid the fate altogether, even if escape from the situation means forfeiting some luxurious fine meals between now and then.

Yet TLT is a very popular fund. It has almost $5.5 billion in assets. Perhaps it’s because we’re dealing with bonds. We know what our returns will be (assuming Congress doesn’t blow away the status of Treasuries credit-risk-free asset with credit risk). If we’re willing to hold to maturity and take opportunity losses (i.e. bypass an opportunity to get better returns later on by investing in different securities), we’ll still, at the end of the day, lock in a certain return (the current yield on TLT is about 2.53%). That might be better than what we can get from stocks if we have a depression or something like that.

But who said anything about bonds, notes, etc.? TLT is an ETF! That means it doesn’t mature – not ever unless its sponsor chooses to terminate it. So the basic bond math is out the window.

Let’s pretend TLT holds only 20-year Treasuries (the shortest-term least volatile instruments it can own). What do you think it will own a year from now? Will it then have Treasuries maturing in 19 years? No. It will have had to adjust its portfolio such that it will again own bonds with a 20-year overall maturity. Ten years from now, it will still be holding securities that mature in 20 years (albeit not the same ones it has today; it’ll have to do a lot of rolling over to keep that constant 20-year average maturity).

How Bad Can This Turn Out?

Very bad, and much worse than would be the case if you simply held bonds that mature in 20 years.

In Table 1, I’ll track an individual hypothetical 20-year treasury security issued today with a $1,000 face value and a coupon of 2.6%, which, I’ll assume, is the market rate. I’ll also track a $1,000 stake in TLT-like ETF assuming its starting portfolio also consists 20-year bonds with a starting yield of 2.6%. For the ETF, I assume a zero expense ratio and that at the end of each year, the portfolio is reconstituted to consist of new 20-year bonds with coupons and yields that match whatever the market rate at that time is. In both cases, I assume interest rates rise by 0.25% per year. I also assuming in both cases, the investor does not reinvest the income.

Table 1 tracks the progress of the bond and the ETF using basic present-value-based bond-price formulae. These are not as precise as possible given that I’m not factoring “convexity” into the interim paths, nor am I making assumptions about the pace at which rates over the course of each year. The big-picture conclusion, however, is not changed by addition of convexity (mainly of issue for the interim prices over the 20-year path of the bond) or by the tweaking of other assumptions.

Table 1: Starting out with $1,000 principal and 2.60% yield

There you have it. After 20 years, the ETF shareholder is staring at a 47% loss while the owner of a bond suffers no capital loss at all. Because fixed income securities have maturity dates, no matter what happens in the secondary market as a result of interest-rate increases, the bond must, at some point, start recovering losses and rising toward the $1,000 maturity value.

In this example, the turnaround for the bond began after Year 12. Before that, when the bond had suffered its maximum loss, 19.4%, the ETF investment was already down 31.8%. The main difference is that with the ETF, I was continuing to assume a 20-year maturity, exactly as I had on day one. With the bond, however, the passage of time had turned a once long-term 20-year security into a now-intermediate nine-year issue. Eventually, it would turn into a short-term issue and then mature.

Because fixed-income ETFs typically do not have maturity dates, there’s nothing that pulls them toward a fixed target such as a par value. Fixed-income ETFs can never recoup losses unless or until interest-rates change direction. This is very different from what we know with stocks, which can recoup even massive losses quickly following some positive announcements, such as increased earnings guidance coming after a few disappointing quarters.

To drive home the point, I illustrated a 20-year security, a volatile issue. Table 2 illustrates the phenomenon with a shorter time frame. For the direct security, I assumed a five-year maturity. I based the hypothetical ETF on the iShares 3-7 Year bond ETF (IEI), and assumed its portfolio would always hold the shortest-term securities it’s allowed to own, those with three-year maturities.

Table 2

In this case, the ETF sustained a very modest 1.8% loss. Compared to what we experience with stocks, that seems easily bearable. But again, compare it to holding the security – we didn’t have to sustain any loss.

Bear in mind, too, that I compared an assumed 5-year security with a less-volatile ETF portfolio that presumed a portfolio of instruments maturing in three years. I also retained my zero-ETF-expense-ratio assumption and cut the annual increase in interest rates down to 0.125%. While this hypothetical was framed such as to look benign, the reality is that the future is uncertain: Deliberately choosing a security that is indisputably programmed to perform worse under the adverse scenario we fear cannot be considered suitable risk management.

So How Can We Invest in Fixed Income

Given the way commentators, brokerage firms and robo-advisers pitch fixed-income ETFs, it’s easy to fall into the trap of assuming these are no-brainers risk-management vehicles. But that is definitely not the case.

Ironically, there are some less-known ETFs from iShares and Guggenheim that seem to solve this problem, target maturity funds. So far, though, investors have shown little interest. Maybe they should start looking. I’ll discuss those, and other potential approaches to robo-risk management in my next post. For now, though, it’s important to recognize that fixed-income ETFs came of age during the course of an epoch fixed-income bull market. The products as designed have been fine as long as those circumstances persisted. But they do not appear to have been designed to effectively withstand the possibility that the core assumptions (continuing gains in bond prices) may change. But now, it is changing.