I’m usually too geeky to hype any investment as being best or worst of anything. Sometimes, though, even I have to get down, get dirty and use the S-word: Sell. I do that today with the iShares 20+ Year Treasury ETF (TLT) and the PIMCO 25 Year Zero-Coupon US Treasury Index ETF (ZROZ) and 12 similar monstrosities.

My Hate List

Here’s my list of ETFs to avoid.

Table 1

Why I Hate These ETFs

I hate them because they all appear in a Portfolio123 screen I created to identify fixed-income ETFs with a long-term focus.

That’s all. This has nothing to do with creditworthiness. It’s long-term I hate, and that applies equally to corporate fixed income and Treasuries. (I took a TIPS bond off the list because that has a different set of dynamics. And by the way, this list isn’t exhaustive: It doesn’t include general-purpose ETFs that have multi-term exposure, but you may want to check on how much exposure they have to long-term and sell if it’s heavy.)

I don’t necessarily expect any of the ETFs in Table 1 to post the worst losses around. In fact, I have complete confidence that the world of equities will serve up many ETFs that wind up delivering far worse returns over the next several years.

My problem with fixed income ETFs is one of suitability. If you hold what one person considers a dog of an equity ETF, at least you can presumably make some sort of argument, to yourself even if to nobody else, that things aren’t really so bad for this group, that these stocks are unappreciated or underappreciated.

Long-term fixed income ETFs are different. The probabilities here are hopelessly tilted against you, due not to analytic opinion but basic math. So even if your goal is diversification, or mitigation of volatility, as it so often is for so many who own these ETFs, you have virtually no chance of getting what you ultimately want.

Starting With the Past

One of the reasons I’ve got steam coming out of my eyeballs regarding this topic is a debate I’m having with a Portfolio123 user in that site’s Community; I’ll call him Jack (not his real name). Jack is a bright guy who has had good success in the past using TLT as a hedge vehicle and he continues to favor it. This is why I’ll focus here on TLT, although the same could be said of all the ETFs in Table 1. ZROZ get special mewtion because 25-year zeros . . . oh my. :-(

Indeed, TLT has served Jack and others magnificently in the past. Figure 1 shows a simple 1/2/99-7/11/16 backtest of an ETF portfolio that holds only one security, TLT.

Figure 1 – TLT

Nice! Considering the volatility we’ve seen in equities, you can easily imagine what a wonderful diversifier this has been.

Those who used the iShares 3-7 Year Treasury ETF (IEI) or the iShares 1-3 Year ETF (SHY) also did incredibly well in mitigating volatility, as shown in Figures 2 and 3.

Figure 2 – IEI

Figure 3 – SHY

But why not go with TLT. It had the highest volatility (standard deviation) of the three but was almost as poorly correlated with the S&P 500 and clearly provided the best average annual return.

It makes sense that TLT was the winner. In a declining-interest-rate environment, long-term bonds are supposed to appreciate better than intermediate-term bonds and intermediate appreciates more than short-term. That’s basic bond math. I didn’t test ZROZ because its history doesn’t extend far enough back to allow a full test. But as a zero coupon bond, we can assume it would have appreciated even more vigorously than TLT. (By the way and for the record, this is really about a bond concept called “duration,” that combines coupon and maturity, rather than maturity per se. But for purposes of today, we need not deal with that beyond saying zero-coupon binds have longer durations and, hence more volatility.)

Figure 4 makes it clear for anyone who’s been distracted for the last 35-or-so years that interest rates have, indeed, moved steadily and aggressively downward.

Figure 4

So everything we see thus far makes perfect sense. TLT and ZROZ were heroes, as were all other long-term fixed-income bonds and ETFs that are, of course, impacted by the same dynamics.

Turn! Turn! Turn!

I used to think of Pete Seeger as a folk singer and songwriter, but perhaps I should start seeing him as having been an especially forward-thinking fixed-income analyst when, inspired by Chapter 3 of Ecclesiastes, he penned the song lyrics “Turn! Turn! Turn!” What else can one say about where interest rates are now; pretty much at zero.

I really didn’t think this was such a complicated notion, but then, neither the author of Ecclesiastes nor Pete Seeger nor I envisioned the nonsense floating around about negative interest rates.

Strictly speaking, yes, they are actually here – but not really. There are cases where banks now pay a fee to make deposits of overnight funds, as opposed to receiving interest. That may spread or even deepen a bit. But that’s not the sort of economy-wide phenomenon we’d need in order to give TLT, ZROZ or any other funds in Table 1 the kind of return characteristics needed to justify continuing to hold.

To understand what it would take to make investments like this work, imagine its 2020 and the yield on the 10-Year Treasury is minus 4%. Since corporates carry credit risk, their yield you be even lower, say minus 7%.

Traders work on the floor of the New York Stock Exchange (NYSE) on March 7, 2016 in New York City. (Photo by Spencer Platt/Getty Images)

What would this world look like? Would bidders at Treasury auctions say how much of a holding fee they are willing to pay for the privilege of getting their hands on Treasuries, which pay nothing until they are redeemed at face vale in the future? What happens to older secondary market securities? Would holding fees high enough to offset pre-existing contractual interest obligations be paid to sellers? What happens when the issue is sold again; which seller gets the holding fee? Or does it all go to the Treasury Will the equity market look like if business can be paid by bondholders to take capital from them? Would they worry about repayment of principal? If rates stay negative, they’d just borrow anew to repay maturing obligations. Why share the equity-holder’s wealth with the public?

What about rates paid by consumers? We can’t expect them to pay 25% on credit card debt any more, or if we do, then for negative rates to get them to spend more, Congress may have to enact a penalty on savings and excess income (good luck defining that) of, say 30%. How else can we motivate them to take on more credit card debt? What about folks building savings for college, retirement, etc.? Are we to impose tax penalties on those who don’t empty these accounts to spend now, whether they want things or not? (That will make battles over rules requiring people to buy health insurance look like a marshmallow roast) Who do you think would sponsor such a Bill in the Senate? Ted Cruz? Bernie Sanders?

This, not fees on overnight instead of payment for deposits is what negative interest rates are really about, at least in the context of giving meaningful positive potential return potential to long-term fixed income. So do yourself a favor and tune out the policy wonks who talk about negative interest rates already being here.

Enough of this jiggery-pokey! Can we just get real and recognize that the prices in the fixed income market (which rise as interest rates fall and vice versa) are butting up against an absolute incontrovertible top, a peak, dare I say, a bubble? And this is not like an equity-market peak, which s never absolute and always debatable. In fixed income, there is no debate. It’s absolute.

Turn? Yes! When? Hmm . . .

Here’s the hard part. When will rates rise? In late 2016? In early 2017? That’s hard to say, given the incredible amount of economic and financial dislocation out there (at least partly caused by Central Bankers who kept slashing rates long after it became apparent to normal people that this sort of thing had lost its former power to influence economic activity).

So it’s possible that rather than chart a “U” or “V” shape, interest rates could, after reaching bottom, as they have, move horizontally along the floor for a heck of a long time. Or not. Honestly, I don’t know.

The Suitability Conundrum

OK. So it would not be correct to say there is a 100% probability of loss in long-term fixed income. Its possible principal/asset value could hold constant for a long time while investors continue to collect interest at current nano-rates. It’s also possible that serious traders can make money catching the squiggles inherent in any open market as it reacts to news events and the like: As Jack pointed out in our forum, TLT is up 8 points since I last bashed it (that apparently being buying in reaction to Brexit).

So now, let’s think about the countless folks who are exposed to long-term fixed income. Are they there for the trading opportunity? I seriously doubt it. My 88-year old mother, who received a proposed plan from Vanguard Personal Advisor Services that recommended a 8% stake in long-term U.S. fixed-income definitely isn’t interested in quick-hit trading opportunities. I suspect none of the clients of Schwab, Wealthfront, Betterment, Wise Banyan etc. who have fixed income exposure either through default (long-term allocations in the all-segment bond funds into which they get shoved, or in dedicated long-term fixed income) are looking for the quick-hit trades.

If you’re a skilled trader and you believe you have a bead on TLT or another long-term fund, have at it. For the rest, I think we have to consider something else.

Risk Reduction

As noted, TLT has been a risk-reducer’s dream for Jack and so many others. But that was then. We can’t go back in time. We have to look forward.

Even in the best-case scenario, the performance seen from TLT in Figure 1 is absolutely positively out of the question. Quick-hit squiggles aside, TLT has no choice but to move sideways, with its 2.24% yield, or fall.

As with all investments, there is a range of probable outcomes, which can be summarized by al illustration of a probability distribution, as per Figure 5.

Figure 5

Many will recognize that this is a nice clean artistically rendered “normal distribution” with bad and good outcomes (red and green respectively) being equally probable. The further away from the middle we go, the better or worse the potential result. The higher up we go vertically, the higher the level of probability we get a particular result. By definition, the highest probability is associated with the neutral outcome in dead center.

We know, of course, that many things, including security returns are not normally distributed, but for purposes of today’s topic, let’s not get picky and just focus on the green and red, and the idea that there are plenty possible outcomes on each side of neutral.

We hate the red portion of the diagram and should be trying to do everything we possibly can to try to set up situations in which the positive probabilities (the green) are larger than the negative (the red). But even if we can’t accomplish that (and in real life, this is easier said than done), we want to at least borrow from the medical community’s Hippocratic oath and “do no harm;” i.e. we need to bend over backwards to make darn sure we don’t do anything crazy that will cause the red to exceed the green.

With that in mind, consider the upside total return for TLT. I’ll say it’s 2.24%, which is the yield (subject to itty bitty variations in both directions that appeal to traders but not most folks). Trading zigs aside, if the return varies from 2.24%, it can only come in get worse as rising rates lead to the reverse of what we saw in Figure 1. Being longer term than IEI or SHY, TLT has to suffer more severely. It’s the same math that caused TLT to outperform on the upside.

What about IEI? The yield here is lower, 1.34%. But if rates rise, IEI will suffer to a lesser degree than TLT, again this being the flip side of what we saw before, for IEI in Table 2. The shorter term SHY is a different picture. It will fall even less in response to rising rates, but it’s here and now yield, 0.61%, is really, really low.

Let’s take SHY off the table. Let’s make this a contest between IEI and TLT.

You have a teeny-weeny upside potential with TLT turning out better than IEI if interest rates stay where they are for many years. That might happen. Or it might not. If, when and as rates rise, TLT will wind up adding downside volatility. Your probability distribution of potential outcomes from TLT in lieu of IEI looks like Figure 5 – if you squint, you’ll see a sliver of green just to the right of neutral.

Figure 5 – Probable Outcomes From Choosing TLT over IEI.

Is that what you want? Really? And by the way, it’s worse with ZROZ. It’s yield, 1.69%, is much closer to that of IEI but its volatility, vulnerability to loss if rates rise, is much more severe.

Perhaps it’s OK for my Portfolio123 friend Jack. He has an engineering background devotes substantial effort to watching the markets and following relevant news and commentary, and perhaps, might be able to squeeze out that extra percentage point or so of TLT yield and switch to IEI or even SHY is he times the trends in interest rates.

But is that you? Is this what you expect when you go into long-term fixed income or are put there by your adviser in order to mitigate the risks of equities? Yes it’s “uncorrelated.” Yes it’s “diversified.” But is it “suitable?”