Why I respectfully reject the harsh take on unconstrained bond funds expressed by Google and LPL co-authors

This may be a good example of the classic story of the three blind men describing an elephant. Not only are the category definitions a bit loose, but we are at a point in the interest rate and economic cycles where it may be particularly problematic to use the rear view mirror to formulate a current course of action.

We don’t invest in categories. We make informed decisions about investing in specific funds. There is so much variance between these flexible bond funds that the decision should always be one of evaluating its suitability for a particular application.

In general, investors should understand the full range of risk and potential return characteristics of a fund before going in. If a fund’s portfolio or strategy is too opaque or complicated to understand, it may be better to look elsewhere.

Because we are likely near an inflection point in the rate cycle, investors evaluating their PIMCO Total Return positions have more than Bill Gross to thing about. Do they want to maintain that degree of exposure (at PIMCO or elsewhere) or do they want to reduce the level of risk in the “safe” end of their portfolios?

Jim,

I accept your criticism with all the dignity possible of an elephant molester.

Brooke

LOL Brooke. What you’re doing with that bag of peanuts is a private matter.

On a more serious note, I think you do a superb job of gathering useful information and making it accessible. It’s hard to find time to stay current on all that is going on in our interesting corner of the investment world, and your site is both useful and interesting.

As this article illustrates, there are multiple perspectives on most investment topics and that makes for a healthy market.

One’s perspective on unconstrained bond funds hinges on one’s opinion on bond funds and the role of fixed income in a portfolio. I think that bond funds have embedded risk in a rising rate environment. Unconstrained bond fund try to mitigate that risk by giving greater flexibility, including the ability to go short. The problem is that the bond market doesn’t always do what you think it will do when you think it will do it. Bill Gross can testify to that on his call to short Treasuries just before S&P downgraded US debt. Should have worked, but it didn’t. And investors paid the price.

I view bonds as an investor’s safe money. And for those in retirement, we use them to generate a predictable paycheck using individual bonds. Why individual bonds? Because individual bonds are a unique security…a legal obligation by the issuer to pay back the investor his or her principal with interest. The exact amount and timing of those payments are known in advance, so individual bonds are predictable. And each investor can buy bonds to match his or her cash flow needs buy matching up maturities. When interest rates rise, neither the principal nor the interest are affected so long as the investor holds the bond to maturity. Thus the downside is limited to the YTM on the bond the day it is purchased (assuming high quality bonds like government agency bonds). And YTM is positive.

Bond funds, on the other hand, are simply mutual funds that happen to invest in bonds. A fund does not have a legal obligation to pay principal and interest. The turnover within a fund means that many bonds are not held to maturity and thus losses caused by rising rates are realized within the NAV. An investor needing to take systematic withdrawals can lose principal when rates rise, unlike individual bonds set to mature when the investor needs cash.

Unconstrained bond funds seek to give more flexibility, especially around duration, so that if rates rise, the fund won’t be a sitting duck. Unconstrained managers are given greater freedom to take bets, but the timing often doesn’t pay off. For investors taking systematic withdrawals, any missed timing creates sequence risk (think worst case reverse dollar cost averaging). Returns within an unconstrained fund come more from the managers bets paying off than the natural, stable nature of the underlying bonds. Figure 2 above highlights the range of returns and ST DEV for a given asset class. That kind of dispersion makes taking withdrawals more like a roulette game than a paycheck. Although lower relative volatility and low correlation might reduce overall portfolio volatility, volatility to the downside for anyone taking withdrawals means that retirees will be drawing out of a portfolio that has less money and therefore less money to recover.

I do not see fixed income as the place to take risk in a portfolio and thus don’t see a role for unconstrained bond funds. I see the role of bonds, specifically individual bonds, not bond funds, as the portion of a portfolio that should stand strong when the rest of the portfolio is under pressure. Consider 2008. When your equity portfolio is seeing huge declines you don’t want your bond fund also declining because it is overweight lower investment grade corporate bonds or has negative duration in anticipation of rising rates. Neither of those bets would have payed off, but an agency bond maturing at the end of ’08 matured on schedule so the investor wasn’t forced to sell into a declining market that was losing liquidity.

You don’t hold bonds for the markets we have seen since ’09. You hold them for markets like ’08, 2000-02, 1973-74. And especially for markets like 1969 when both stocks and bonds were down. Unconstrained bond funds just add another risk factor to bond funds, which are already risky enough in a sustained rising rate environment. I would gladly give up the chance at alpha (which any SPIVA scorecard will show is challenging enough) for the predictability and certainty of individual bonds when markets are falling apart (or the manager guessed wrong). I’ll take my risk on the equity side where investors are better compensated for taking on uncertainty.

Very well put. Over 90% of our business is DC plans, where individual bonds are not available, so we are particularly sensitive to bond fund selection for our plan menus, taking into consideration the plan demographics and our capital market expectations. Bond fund selection is no less complicated than selecting stock funds, although I’m not sure how many people appreciate that.

Jim,

Thank you for your kind comments about RIABiz and I agree with you about Brent’s comment. In fact I’m trying to convince him to add a few things and we’ll run it as a column.

BTW, I like the elevator pitch on your site and the fact you even call it that…on your site. It has some plainspoken language and an air of humility.

.”.more of your employees will retire with greater security than would otherwise be the case….you will sleep better at night knowing that your retirement plan is in good hands and in good shape.

Brooke

Thanks Brooke. Our success is largely attributable to transparency, clear communication, and commitment to making a difference. Pretty straightforward, yet it seems to differentiate us…

Three comments:

1. Unconstrained bond strategies are just that unconstrained and able to invest where the manager identifies the best risk reward profile coupled with the ability of the manager to execute. Managers will have differing views on risk/reward and different capabilities to execute hence the wide dispersion of performance within the category. Simply put some managers might have better corporate credit resources, some emerging market, currency etc. Utilizing historical data to assess the veracity of such strategies is inadequate at best and perhaps misleading. Really should have a qualitative assessment of the strategy – sitting down with manager(s) to understand what they do, where they go and how they do it to get a real sense.

2. In theory, the comments about buying individual bonds makes sense- in practice it is much harder for average investors or advisors. The bond market is an auction where size is your friend- believing you can get similar execution trading 25,000/50,000/100,000 lots as Blackrock is not gonna happen. Further- you have no idea regarding implementation shortfall. Also- assuming you buy individual credits- taxable or tax exempt- you really do need a robust research group particularly today. Finally- how do you benchmark your activity? Is there a GIPS compliant or audited track record to point to? While it sounds good…as does an unconstrained bond strategy ironically…the devil is in the details.

3. I do agree that assessing bond funds is more difficult than one might expect- particularly anything related to any form of credit. Another reason why you should have a qualitative evaluation rather than rely on quantitative/ statistical only.

FAA, although still behind the equity markets in terms of liquidity, the newer ATS systems like Tradeweb Direct have really compressed the bid/ask spreads. When building out a portfolio of individual bonds, we look to match the spending needs for those in retirement (those in accumulation are a little different). Think paycheck portfolio. We aren’t trying to generate total return. We are matching cash flows using a liability driven (LDI) approach. Thus, the target cash flows are the benchmark. Actually LDI is a special case total return where the downside is limited to the YTM on the bonds, but upside can be realized if the gains worth recognizing when rates fall. The purpose of the portfolio, however, is to manage sequence on both the equity and bond portfolios.

By holding bonds to maturity, interest rate risk can be eliminated and by matching the timing of the maturities to the cash flows, shortfall risk is eliminated. Credit risk is usually not rewarded in an LDI scenario, so we tend to stick with CDs and Agency bonds unless munis are warranted. Munis require a lot more diligence and monitoring, but the tax benefit can be worth the extra effort.

A goals-based/LDI approach also changes the bond allocation conversation from a nebulous percentage to a concrete time horizon, say 8 years. Thus, the portfolio would generate 8 years of predictable income, which provides a time buffer to ride out poor equity market conditions and give the portfolio time to recover.

Brent Burns- so effectively you are dedicating the bond portfolio vs anticipated cash flow needs right? Have some background in this and this approach was a popular strategy for pension liabilities/ insurance contracts (GICS) when the yield curve was inverted and rates were high- obviously the discount function on the liabilities had a lot to do with that. LDI is really a reiteration of what was done in the ’80’s/90’s. So I think I get what you are doing and seems reasonable to me-two points I might add. Dedicated bond portfolios tend to be more expensive than immunized bond portfolios- match duration, skewness and minimize month squared deviation. Not to say your approach doesn’t work just an observation. Secondly- you still have to trade it in odd lots I would imagine. That’s the tough part- but good for you folks if you can work through that.