When I reviewed Bernstein's Rational Expectations book a few months ago I promised readers a post discussing fixed income (bond) options for your portfolio. Specifically, there were three recommendations Bernstein made that are at least somewhat controversial that I'll discuss here.

  1. Buy minimal risk bonds
  2. Buy individual bonds
  3. If you must use a fund, choose an open-ended one instead of an exchange traded fund (ETF)

I'm going to discuss each of these, including pros and cons in turn, then you can decide what you want to do with your bond portfolio.

Keeping Bonds Short-Term and High-Quality

Bernstein, like Larry Swedroe, has been an advocate of taking minimal risk with the bond portion of your portfolio for a long time. The theory is to take your risk on the stock side, where it can be taken more efficiently. Here's what Bernstein says:

Critically, your riskless assets should retain their value in a crisis. Corporate bonds, in particular, have a modest amount of stock-like behavior and can see price falls independent of the rise or fall in overall interest rates. Municipal bonds can also behave this way, and if you're depending on either corporates or munis for liquidity during a crisis–precisely when you're likely to want or need it the most–you may have to take a substantial haircut to realize it. Tax-sheltered investors should keep almost all of their fixed-income assets in government guaranteed securities: Treasury bills, and notes and CDs. And while taxable investors should own at least some municipal bonds, they should still hold a fair dollop of Treasuries and CDs as well….At present, very low rates might make you consider using….Short term [bonds rather than intermediate bonds.]

The alternative argument is to look at long term returns. Both longer-term bonds and riskier bonds (corporates, junk, even P2P loans) have much higher long term returns. Taking a look at the Vanguard site, we see the following 10 year annualized returns for various asset classes:

  • Short term treasuries: 2.70%
  • Intermediate term treasuries: 4.51%
  • Long term treasuries: 5.93%
  • Intermediate term munis: 4.06% (federal tax-free)
  • Intermediate term corporates: 5.20%
  • Intermediate term junk: 6.57%

As can be clearly seen, the more term and credit risk you take, the better your long-term returns. So how do you reconcile a recommendation that seems to suggest you accept lower returns? Well, there are two good reasons to accept what seem to be lower returns. The first is if you actually might need to sell some of this stuff in a crisis to live on. You only get those higher returns if you DON'T sell. Secondly, Bernstein would argue that you must look at portfolio performance as a whole, rather than as individual parts. In a crisis, a liquid, high-quality bond is very helpful to the overall portfolio. Not only does it give you an asset that is liquid in case you need to live on it, but it is also an asset that has likely risen in value. In 2008, even Vanguard's relatively high quality intermediate corporate bonds lost 6.1%. But the intermediate treasuries were up 13.5%. Since a crisis is the time to sell bonds and buy stocks, you're a lot better off selling bonds that have appreciated, rather than bonds that have depreciated less than your stocks in order to buy more stocks.  This effect may overcome the effect of the higher long-term returns of lower quality bonds on your portfolio, resulting in higher overall portfolio returns, which is what really matters.

The recommendation for short term bonds is simply because they are far less volatile and retain their value much better in an environment of rising interest rates. That wasn't the case in 2008, but would have been in 1973-74 when both stocks and bonds underperformed cash.

So who can take on additional bond risk in order to get those higher long-term returns? Generally, a young, accumulating investor with a secure job. This is because this investor doesn't have to sell his bonds in order to live and his savings to portfolio ratio is still quite high, so he can simply rebalance the portfolio with new money. Personally, I do a little of both. I hold a very safe bond fund (the TSP G fund) which essentially gives intermediate treasury yields while taking T-Bill risk. I also hold TIPS, which are quite safe from both a credit perspective as well as an inflation perspective. However, I also hold Peer to Peer Loans, which are the ultimate in high credit risk since they are essentially unsecured loans to the financially inept, primarily for their high returns (12% a year over the last 3 years for me personally.)

Individual Bonds Are Better Than Bond Funds

Bernstein is also a huge fan of individual bonds rather than bond funds, but this recommendation grows out of the prior one- to basically only hold treasuries. He readily admits that if you do choose to hold corporates or munis that you DO use a fund to achieve adequate diversification.

Unless your account size is tiny, it makes no sense to own a Treasury or government-bond fund, since you can buy these securities at auction and hold them at no cost. The same goes double for TIPS…you can, with a little effort, tailor a ladder…If on the other hand, you hold a TIPS fund, not only do you pay unnecessary fund fees, but there will likely be periods when you will be forced to sell these securities at disadvantageous prices, as happened to many investors in 2008-2009. Much as I love Vanguard's low fees, there's almost no reason to pay them even a penny for their TIPS and government-bond funds….The only bond funds you should own are open-end municipal and corporate bond funds (to the extend that you do own these two asset classes.) The reason for this is simple. Without a great deal of effort and expense, the individual cannot put together a well-diversified low-expense mix of municipal or corporate securities.

The counter argument, of course, is that it is really convenient to use a bond fund, and they don't cost that much anyway. For example, the Vanguard TIPS fund charges just 20 basis points (10 if you have at least $50K in the fund.) That's hardly a huge drag on returns. If you're putting $20K into bonds every year for 30 years and you earn 2.8% instead of 3%, you only end up with $33K less (about 3.3% of the total.) Not exactly life changing. If we're talking about something like $100K, 20 basis points is only $200 a year. That might just be worth it to avoid the hassles of having to keep track of treasury auctions and dealing with brokerage accounts, not to mention avoiding the cash drag issue (although to be fair, the fund also has cash drag, but probably less than you.)

Many investors consider buying treasuries directly through the appropriately named government website Treasury Direct, only to discover that they can only do it in a taxable account. Since treasuries, and especially TIPS with their phantom income issues, are generally held in tax-protected accounts, especially by those in a high tax bracket, this turns the investor off from holding TIPS individually. However, you can purchase TIPS and other treasuries at auction in your IRAs (and possibly even 401(k)s) at any of the main brokerages, often for no fee at all.

The bottom line? Once you have a decent sum of money (which I'd define as between $100-500K in treasuries,) and assuming you're willing to deal with a little bit of hassle, buy individual treasuries instead of using a treasury bond fund. But stick with a fund for other types of bonds.

Don't Use Bond ETFs

Perhaps the most interesting of Bernstein's recommendations was that if you use a bond fund you should choose an open-ended version, rather than a bond ETF (exchange traded fund.) Here is his explanation:

I highly recommend that you avoid all ETF bond funds. To understand why, I'll need to explain some of the trading mechanics involved. An ETF, unlike an open-end fund, trades throughout the day at a discount or premium relative to the net asset value of the underlying shares. In most cases, the spread between the two is minimal because shares can be created and liquidated by independent agents: “authorized participants” who buy up the securities underlying the funds and bundle them into ETF shares that are then delivered to the fund company….This mechanism works well with stocks, which are highly liquid, but not with bonds, which are not. There is, for example, only one commonly trade class of Ford Motor Company stock. By contrast, Ford has a range of bonds of varying issue dates, coupons, and maturities. Since there are so many more individual bonds than stocks, the bonds can be highly illiquid. During a financial disturbance, when liquidity becomes even thinner and most corporate bonds trade only “by appointment” the authorized participant mechanism fails, often at considerable disadvantage to the shareholder. The open-end fund holder, who can always buy and sell at [the end of day price], has no such problem.

I confess I had never considered this issue with bond funds before, and I find his argument compelling. However, this issue is far less important with a treasury bond ETF, since those securities are far more liquid than corporate or municipal bonds. It would have to be a pretty terrible crisis to run into a liquidity issue with the treasuries that make up treasury bond ETFs.

Making Changes

So what am I going to do with all this information? Very little, and what I am going to do, I'm going to do slowly. I intend to continue to hold the TSP G fund as a major part of my bond holdings. This is a major reason I didn't roll my TSP money elsewhere when I left the military. While it is a bond fund, the expenses are incredibly low at 2 basis points a year, and these particular treasury instruments cannot be bought individually anyway. I also intend to continue to hold TIPS as a major part of my bond holdings. However, for the last couple of years I've been holding these as the Schwab Bond ETF in my 401(k). I'm getting close to the point where I think it makes sense to buy the individual TIPS, so will probably gradually convert my ETF shares over to individual TIPS. I plan to continue to hold P2P Loans as a small part of my fixed income portfolio. At this point, I expect higher returns going forward from those than I do from my stocks, and my portfolio is still small enough that I could really use the growth. My lifetime capital to investment capital ratio is still quite high, so I can afford to take that type of risk through a crisis.

What do you think? Are you going to change your fixed income strategy based on this information? Why or why not? Do you use bond funds or individual bonds? Do you use open ended up funds or ETFs? Do you take significant risk with bonds, or keep them short and high-quality? Comment below!