Optimizing Fixed Income Portfolio

Reader Question

Hello LTR, Thanks again for your blog. I find it very useful and your insightful postings have already motivated me to buy my first I-bonds this year. I'm confident I have many more I-bond purchases in my future. I'd like to hear your thoughts regarding long-term vs medium and short-term bonds in this low interest rate environment. From reading your blog, it seems that you favor longer term bonds (EE bonds, etc.) over shorter term investments but could you share more of your analysis on this topic? My percentage of equity holdings is larger than what is normally advised for someone in their mid-50s. I've weather the stock crashes well and was an enthusiastic purchaser of equities in March 2009. My net worth is large enough that I feel comfortable taking a little more risk in my portfolio. For fixed income, I have chosen (so far) to stay with short and medium-term corporate funds from Vanguard, medium-term municipals (Vanguard), I-bonds, and CDs. I've stayed away from Treasuries and TIPS because I believe they are overvalued. I realize that interest rates may stay low for a very long time as they did after WW2 or they may rise significantly when inflation picks up. I just haven't yet been willing to lock into 20 or 30 year fixed-income investments at these historically low rates. I would be interested in your analysis on fixed-income duration in today's environment. I think a lot of investors are struggling with this nowadays. Thanks!

My Reply

Thank you for the kind words about my blog. I'll be happy to share my thoughts on the state of the bond market, but I doubt you'll find much new insight in them. It sounds like you are already well aware of the (unfortunately meager) potential upside and potential downside of all the different bond choices today -- especially the obviously poor values of nominal treasuries and TIPS compared to many alternatives.

I wouldn't say I favor long-term bonds over medium or short ones today. I'd say I'm neutral on the issue. If my cheerleading for EE-Bonds made it sound like I advocate locking into long-term bonds for all investors, I apologize. I mean to only convey the message that if you are investing in long-term bonds in the first place, then you should make full use of EE-Bonds. I-Bonds are the only bonds that I would recommend to anybody today without a second thought. But a lot of investors do invest in long-term bonds without even realizing it -- I'll get to that point in a bit -- so they could benefit from "upgrading" to EE-Bonds if they don't mind jumping through some hoops.

I definitely favored all sorts of long-term bonds for all investors back in late 2010 and early 2011 when I started this blog and made numerous recommendations to that nature. It was not because I foresaw the drop in interest rates. I recommended those bonds because I saw their yields as at least OK or better, both by historical standards and especially compared to the near-zero return from short-term investments. Unfortunately that is no longer the case. Long-term rates dropped 1-2% since those days (which is quite a major move for long-term rates) and it's hard to classify today's long-term bonds as anything better than mediocre.

As some points of reference, historical average yields are about:

• 3.5% for cash
• 5.0% for intermediate-term (10 years) treasuries
• 5.5% for long-term (20+ years) treasuries
• 6.0% for long-term (20+ years) investment-grade corporates

Compare these with what's available today:

• ~0.5% for cash, depending on where you park it (with a clear bargain in I-Bonds that could be expected to get 1.5-2.0%)
• ~1.5% for intermediate-term treasuries
• ~2.5% for long-term treasuries (with a clear bargain in EE-Bonds that will get 3.5%)
• ~4.0% for long-term  investment-grade corporates

The question becomes whether one should lock into mediocre returns of today's long-term bonds, accepting the risk of substantial (if ultimately temporary) drop in value along the way or whether one should suffer the almost literally zero return from today's cash/short-term bonds while hoping for better yields in the coming years.

History isn't overly helpful in answering this question. Last time interest rates on treasuries were in this neighborhood was in 1940s-1950s (corporate bond yields were even lower back then). Towards the beginning of that period you would have been better off holding long-term bonds for 20 years than trying to ride it out with cash or short-term bonds -- and this is despite the fact that cash yielded more back then than it does now. Towards the end, cash and short-term clearly would have been preferable -- as long as you were clever/lucky enough to time the switch back into long-term. After the back of inflation was finally broken in early 1980s long-term bonds became (in hindsight) bargains of anyone's lifetime -- a chance to lock in 15% returns for 20-30 years! Of course, vast majority of investors of the day were too badly burned by the previous decade's inflation to consider locking into a fixed rate for that long.

Are we in the beginning or middle or end stages of this round of low interest rates? Are we due for a 1970s-like inflationary blow-up or will the rates simply gradually return to normal? Nobody knows (though a lot of people sure have strong opinions!) And even if the history was clear on the short-versus-long term issue it still wouldn't have settled anything definitively -- there is no reason for things to play out the same way this time around.

So it comes back to individual investors to decide what to do. One extremely reasonable course of action is to simply tune out all this talk of interest rates and market timing and to continue holding and investing in VBMFX which holds all investment-grade taxable bonds of all maturities in approximately market-weight proportions. I would never say that sticking with VBMFX is a bad idea. And for investors in higher tax brackets forced to hold fixed income in taxable accounts, Vanguard's municipal bond funds are perfectly appropriate as a set-it-and-forget-it alternative to VBMFX.

For those who like to tinker and/or try to squeeze every last penny from their investments, there are some obvious ways to optimize their fixed income portfolios without sacrificing their risk profile (by reaching for yield in lower quality or longer maturities). Ways to accomplish this optimization is what I try to focus on in my posts. I'll repeat some of these ideas here even though it sounds like you are fully aware of them already.

An Idea

Again, the most obvious no-brainer optimization that investors can do today is to max out their I-Bond purchases every year until the interest rates change substantially. This can even be done using "emergency fund" cash because after one year your I-Bond becomes redeemable at very small penalty and from that point forward it doubles as a higher-yielding, tax-deferred emergency fund. I-Bonds are clearly preferable to any TIPS of any maturity at today's prices. On top of that, by investing in I-Bonds in taxable space you can sell off the equivalent amount of less attractive tax-sheltered bond holdings (e.g. TIPS) and use that space for equities or maybe more appealing fixed income options.

EE-Bonds are a much bigger commitment since they have maturity of 20 years. But if you look over your entire portfolio in detail, chances are pretty good that you already invest in something with similar maturity/duration. Do you hold VBMFX or BND or any other "total bond market" funds from Vanguard or other fund companies? A quarter to a third of those funds' holdings is securities with maturities of 20 years or more!

So what you could do is replace VBMFX with a roughly 3:1 or 2:1 mix of short-term fund such as VBISX and EE-Bonds and have pretty high degree of certainty that for the next 20 years your new fixed income portfolio will have higher return with less risk than your original VBMFX. Obviously the limit of purchases of EE-Bonds puts a damper on this strategy for very large portfolios but right around now, assuming you are married, you could be putting $40,000 into EE-Bonds ($10,000 each for you and your spouse in 2012 and then again in 2013). That's enough to transform a $160K fixed income portfolio from all-VBMFX to $120K VBISX and $40K in EE-Bonds.

But wait, there's more! Then you realize that US nominal treasuries on the short end of the yield curve (such as those held by VBISX) are priced for zero returns and even short-term corporates (such as ETF VCSH) have yields to maturity lower than some available 3-year CDs. So you sell your newly acquired VBISX and put the proceeds into a PenFed IRA with a 3-year 1.7% CD (although PenFed is a military credit union, its membership is open to civilians who make a small donation to a military charity) or into any one of other CDs available at brokerages such as Fidelity.

Let's analyze what we just did. We exchanged $160,000 in VBMFX/VBTLX with 1.5-1.6% yield and maturity of 7 years with 5-year duration (in practice duration should be thought of as being higher, closer to 7 years -- a good chunk of VBMFX holdings are in mortgage-backed securities whose duration effectively increases with higher interest rates) for $120,000 in 1.7% PenFed CD with 3-year maturity/duration and $40,000 in 3.5% EE-Bonds with 20-year maturity/duration.

• Average weighted maturity/duration of our new portfolio is just about 7 years -- equal to VBMFX maturity and duration for all intents and purposes. 
• Our average weighted yield to maturity is now full 0.5% higher. 
• All our fixed income holdings are now 100% backed by the full faith and credit of the US government (as opposed to only about two thirds of VBMFX holdings).
• In case of unexpected rise in interest rates in, say, 3 years (whether due to inflation or other causes), no matter how high they go we'd "lose" no more than about 10% (3.5% per year over 3 years) of our EE-Bonds value and nothing at all in our CDs. That translates into a loss of just 2.5% of our overall portfolio value and it doesn't matter how high the rates go! By comparison VBMFX will lose around 5-7% of its portfolio value for every 1% rise in interest rates! Feel free to run similar analyses on time periods other than 3 years.

As you can clearly see with some creative manipulation we slightly improved the yield of our "standard" VBMFX-based fixed income portfolio while significantly improving its safety, both from potential defaults and from potential rising interest rates.

If this is too complicated, then you could have also skipped the EE-Bond portion of our exercise altogether and simply gone for a 7-8 year FDIC-insured CD at around 2.0% APY.

Note that our ~2.0% return over 7 years from FDIC-insured/US-govt-backed investments is approximately the same as available from corporate and municipal bonds of the same maturity -- neither of which, of course, enjoy the same backing. Though municipal bonds have higher after-tax returns.

A Thought

Changing gears from optimizing the portfolio let's consider how rising interest rates may end up not harming long-term bonds. The key here is to realize that as long as yield curve retains its positive slope (meaning shorter maturities yield less than longer maturities), it is perfectly possible for interest rates to gradually rise without any drop in existing bond values.

Suppose there exist a 10-year bond with 3.0% YTM (yield to maturity) and a 9-year bond (by the same issuer) with 2.5% YTM. Suppose that interest rates gradually rise by 0.5% over the course of one year, across the entire yield curve. After a year an average 10-year bond would have 3.5% YTM and a 9-year bond would have 3.0% YTM. What do you think happened to our 10-year bond from a year ago over the course of this year? Nothing! It simply threw off its expected 3% interest and glided along the yield curve into becoming a 9-year 3.0% bond! It experienced absolutely no drop in value even though the interest rates rose by 0.5%. The mechanics would be somewhat different, but the end result would be the same for a bond fund.

The key point is that it is perfectly possible for a rise in interest rates to not negatively affect the value of existing bonds. It is even possible for existing bonds to rise in value simultaneously with rising interest rates! If in our example we only experienced a 0.3% rise in interest rates across the yield curve over one year's time, then our original 10-year bond would have actually appreciated slightly by the time one year passed and it morphed into a 9-year bond! Interest rates impact bond prices negatively only if the rise is relatively sharp and/or the yield curve is flat or inverted.

Today's yield curve still has a firmly positive slope, even if it's not as steep as it was a couple years ago. If interest rates happen to rise at the snail's pace of about 0.05% per year at the long end of the curve, then none of today's long-term bonds will suffer any noticeable losses as they "slide down" the slowly rising yield curve. At the short end the rise could be even steeper. If the 5-10 year section of the yield curve rises by 0.20% annually for the next five years, today's 10-year bonds will still be safe from any losses in value in that time span.

Now, granted, interest rates can move at much faster pace than that. As I noted in the beginning of this post, the yield curve dropped by nearly 2% in just about 2 years on the long end! It could go back up just as quickly. But the point holds that a positively-sloped yield curve still carries some "built-in protection" for longer-term bonds. Not every rise in interest rates dooms long-term bonds to losses.

Conclusion

This post ended up being a mish-mash of thoughts and ideas, but I suppose that's what you asked for. Hopefully it proved at least interesting, if not actionable. I will simply re-iterate that I don't think there are any bargains in the fixed income market in absolute sense, but there are very obvious relative bargains (I/EE-Bonds, CDs) and relative duds (treasuries/TIPS). If you wish to optimize your fixed income portfolio, you should definitely focus on exchanging the relative duds for the relative bargains as much as possible (which, of course, you have already done).