Wednesday, December 26, 2012

EE Bonds Versus Other Long-Term Bonds

Yet another post on why I feel that EE Savings Bonds are currently preferable to any other long-term bond investments... Before I proceed with EE-Bond discussion, remember that I Savings Bonds are an even easier decision than EE-Bonds since I-Bonds don't require the 20-year commitment that EE-Bonds do and readily double as "high-yield" tax-sheltered emergency fund. But that's all you'll read about I-Bonds in this post...

EE-Bonds Versus Long-Term US Treasuries

This one is very simple. 20-Year US Treasuries, which equal EE-Bonds most relevant maturity point, have yield to maturity of about 2.4% versus EE-Bonds' 3.5%. Today's 30-Year US Treasuries, which roughly equal EE-Bonds duration (remember that today's EE-Bonds are most easily thought of as a zero-coupon bond with 20-year maturity and, therefore, 20-year duration) have yield to maturity of about 2.9% versus, again, EE-Bonds' 3.5%. Whatever you wish to choose as point of comparison among long-term US treasuries, EE-Bonds easily come out on top in yield to maturity. And this is before we consider EE-Bonds deferred taxation and their built-in put option which might prove very valuable in interest rates rise significantly in the next few years.

EE-Bonds Versus Long-Term Investment Grade Corporate Bonds

The comparison against corporates is not quite as cut-and-dry, but EE-Bonds still come out on top. Vanguard's long-term corporate bond funds have maturity of more than 20 years and duration of less than 20 years which, to me, makes them a sufficiently apples-to-apples comparison with the 20-year maturity/duration EE-Bonds. From http://www.longtermreturns.com/2012/03/q-corporate-bonds-or-us-treasuries.html we know that historically 20-year cumulative (not annual!) default rates on corporate bonds have averaged:
  • 1.7% for Aaa
  • 5.3% for Aa
  • 6.8% for A
  • 13.2% for Baa (which the lowest investment-grade rating)

Vanguard's slightly-higher-quality long-term corporate bond fund VWESX/VWETX (VWETX is the Admiral version) has SEC yield of about 4.0% and portfolio mix of about:

  • 4% Aaa
  • 16% Aa
  • 55%% A
  • 25% Baa

Assuming the next 20 years bring us historically average corporate default rates we'd expect cumulative defaults of about 7.9% which works out to about 0.4% annually. Subtracting that from headline SEC yield we get expected return of 3.6% from this fund.

Vanguard's slightly-lower-quality long-term corporate bond ETF VCLT has SEC yield of about 4.3% and portfolio mix of about:

  • 1% Aaa
  • 7% Aa
  • 46% A
  • 46% Baa

Assuming the next 20 years bring us historically average corporate default rates we'd expect cumulative defaults of about 9.6% which works out to about 0.5% annually.  Subtracting that from headline SEC yield we get expected return of 3.8% from this fund.

As a reminder, expected annual return on EE-Bonds over 20 years is slightly above 3.5% (2 ^ (1/20) - 1 if you like math) which is almost in line with VWESX/VWETX and slightly lower than VCLT. But what you get in return is best available protection in case the next 20 years turn out to be worse than average as far as default rates go, as well as deferred taxation and that same potentially very valuable put option to redeem EE-Bonds at more-or-less (actually a hair higher than) principal value should interest rates unexpectedly rise soon. To me those characteristics are more than worth the tiny relative difference in expected returns. To be fair historically corporate bonds recovered about half of their value even in defaults, but that extra 0.2% annually does not change the picture substantially for me.

EE-Bonds Versus Long-Term Municipal Bonds

Vanguard's "long-term" municipal bond funds have maturity and duration of well under 10 years which means that they are not suitable for an apples-to-apples comparison against EE-Bonds. However several other companies, including PIMCO, Invesco PowerShares, and State Street offer so-called "Build America Bonds" (BAB) ETFs which can be thought of as taxable municipal bonds and which, conveniently for our analysis, offer maturities and durations similar to those of Vanguard's long-term corporate bond funds: maturities of over 20 years and durations of under 20 years. This makes BAB ETFs suitable for a comparison with EE-Bonds.

BAB ETFs vary in their composition, but generally have slightly higher credit ratings than long-term corporates: something like 10% Aaa, 45% Aa, and 45% A. Historical default rates for municipals have been lower than for similarly rated corporates, so we'll somewhat generously assume zero losses due to defaults on our BAB ETFs. Their SEC yields are about 4.0%. Those characteristics combined make BAB ETFs toughest competition yet to our EE-Bonds. The question becomes whether the approximately 0.5% in extra yield of BAB ETFs is enough to overcome the extra perks of EE-Bonds. In taxable accounts the answer would be a pretty clear "no" (since, remember, BABs interest is taxable). But in tax-sheltered accounts you could make a case for them. Personally though, I would forego that 0.5% extra and go with EE-Bonds for their ultimate safety and to save the tax-sheltered room for other investments. I might lose a tiny bit relative to BABs in the most BAB-friendly scenario but can come out quite a bit ahead in BAB-adverse ones.

The above choice is underscored by the yields available from truly safe munis: 2.75% on 30-year Aaa as can be seen at http://www.bloomberg.com/markets/rates-bonds/government-bonds/us . Note that now we're talking about tax-exempt municipal bonds. Still, a 30-year 2.75% has duration of slightly longer than 20 years (and obviously a 30-year maturity), but would still lose to the after-tax rate on today 20-year EE-Bond for up to 28% tax bracket. And this is again, not counting the considerable value of EE's built-in put option.

Conclusion

Hopefully the above comparisons make it clear why I view today's EE-Bonds as best available long-term fixed income choice. In no way are they bargains in absolute terms, but they are your best available option if you want to hold long-term bonds in your portfolio. Remember that EE-Bonds require a 20-year commitment to become a worthwhile investment and if you don't have that long an investment horizon you should not be considering them. Of course, you can also make the choice of foregoing long-term fixed income investments altogether until interest rates rise. This is not unreasonable. But, unfortunately, there is no guarantee that the interest rates will rise any time soon. Around the middle of 20th century they stayed at these levels for a couple of decades!

I wish I could tell you whether today's long-term fixed income investments are to be jumped on or avoided, but I cannot. What I can tell you with high degree of certainty is that of what's available in this space, EE-Bonds are the best deal going. If you wish to hold long-term bonds today, and especially if you have to invest in taxable accounts, then you have to very strongly consider EE-Bonds.

Remember that you can buy EE-Bonds only electronically through TreasuryDirect.gov at up to $10,000 per calendar year per Social Security number. It takes just a few minutes to sign up online and make the purchase. And not to put any pressure on you, but you only have a couple days left if you wish to buy your 2012 allotment of EE-Bonds.

15 comments:

  1. Sold! I bought my I-Bonds earlier in the year and took the plunge into EE's last week. Your posts (especially the graphs) have helped me convince myself that its a prudent decision. (They helped bring my wife on board, too.) I'm 50 and I figure I'll buy the EE's for a few more years and they'll mature early in our retirement. At some point I'll stop buying the EE's because I worry about our heirs selling too soon and about me going senile and making a poor decision myself! Anyway, your writing is influencing people in a positive way. Keep up the good work and have a great New Year!

    PS - Is there a good reason to buy EE-Bonds or I-Bonds quarterly or semiannually vs. once per year? (Managing the eventual sale of these would seem to be easier if there was only one purchase per year.)

    ReplyDelete
  2. I'm always very happy to hear that my blog was helpful to somebody -- thank you for letting me know! And Happy New Year to you too and to everyone else who reads this blog!

    As far the exact timing on EE/I purchases, it doesn't matter a whole lot. I-Bonds get adjusted in value monthly. So each month you delay the purchase will be one less month for which you get interest. There is also a slight (and I do mean slight) advantage to delaying the purchase of I-Bonds till the end of the month since you capture the full return for that first month regardless of the day of your purchase. Of course, to profit from this delay you'd need to have a profitable way to hold the funds for the rest of the month -- and no such particularly profitable short-term investment exists these days. Maybe you could squeeze a Starbucks coffee's worth of profit per year by optimally timing the I-Bond purchases. Not worth stressing about in my opinion.

    With EE-Bonds the situation is even simpler. You should be buying them for one reason only: to see them double in value on their 20th anniversary giving you that virtual 3.53% annual yield. The longer you wait to buy, the longer you'll wait for that 20th anniversary.

    ReplyDelete
  3. As I suspected - it's just nice to have independent confirmation. Thanks!

    ReplyDelete
  4. Hi Grateful reader here. But I'd like to put some extra thought in this.

    If you compare EE bond and long term treasury isolated, after 20 years. EE bond is a no brainer winner. However, EE bond price doesn't fluctuate with the interest rate. I'd like to view it as a 0.2% CD for 19.9 years, and a big bonus in the end. As a pessimistic person, I'd rather view EE bond as an alternative of cash, especially during early years of holding.

    If we relive 2008. EE bond won't offset the huge stock drop, as long term treasury would, which gives me a peace of mind.

    ReplyDelete
    Replies
    1. You are correct. If for some reason you were forced to redeem the EE prior to its 20th anniversary you would be stuck with essentially zero return on your investment. For that reason EEs are not suitable for re-balancing, such as in a redux of 2008.

      The flip side to this is that EEs are not particularly vulnerable to a major hike in interest rates in their early years. You can redeem them, get your principal plus a few pennies back, and re-invest the proceeds at a higher rate. This is much preferable to long-term treasuries which would be subject to large price drops in the same situation. Of course this is more of an advantage for the first 5 years or thereabouts of EE's life since its yield to maturity increases every year (dramatically so in final years) which means it only makes sense to redeem it early if there are equivalent investments with even higher yields to maturity available.

      Which of the factors ends up more valuable depends wholly on the future events. If we experience another 2008-style meltdown where nominal treasuries skyrocket and everything plunges, then EE will prove less appealing than nominal long-term treasury. If instead we experience a large spike in nominal interest rates (whether due to inflation or anything else), then EE's early redemption option will give it the edge.

      Delete
    2. I somewhat agree with you. But I'd also like to make a point that raising interest rates is not necessarily that bad for long term bond fund. As we know, in a bond fund, it will constantly sell short term bond for long term bond. So the interest rate raise, at least early in the holding, is not a bad thing.

      To make a comparison of EE bond to a super simplified LT bond fund. Suppose the bond fund has only 1 bond, which is 20 year zero coupon treasury. After 10 years, we sell that (note that in 2023 it will become a 10-year zero coupon), and use the money to buy another 20-year zero coupon, and sell all in 2033.

      We imagine a hypothetical interest rate curve future:
      today: 20-yr 3%, 10-yr: 2%
      2023: 20-yr 10%, 10-yr: 6%
      2033: 20-yr 20%, 10-yr: 12%

      1000$ investment in the "bond fund" today would end up with 2164$ in 2033, outperform 2000$ in the EE fund (without thinking about tax impact).

      Of course if we change some numbers in the interest rate curve future, the result will differ. EE bond will win in many situations, including the interest rate keep declining.

      I guess what I'm really trying to say is that if one has asset allocation with LT bond (e.g. the 4x25 permanent portfolio), he/she should think more on replacing the LT bond fund with EE bond. Use EE bond as cash equivalent seems to have not much downside though.

      Best,
      Anonymous

      Delete
    3. Agree with everything you said. In the end it will come down to how the yield curve moves over the next 20 years. Since it's hopeless to try to assign probabilities to all the different possible scenarios, I fall back on yield to maturity and maturity/duration comparisons. The EE has higher YTM than any other treasury with either maturity of 20 years or duration of 20 years and that's good enough for me to go with the EE over other long treasuries.

      Delete
  5. I currently have a 3.375% 7-1 ARM mortgage. Should I use extra cash to pay off the mortgage or buy EE-bond?

    ReplyDelete
    Replies
    1. Paying off the mortgage would take out any uncertainty of the 7-year rate reset and (hopefully) provide some psychological benefit of owning the home free and clear. It's likely that (depending on exact tax rates and future yield curves) the EEs end up returning ever so slightly more than the savings of paying off the mortgage, but the difference will probably be miniscule. I would pay off the mortgage.

      Delete
  6. This comment has been removed by the author.

    ReplyDelete
  7. If buying individual bonds, yes EE is preferable to LT Treasury bonds bought today and held 20 years. But given we can pretty much assume rates will rise during the next 20 years. An intermediate bond fund or a bond ladder will have rising returns over time as rates rise, likely over 3.5 percent. Plus they have the liquidity to sell before then getting back more than principal. Ibonds too don't have this problem, since you only loose 3 month's interest.

    ReplyDelete
    Replies
    1. I disagree that we can pretty much assume rates will rise. They have been lower and stayed lower for nearly two decades in 1940s and 1950s. Not only that, but that period has some funny parallels to today: high national debt (post-WW2) and artificially low interest rates due to actions of the central bank.

      I think a more prudent way to approach the situation is to assume that we know nothing about future rates and then to invest in the (relatively) obvious bargains and avoid the duds in today's fixed income space. Both I-Bonds and EE-Bonds are very clearly in the bargain category relative to alternatives of similar maturities and/or credit quality. If you hold long-term bonds -- and if you invest in something like the total bond market, then a quarter to a third of your holdings is in long-term bonds -- then EE-Bonds should be near the top of your list of ways to squeeze a bit more out of today's fixed income choices.

      You could of course simply refuse to invest in long-term bonds entirely due to below-average interest rates. But in my opinion that's not at all a sure bet. It's the short-term rates that are sure to rise, not the long-term ones.

      Delete