At What Yield Do Bonds Begin To Make Sense Again?

by: Jim Sloan


Bonds may not be what you think they are. Extremely low yields call their role in portfolio risk reduction into question.

Historical studies show that over long periods a bond allocation reduced the severity of drawdowns and frequency of down years significantly while only slightly reducing the rate of return.

Low rates both lower the return from bonds and make them less safe; a linked table/chart shows that rates under 3% for the 10-year Treasury occur about 20% of the time.

For the above reasons, I haven't owned bonds since the 2009 crisis, holding mainly cash and some I Bonds, the zero duration providing better safety and convenience.

I would consider a small position if the 10-year Treasury yield reached 3%, adding in stages at the 136-year median (3.87%) and mean (4.58%).

Bonds may not be what you think they are. They once were the principal source of portfolio ballast. They helped stabilize and offset the risk of stocks. They provided safe income. They haven't really been those things for a few years now, but they may be again in the future. Let me explain.

Bonds were once the great friends of retirement investors and insurance companies, which have to defease future liabilities. Retirees and insurance companies may sound like an odd couple, but they actually have a good bit in common. Both need dependable and predictable income. If you run an insurance company, bonds are the asset which combines safety with the predictable cash you need at known future moments. It's the same with retirees. As you get older, your needs are focused on getting income at regular intervals, and it must be dependable.

It used to be doctrinal: The older you get, the more bonds you should have in your portfolio. That's why most advisors, human or robot, have you increase your allocation to bonds at a steady pace as you grow older.

Bonds are also the anchoring asset for risk. The equity risk premium is a simple number for the amount of excess return you should expect to receive from equities in order to justify their additional risk. The measure to which it is added is usually the rate available in 10-year Treasury bonds. That's the basis for the canonical but somewhat misleading Fed model, which argues that stocks are okay to buy when their earnings divided by price (inverse of the PE) exceeds the yield on the 10-Year Treasury - possibly with room for some risk premium. The glaring flaw of this model is that it ignores the fact that stocks have very long duration and the 10-year yield is a moving target.

When you measure risk as variance, bonds have historically had a lot less risk than stocks. Bonds therefore provide safety, a modest return, and a reasonable amount of flexibility in case you need to realize cash. They used to anyway.

In recent years, bonds have not provided an easy solution for either insurance companies or retirees. In fact, they have become something between a challenge and a nightmare.

What Bonds Are Supposed To Do For You

Bonds backstudy well.

It is a major pillar of market doctrine that bonds dampen the volatility of your portfolio. In theory, they move in the opposite direction of stocks, at least at moments of market stress. This should smooth the path of long-term returns. The price for this smoother ride is a modest reduction in absolute return.

And it works. It has worked in the past, anyway. Many back studies of various stock/bond mixes demonstrate that this has been true over long periods - reduced volatility, fewer down years, and only modestly reduced returns at all levels of stock/bond combination.

The reduction in the number of down years and the reduced severity of drawdowns are especially important to retirement investors. For a useful table, Vanguard provides this backstudy using various indexes. It illustrates the modest diminution of total returns at each step as the ratio changes by stages from 100% stocks to 100% bonds. When you hold at least some bonds, the distribution of outcomes looks much less scary than when holding stocks alone. For this reason, portfolios holding some bonds are less dangerous to retirees who need to keep an eye on "sequence risk" (the risk of crippling down years coming early in the distribution phase).

The benefit of risk reduction provided by bonds is not spread evenly over all market conditions. This is both good news and bad news. A good bit of the risk control and portfolio smoothing done by bonds stems from a few crucial dates. You can infer this by looking closely at the data in the Vanguard study. The year 1931 recurs repeatedly as the worst one-year drawdown. This is true for every allocation except holding 100% bonds.

In extreme economic crisis, bonds do wonders for a portfolio. A move from 100% stocks to 50% in bonds reduced the worst annual drawdown (1931) almost in half (from 43.1% to 22.5%). To make the outcome sweeter, it did this mitigation of an extreme drawdown while reducing the long-term return by only 1.4% (10.1% to 8.7%). Owning some bonds provided a huge reduction of risk with a very slight reduction in returns. What's not to like about that?

So bonds were great in the early 1930s. They were also great during the brief 1987 Crash, the dot-com crash of 2000-2003, and the real estate based meltdown of 2007-2009. Fine.

But what are they like in more normal times? And what have they done for us lately?

What Went Wrong With Bonds?

Bonds are all about the coupon.

I suspect that many investors in bonds and bond funds are not aware of that fact. When you open your statement and see the change in value of your bond holdings, much of it represents the change in value of the bond(s) because of interest rate fluctuations. For holders to maturity, however - unless you bought the bond at a discount or premium - you get the coupon, which is interest paid at semi-annual intervals. That's the real investment return. Anything you receive from selling a bond before maturity is speculative return. Speculating in bonds is not recommended for individual investors. It's the coupon. That's the way you need to think about bonds.

The portion of your statements reporting bond returns will suggest something different from the facts set down in the above paragraph, but what they tell you about your return is unhelpful in thinking about bonds. The fluctuations in price over a given period of time distort your sense of what you are actually getting.

A good friend of mine who runs an insurance company wrings his hands when he sees the large capital gains in its bond portfolio because it means that money he must put in now will be inadequate to offset liabilities for business he would like to write in the present moment. Almost all insurance companies have had this problem in recent years. There have been various ways to address it, but the most common among the leaders has been simply to pass up the less profitable business, let their revenues decline, and increase earnings by using cash flow to buy back their own stock. Traveler's (NYSE:TRV), for example, has increased earnings and dividends per share regularly not by growing but by cutting its share count in half over the past decade. Unfortunately, if you are nearing retirement, you can't cut yourself and your needs in half.

We are now at or near the end of a 35-year-old bull market for bonds. Put another way, interest rates have broadly declined for 35 years. Most investors at most points along the way have received statements reporting higher returns than their coupon because the bond itself increased in price. For buy-and-hold investors that gain is illusory, because they will simply receive their original investment back at maturity. Only trend-following speculators have really profited from the apparent capital gains.

That being said, the coupon return on bonds which was quite good at the beginning of the 35-year period has shrunk in recent years to near invisibility. I think many bond investors are not very well informed on this history. I'm certain that many bond investors have not grasped how radically that changes the risks and rewards of diversifying a portfolio with bonds.

For the purpose of this discussion, let's continue to use the Treasury 10-year bond. It's a good place to start because the nominal risk is as low as it gets - virtually zero. The government will make its coupon payments and return your money in ten years. The duration is just long enough to make it a true long-term investment. Also, as mentioned above, it is one of the anchors for thinking about stock valuation.

To put the current and recent return on the 10-year Treasury in historical perspective, it is worth taking a few minutes to study this table. You should also flip to the accompanying chart. Several things about rates on the 10-year Treasury will jump out at you. For one thing, going all way back to 1871, the 10-year rate was never less than 3% until 1934.

Once the 10-year fell below 3%, however, it stayed there until 1956 - a total of 22 years. You would have done well to hold a lot of bonds during the 1930s. That was the decade of the Great Depression. There was deflation, which made bond return even more attractive. Stocks were going nowhere and they were going there in an erratic and volatile way. The low moment was 1941, when the 10-year yield dipped under 2%, the lowest ever until the crisis of recent years.

From 1941 forward, however, stock returns including dividends did better than bonds almost every year, and with only occasional major volatility. You would have done better to own just stocks. This was true not only through the 1949-1966 bull market, but through the miserable inflationary 1970s all the way to 1982. Those were 41 lost years for bonds in which they increasingly dragged down portfolio return and their contribution to portfolio stability was modest and infrequent. Throughout that period, rates trended gradually upward. No wonder bonds came to be scorned in the 1970s as "certificates of confiscation."

The next time the 10-year rate fell below 3% was in December 2008, the first of 8 instances up to the present. Altogether there have been 30 years in which the 10-year yield was under 3%. That's just 20% of all years from 1871 to 2017. Put another way, current yields on the 10-year are firmly within the lowest quintile. The coupon return offered by 10-year Treasuries was better almost 80% of the time from 1871 to the present moment. Another way of thinking about 10-year yields under 3% is that the next major move, whenever it comes, is likely to be a reversal.

That reversal may already have taken place. The moment may have been July 8, 2016.

Why I Don't Own Bonds And Haven't Since 2009

I don't own any bonds. I haven't owned bonds for several years. The lone exception is a portfolio of I Bonds, which are a form of Savings Bond which provides long-term inflation protection. A key facet of I Bonds is that they can be cashed in at any time after a year. For that reason, I Bonds have the same duration as a one-year CD. Most of my I Bonds were purchased more than a year ago and have a duration of zero.

The major asset with a zero duration is cash. At the present time, my fixed income allocation, except for those I Bonds, is entirely in cash. For those who wanted to improve the return by holding "near cash," I wouldn't argue. Because I occasionally use my cash reserve to arbitrage relatively safe deals, the inconvenience of "near cash" isn't worth it in my case. The point is to keep your fixed income allocation in very short duration. The major reason is that the paltry coupon just isn't worth the risk of wanting cash at an inopportune moment.

Let me explain using a single moment in the 10-year Treasury yield. On July 5, 2016, the 10-year yield dipped to 1.37%. On July 8, 2016, it repeated that 1.37% print. That was the lowest 10-year yield ever. Ever. At that moment everyone who held a 10-year Treasury was sitting on a capital gain. That was great, right? It probably felt great on that day.

Then something terrible for bondholders happened; everything reversed. The yield on the 10-year started to rise, and the price of the bonds started to fall. That July 8 price proved to be the exact moment of reversal, and the yield has since risen above 2.5%, touching 2.6% at two points. The consequences of that move are very significant when thinking about the risks of holding bonds at the current low level of rates.

Everyone who bought a 10-year Treasury bond on July 8, 2016, could expect returns of exactly 1.37% per year over 10 years if they held the bond to maturity. Obviously, there were some people who thought that tiny return was okay.

Here's the reason that 1.37% yield may not have been okay. When rates ran up to 2.50%, the price of the bond bought on July 8, 2016, traded on the market at a price decline close to 10%. That loss amounted to about 7 years of its coupon payments. You could always come out whole by holding the bond for 10 years, but that meant accepting the measly rate of return of 1.37% annually for 10 full years. Your yield to maturity would be 2.5%, but only after that 10% haircut on your original capital. If you needed the money for any reason in the meantime, you would have to sell the bond at a loss.

That's not the kind of safety and stability I want in my portfolio. It's why at anything like the current rates, I'll take cash or "near cash."

There's obviously a cutoff - a point below which under most circumstances it isn't worth it to own bonds of 10-year duration. It's the same point where rising yields make it worthwhile to own those same bonds again.

Once you have thought about the various angles, you can arrive at the point where return, risk, and convenience add up to meet your needs. I'll share my own thoughts about it.

What Yield Would Interest Me In Bonds Again

The real question is this: At what yield do bonds of intermediate to long duration become helpful again by providing the safe side of a balanced portfolio?

The answer to this depends to some degree upon the individual, especially as recent years with meager bond yields have pushed many individuals to seek yield in "bond proxies" such as REITs, utilities, telecoms, and consumer staples or by descent into bonds with greater credit risk such as emerging market or high yield bonds. While all of these assets have their place and their moments in a portfolio, none are exact substitutes for the income with safety provided by Treasuries. The risks of some of them, in fact, have become more apparent as bond yields rose after July of 2016.

You could return for a starting point to the table highlighted above. A good initial observation is the one previously cited: Yields have been below 3% only a little over 20% of the time. For that reason, I would personally not be inclined to buy longer duration bonds when the yield on the 10-year are lower than 3%. Below that, I would want to sit in cash and wait. I can understand many reasons that others might differ, including the views of some of the gurus who got the long-term bond bull market right and who still believe that rates will have another down leg. I don't agree with that view, but I don't discount it either.

A key question to ask is just how much it costs to turn down a yield under 3% for a higher level of safety. It means more to some investors than others. Personally, I would consider longer duration bonds again if the 10-year Treasury reached a yield of 3%, but only for a fraction of a normal bond allocation - perhaps a third.

To raise the level of bond allocation toward "normal," I would note two other numbers provided at the bottom of the chart. The median yield over the 136 years from 1871 to 1917 was 3.87%. That's the dividing point on the chart with half of the years higher and half lower. At that point, I might raise my allocation to two-thirds of "normal" for a person of my age and particular circumstance. The final number is the mean (average) yield over the 136-year period: 4.58%. At that point, I would move to a full "normal" bond allocation, with one caveat. The caveat, an important one, is that I would look closely at the trend and absolute level of inflation. A high level of inflation means lower real yield and would cause me to think about bonds much less favorably.

The chance to buy longer-term bonds with the 10-year yield at 3%, 3.87%, or 4.58% may or may not occur in the near future. Remember that starting in 1934, the yield on the 10-year remained under 3% for 22 years. For the last 15 of those years, however, stocks had no crash-like drawdowns and bonds were mainly a modest drag on returns.

It is also possible that my point of view is somewhat displaced by the fact that I am fortunate to have pension and Social Security income which even John Bogle has suggested might be considered as a "phantom bond allocation."

The great second best, as with stocks, is probably to use inexpensive funds which roll their bonds over and simply stick with your age-based allocation. A slow rise in yields would be a drag on returns but maybe not so much in a fund with diversified maturities. This is still the approach recommended by Vanguard, which basically says to work out proper stock/bond allocation for your age, diversify in both stocks and bonds, rebalance, and don't think.

Vanguard's mechanical low-expense approach is probably not bad for investors who aren't well versed in markets, but it's not what I do. I actually use Vanguard to hold most of my money, and they are absolutely great, but I only look at their age-based allocation suggestion every once in a while because it provides useful anchoring as to what they would have had me do.

The alternative strategy, which I don't endorse is owning a large overweight of "bond proxy" stocks and talking yourself into the idea that they are really just bonds. They aren't. All stocks have risks that bonds don't, and those particular stocks appear at present to have the primary risk of being overpriced.


Bonds have in recent years seemed much less able to perform their traditional roles in a balanced portfolio of reducing risk without greatly diminishing returns. This has made a particular problem for investors near retirement, who in the past have been able to depend upon bonds to mitigate "sequence risk" for annual withdrawals. Because of their recent low yields, bonds do not perform either function as well as they historically have - offering poor returns and having the possibility of declining at the same time as stocks, and for the same underlying reasons.

My personal solution has been to hold a significant amount of cash. Some of it might be deployed into bonds in stages with 10-year Treasury yields at or above 3%, 3.87%, and 4.58%. One's approach to this problem may vary considerably with personal circumstances. For those not confident of their ability to do it, I recommend using the age-appropriate asset allocation sketched out in the link provided above.

Disclosure: I am/we are long TRV.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.