Page 1 of 1

Bond duration compensation

Posted: Sat Jan 12, 2019 10:06 pm
by thisisallen
Is there a general rule regarding how many additional basis points are recommended to compensate the investor for the risk of each extra year of duration?

Say that the interest on a 2 yr note is 2.80. On a 5 yr it is 3.4. Difference is 60 basis points. There is 20 basis points extra for each of the 3 yrs difference between 2 and 5 yrs.

Thx and regards,

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 7:38 am
by ochotona
Allen it's all determined by Mr. Market. There is no rule or formula. And it changes continually and lately really quickly.

Look at this website, it updates daily.

https://www.treasury.gov/resource-cente ... data=yield

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 9:14 am
by buddtholomew
thisisallen wrote:
Sat Jan 12, 2019 10:06 pm
Is there a general rule regarding how many additional basis points are recommended to compensate the investor for the risk of each extra year of duration?

Say that the interest on a 2 yr note is 2.80. On a 5 yr it is 3.4. Difference is 60 basis points. There is 20 basis points extra for each of the 3 yrs difference between 2 and 5 yrs.

Thx and regards,
If I recall Larry Swedroe has a formula for term risk, but this is not applicable to the PP investor. We want the investment with the highest duration even though the rates between 30 and 10-year Treasuries do not compensate for the additional risk.

A 1% decline in interest rates will result in approximately a 17% increase in LTT’s and only a 5-6% gain in ITT’s.

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 10:16 am
by thisisallen
ochotona wrote:
Sun Jan 13, 2019 7:38 am
Allen it's all determined by Mr. Market. There is no rule or formula. And it changes continually and lately really quickly.

Look at this website, it updates daily.

https://www.treasury.gov/resource-cente ... data=yield
That’s a better/more accurate site for the rates than the one I was using at Fidelity. Thanks!

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 10:25 am
by thisisallen
buddtholomew wrote:
Sun Jan 13, 2019 9:14 am

If I recall Larry Swedroe has a formula for term risk, but this is not applicable to the PP investor. We want the investment with the highest duration even though the rates between 30 and 10-year Treasuries do not compensate for the additional risk.

A 1% decline in interest rates will result in approximately a 17% increase in LTT’s and only a 5-6% gain in ITT’s.
The term risk may not be necessary for the PP investor at the long end, but perhaps it would be useful for the short term slice?

Gaining/losing 17% for LTT’s change of 1% is something to seriously consider. I help my Father, in his 90’s, and it’s a bit scary to think he could lose that much principal at a time when he may need to sell bonds to pay for medical expenses. At this stage, he is set up almost all in bonds.

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 11:51 am
by ochotona
Don't have bonds for your Dad past his maximum life expectancy. That's how my Mom is set up. She's 89, her longest bond matures in 11 years. Why go longer?

She has gold to compensate for any inflation in the next decade, and a little bit of equity index funds in her Roth.

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 5:10 pm
by Kevin K.
I may well be being obtuse here (wouldn't be the first time) but I'm not seeing any advantage in extending Treasury bond maturities at current rates. The chances of a 1% (or even a .25%) decrease in rates is small, while the odds of a .50% increase this year are high. And yeah, I do remember how LTT's saved the PP during 2008's flight to safety, but baseline interest rates were much higher then, not low and flirting with inversion as they are now. I've read several articles recently in which seasoned financial advisors who'd normally have their clients in intermediate term bonds are instead recommending at least 25% of the total portfolio in cash.

Gold absolutely does not provide inflation protection - only "SHTF" protection. Stocks can help offer such protection in the long run, but as Tyler points out cash is actually the most consistent inflation buffer:

https://portfoliocharts.com/2017/05/12/ ... -investor/

Vanguard Treasury MM fund is yielding 2.32% right now if you can meet the 50K minimum, or one can buy 6-12 month T bills commission-free at Vanguard, Schwab and (I think) Fidelity. Plenty of volatility and potential upside in equities and gold without risking catastrophic losses on the bond side.[/i]

Re: Bond duration compensation

Posted: Sun Jan 13, 2019 10:03 pm
by thisisallen
Kevin K. wrote:
Sun Jan 13, 2019 5:10 pm
I may well be being obtuse here (wouldn't be the first time) butt I'm not seeing any advantage in extending Treasury bond maturities at current rates. The chances of a 1% (or even a .25%) decrease in rates is small, while the odds of a .50% increase this year are high. And yeah, I do remember how LTT's saved the PP during 2008's flight to safety, but baseline interest rates were much higher then, not low and flirting with inversion as they are now. I've read several articles recently in which seasoned financial advisors who'd normally have their clients in intermediate term bonds are instead recommending at least 25% of the total portfolio in cash.

Gold absolutely does not provide inflation protection - only "SHTF" protection. Stocks can help offer such protection in the long run, but as Tyler points out cash is actually the most consistent inflation buffer:

https://portfoliocharts.com/2017/05/12/ ... -investor/

Vanguard Treasury MM fund is yielding 2.32% right now if you can meet the 50K minimum, or one can buy 6-12 month T bills commission-free at Vanguard, Schwab and (I think) Fidelity. Plenty of volatility and potential upside in equities and gold without risking catastrophic losses on the bond side.[/i]
Your point about “not seeing any advantage in extending Treasury bond maturities at current rates” is the same advice the Fidelity bond reps give me when I ask about purchasing bonds in my Father’s account.

Fidelity does offer commission free purchases of T bills, notes, bonds at auction.

Re: Bond duration compensation

Posted: Mon Jan 14, 2019 2:30 am
by boglerdude
> The chances of a 1% (or even a .25%) decrease in rates is small

Predicting the future eh. How long will Japan's 10 year stay at 0

Re: Bond duration compensation

Posted: Mon Jan 14, 2019 2:39 am
by ochotona
We can predict... Most retirees won't live past 100

Re: Bond duration compensation

Posted: Mon Jan 14, 2019 10:27 am
by Kevin K.
boglerdude wrote:
Mon Jan 14, 2019 2:30 am
> The chances of a 1% (or even a .25%) decrease in rates is small

Predicting the future eh. How long will Japan's 10 year stay at 0
I don't have to predict the near-term future since the Fed is doing it for us. They've done exactly what they said they'd do with rate increases and while they may or may not follow through with the two further bumps planned for this year interest rate decreases are not on the horizon. Vis-a-vis Japan and the bigger picture, this excerpt from Todd Tresidder's excellent 2013 article "The Great Bond Bubble" still seems very timely:

"Further proving how extreme our current point in history really is, Timely Portfolios did a fascinating study on 10 Year U.S. Treasuries (constant maturity similar to IEF) evaluating what would happen if interest rates went to their theoretical minimum – zero.Are you ready – drum roll please – a paltry 17% gain. Shocking, but true.

That means that if interest rates went as far as they could theoretically go on the downside, the entire upside left on 10 Year Treasuries is a mere 17%.That’s not much considering it’s the mathematical extreme case limit for how far it could go. Your actual returns would depend on how long it took to reach this limit.
For example, Market Sci blog showed that if it took 4 years to reach zero percent interest rates, then the annualized return would be just 5%, and the total return would be 21.6%.

This may not sound too bad until you realize that 0% yield is not very likely. Given that Japan’s record low yield for their 10 Year Note was .47%, it’s interesting to note the maximum upside is reduced to a mere 13.8%.If it takes 4 years to get there, your annual return is just 4.1% with a total return of 17.6% – before inflation! If you net out inflation, the real return is minuscule at best.

The implication is clear – the upside potential in bonds makes no sense compared to the downside risk.

The problem is bond prices move inversely to interest rates, and interest rates are approaching their theoretical floor. In other words, as rates decline, bonds rise in value. As rates rise, bonds lose value. There is little room left for interest rates to fall, and tons of room for rates to rise, creating an unfavorable risk/reward ratio. This problem is further exacerbated by the fact that current low interest rates would cause a modest rise in rates to cause disproportionately large losses that could dwarf any income received in the interim.

For example, as of this writing, a mere 1% rise in interest rates on the Treasury long bond should equate to a roughly 20% price decline, wiping out 7 years of income at current interest rates.

Do you think it’s reasonable to expect a mere 1% increase in interest rates from these historically low levels over the next seven years as the above example illustrates? After all, far worse has occurred in the past when markets were less volatile.
For example, the 30 year Treasury yield rose 240 basis points in just 9 months back in 1994. Just imagine what a 2 – 3% rise (or more) would mean to investor portfolios given the above example.

This is critically important because fixed income’s traditional position within asset allocation is as a “safe investment“. In fact, we have entered one of those rare points in history where the risk/reward analysis on bonds could conceivably be more dangerous than equities, because the historically low coupon implies historically unprecedented volatility and downside price risk.

In other words, capital loss risk to bonds is highest when starting yields are lowest. Given that yields are at all time historical lows, many historical benchmarks for capital losses in bonds are unrealistically conservative. The future could easily be far worse than the past.

For example, according to Welton Investment Corporation the deepest (-15.3%) and longest (8+ years) Aaa corporate bond drawdown occurred from 1954-1963 because of a tiny 1.8% increase in interest rates – a hiccup by today’s volatile standards. The reason is because the starting yield in 1954 was an equally tiny 2.85%.

In other words, the drawdown severity and duration isn’t determined exclusively by the magnitude of the interest rate rise. It’s determined by the relationship of the interest rate increase compared to the starting yield. Today’s record low yields imply historically high risk of capital loss.

For example, Welton also analyzed what could happen to Aaa corporate bonds under different interest rate increase scenarios:

A 6% increase spread over 5 years would result in a 36.2% drawdown and a 6.4% annual loss.
A 4% increase in just 1 year would result in a whopping 34.8% drawdown and a 34.8% annual loss.
Even a modest increase spread over many years could cause zero return (or worse) for more than a decade."

Putting 25% of the portfolio in LTT's that have a strong possibility of experiencing sustained major losses and then offsetting that gamble with 25% cash returning 2.5%, 25% overvalued market-cap weighted U.S. equities and the remaining 25% in a volatile asset with no inherent rate of return certainly doesn't look like the all-weather bunker the PP is supposed to be.

Re: Bond duration compensation

Posted: Tue Jan 15, 2019 11:54 pm
by boglerdude
Japan's holding the 10Y at zero
http://bilbo.economicoutlook.net/blog/?p=40250

Rates are historically low because the real world evolves. Millions of Asians are moving out of poverty and there's a savings glut.

Rates can go negative, and more easily in the future, as we go cashless.

That said, I'm still butthurt about buying EDV at 2.3%, before Trump's election increased inflation expectations. If there's another flight to safety I'll probably sell some and underweight LTTs

Re: Bond duration compensation

Posted: Wed Jan 16, 2019 2:49 pm
by WhiteElephant
Thanks for your post Kevin K.
Because of concern about long term bonds, I adjusted my portfolio by lowering duration. Currently you gain very little by holding on to long-term bonds while the potential downside risk is incredible. It's probably even worse for a Eurozone PP, like mine.

Maybe rates will stay this low for 20+ years, that’s possible of course. But if they don’t it’s possible to suffer very severe losses. Especially if stocks and gold decide to go down as well.

I couldn't feel comfortable anymore by holding on to my eurozone bonds so I've traded them for short-term CD's. Because the duration of my fixed income is much lower now, I've lowered my gold exposure as well, and increased the equity part a bit.

I've basically transitioned into a low duration Desert Portfolio now. Not perfect, but it definitely feels safer than my previous allocation.

Re: Bond duration compensation

Posted: Wed Jan 16, 2019 4:27 pm
by buddtholomew
Interestingly, when there is a flight to safety 3.07% (today’s yield) looks appealing. Not so long ago investors were holding onto LTT’s in the 2.5% and lower range.

Don’t forget we want LTT’s not for the increased yield but rather the increased duration. A 1% rise is a 17% loss, BUT a 1% decline is a 17% gain...

Re: Bond duration compensation

Posted: Thu Jan 17, 2019 1:10 am
by WhiteElephant
A 3% yield might still look appealing, but what about long-term yields <1%, like Germany? When are yields too low?

30year German bonds are yielding 0.8% currently, that’s comparable with a 2-year CD with basically zero risk.

The volatility of yield changes at such a low starting point will be explosive. I didn’t feel comfortable with this anymore, and adjusted my portfolio accordingly.