Upside volatility and long term bonds

[Gumby]

Gumby,

Your calculator seems to be doubling the value of the bond for every halving in yield.

That would indicate, to me, a hypothetical "forever bond," definitely not a 30.

Ah. ok, I see. It seems that particular bond calculator was defective. Sorry about that!

[MachineGhost]

The distinction is important because "disinflation" is "normal" whereas a "deflation" is a "fat tail event" that is usually brought about by sheer government/regulatory incompetence in the economy (which goes both ways, i.e. enforcement as well as cryonism).  If the marketplace is actually free to adjust to previous malinvestments or inflationary excess or whatever, then the resulting "disinflation" would unlikely to turn into "deflation", or at least not be perpetuated for very long.  The difference between Japan and the USA is the difference between "deflation" and "reflation" due to cultural factors.

Stocks always outperform bonds in every economic climate except during "deflation".  But what matters more for what duration risk one should take in bonds is whether the trend of interest rates is falling or rising, not the absolute level per se, again except during "deflation".

MG

Deflation (that many confuse with disinflation) is when the absolute rate of change in inflation is negative, i.e. below 0.

I get what you're saying, but why is the distinction important? 

[MediumTex]

Stocks always outperform bonds in every economic climate except during "deflation".  But what matters more for what duration risk one should take in bonds is whether the trend of interest rates is falling or rising, not the absolute level per se, again except during "deflation".

Doesn't the statement above need some qualification?

It seems to me that there have been multi-decade periods where your statement is not accurate, and these periods were not necessarily deflationary. 

Sometimes simply buying stocks at the wrong time can cause bonds to outperform for many years.

As I recall, Japan has seen mostly flat inflation for the last 20+ years (there hasn't been much actual deflation), and yet the Nikkei is 75% off of levels it was at 23 years ago.  Where are the superior stock returns compared to bonds in Japan? 

[MachineGhost]

The last 30 years bond market bubble is truly an aberration to make one think bonds could outperform stocks.  Here are the average real returns from 1900-2010:

Severe Inflation: Stocks -7.3%, Bonds -12.2%
High Inflation: Stocks 2.5%, Bonds -2.8%
Moderate Inflation: Stocks 11.1%, Bonds 3.6%
Low Inflation: Stocks 5.9%, Bonds 4.5%
Deflation: Stocks 16.7%, Bonds 9.5%
(Mild Deflation: Stocks 23.69%, Severe Deflation: Stocks -1.34%)

[align=center][/align]

But I agree with you that starting valuations also matter.  The U.S. experience is exceptional and other countries will not necessarily follow the same "rules" as above.

MG

Stocks always outperform bonds in every economic climate except during "deflation".  But what matters more for what duration risk one should take in bonds is whether the trend of interest rates is falling or rising, not the absolute level per se, again except during "deflation".

Doesn't the statement above need some qualification?

It seems to me that there have been multi-decade periods where your statement is not accurate, and these periods were not necessarily deflationary. 

Sometimes simply buying stocks at the wrong time can cause bonds to outperform for many years.

As I recall, Japan has seen mostly flat inflation for the last 20+ years (there hasn't been much actual deflation), and yet the Nikkei is 75% off of levels it was at 23 years ago.  Where are the superior stock returns compared to bonds in Japan? 

[Greg]

But that's kind of the point.  If you had a lot of money, you'd probably be better off taking a small negative return than you would be storing it in the Himalayas.  

Dude, I have a drop bear.  This eliminates virtually all risk of theft.  :)

Reading through past info stated by wise sages:

I know this is an old thread but I couldn't stop laughing after I googled what a drop bear was. Hilarious stuff.

I worry about these lowering rates as well for starting my PP but the future is unknown and that seems to be kinda the point for this portfolio. I'll have to see what my gut tells me though about if LTT keep dropping. Perhaps one of the posts in here of a variety of discontinuous jumps like a piecewise function that if LTT is between 2.5 and 3.0% to have 30% in STT and 20% in LTT. Or something like that.

[sophie]

Seems to me that the risk is progressively greater than the benefit as the yield approaches zero. So I'm thinking something more like:
>4%  25% LTTs
3.5-4  20% LTT/5%STT
3-3.5  15% LTT/10%STT
2.5-3  10% LTT/15%STT
2-2.5  5% LTT/20% STT
<2      25% STT

and

I worry about these lowering rates as well for starting my PP but the future is unknown and that seems to be kinda the point for this portfolio. I'll have to see what my gut tells me though about if LTT keep dropping. Perhaps one of the posts in here of a variety of discontinuous jumps like a piecewise function that if LTT is between 2.5 and 3.0% to have 30% in STT and 20% in LTT. Or something like that.

Well.  Here we are, with 30 year treasuries at around 2.8%, and the yield curve is flattening (short term rates have stayed stable).  The question isn't so much how far treasury yields can theoretically drop, but whether treasuries are beginning to hit a ceiling that prevents them from fulfilling their role in the PP.  They haven't been able to counter the recent drops in stock and gold (painful to watch by the way).  Negative yields for 30 year bonds would truly be an insane scenario.  Yes, German bonds are going negative but only short term (2 years or less).  If I were trying to protect large amounts of money I think the last thing I'd want to do it buy a long term bond with negative rates, that I'd have to get someone else to buy if I wanted to get rid of it in a rising interest rate environment. 

I am probably not qualified to make this kind of judgment, but I think that I would likely start putting new money into short term bonds instead of long bonds if the yields go below 2.5%.  It would be nice to do that while short term rates (1-5 years) are still positive.  Is anyone else considering this?

[AdamA]

They haven't been able to counter the recent drops in stock and gold (painful to watch by the way).  

LTT's are up roughly 30% over the past year whereas gold and stocks have been roughly flat (since May 2011), so I'd say that LTT's are functioning as advertised.  

If I were trying to protect large amounts of money I think the last thing I'd want to do it buy a long term bond with negative rates, that I'd have to get someone else to buy if I wanted to get rid of it in a rising interest rate environment. 

But, if you believe interest rates on LTT's can become negative, what's to say they can't become more negative.  Maybe you'd make a lot of money as rates dropped from -2 to -3%...who knows.

[Alanw]

If inflation is at -4%, whose to say LTT's aren't a good value at -2%.  Anybodys guess!

[Tortoise]

At lower yields, long dated treasury's are likely to be more volatile, and potentially zigzag around in price more wildly than at higher yield levels...

Is there a good explanation for that effect, other than based on what has happened in the past? I double-checked this using Yahoo Finance data for ^TYX going back to 1977, and sure enough, the standard deviation of the daily movements in both yield and bond price (expressed as a % of the previous day's value) has increased as yield has declined over the years.

I'm interested in knowing why this effect occurs. Is it mathematical, psychological, or unknown?

Perhaps a ballpark guide might be to x10 the yield to identify the long duration holding. At 1% look to hold 10 year maturity, at 2% use 20 years, at 3% go for 30 years, 4% 40 years, 5% 50 years, at 6%+ undated.

Extending the left tail, at 0% use short term T's, at -1% look to start shorting 10 year T's...

Would a strategy like that effectively be attempting to keep the volatility constant (or as constant as possible) as yield changes?

[sophie]

Clive, thanks for a very informative post.

The increased volatility at low yields is interesting indeed - I'd have predicted the opposite on the assumption that if you're nearing the floor, there's less room for price changes.  Guess I should remember what the PP is all about and stop predicting