The challenge of an EU PP at current bond yields

[belgo]

Hello everyone,

After some frantic reading during February, I took the plunge last month and started a EU PP.  I can't complain as it is 2% up already so far.
But my worry is: am I taking unreasonable interest rate risk?  The correct implementation of the bond leg of the European PP involves investing 25% of your assets into the 30y Bund currently yielding 0.89%.  I did exactly that and tried to persuade myself that this is what the theory says: no market timing, buy all assets at the same time, even if there is one you hate and even if you are convinced is a bad investment.  I have also been mailing with Barrett on this forum who was so kind to show a discussion on this in which Craig himself, the co-author of "the book", says that for him there is no point to hold the bond part at interest rates below 1%.  So now I don't know anymore:-)  The bond has a 23 y duration so interest rate is huge in case of a rate rise.  And there does not seem to be much more upside left although it has to be said that one year ago this bond even traded briefly at around 0.5% yield (a 10% price difference).  But on the other hand, the difference with cash at 0% and a 30y bond at 0.89% is not a lot of potential juice.  So I guess my question is, should I reduce the duration of the bond part or reduce it to 20% given that you could argue that this + cash already gives me enough deflation protection.  Or should I keep the bonds and pray there is no sudden jump and only throw them out ic case they reach 0.5%?  But Harry was still alive when the Japanese situation also had long term yields below 1.5% and I never read that the "theory" should be changed in case we have  historically very low interest rates.    All advice welcome.  Best regards.
 

[barrett]

Thanks for posting, Patrick.

I'm no authority on Harry Browne but I believe that he developed the PP concept so that it could, at least in theory, work pretty much anywhere.

One thing I keep coming back to when thinking about whether or not an asset class is fairly valued - at which point, in my mind, at least, there is an equal chance that it will go up or down - is that there's an equal amount of sentiment on both sides of the bet. It's the old argument that assets are "always fairly valued". In your case, you are essentially taking 25% of your assets and "betting" that inflation over the next 30 years or so could be extremely low or that we could have outright deflation. Given that worldwide growth seems to be slowing, that would seem to be a possibility. The other thing is that your "prediction" doesn't need to hold true for 30 years. Even if there is extremely low to negative growth in Europe for the next 5-10 years, those German Bunds could move up in price nicely and you could reap some rewards on rebalancing.

I believe that narrower rebalancing bands are necessary when rates are so low because it's harder for bonds to move up to 35% of a portfolio's value. But I'd be happy to have people poke holes in that idea.

[MachineGhost]

Hello everyone,

After some frantic reading during February, I took the plunge last month and started a EU PP.  I can't complain as it is 1.5% up already so far.
But my worry is: am I taking unreasonable interest rate risk?  The correct implementation of the bond leg of the European PP involves investing 25% of your assets into the 30y Bund currently yielding 0.89%.  I did exactly that and tried to persuade myself that this is what the theory says: no market timing, buy all assets at the same time, even if there is one you hate and even if you are convinced is a bad investment.  I have also been mailing with Barrett on this forum who was so kind to show a discussion on this in which Craig himself, the co-author of "the book", says that for him there is no point to hold the bond part at interest rates below 1%.  So now I don't know anymore:-)  The bond has a 23 y duration so interest rate is huge in case of a rate rise.  And there does not seem to be much more upside left although it has to be said that one year ago this bond even traded briefly at around 0.5% yield (a 10% price difference).  But on the other hand, the difference with cash at 0% and a 30y bond at 0.89% is not a lot of potential juice.  So I guess my question is, should I reduce the duration of the bond part or reduce it to 20% given that you could argue that this + cash already gives me enough deflation protection.  Or should I keep the bonds and pray there is no sudden jump and only throw them out ic case they reach 0.5%?  But Harry was still alive when the Japanese situation also had long term yields below 1.5% and I never read that the "theory" should be changed in case we have  historically very low interest rates.    All advice welcome.  Best regards.

The PP works on a rough approximation of having equal risk contribution from three risk assets and capturing the volatility, momentum and mean reversion among them with the rebalancing bands (which is the "market timing").  It has nothing at all to do with the level of yields.  If you start deviating from equal risk contribution by adjusting duration without adjusting the relative portfolio weight, you're breaking the synchronization.  You do realize a 23-year duration has probably around a 17.5% annual rate of volatility?

Lets say German Bunds go totally belly up which is actually a realistic probability since Germany isn't a currency issuer (doesn't have any fiscal power to create money via fiat).  That 25% of your PP goes completely to -100%.  What offsets it?  Stocks and gold.  Cash is a trickier question.  It really should be in USD and not local currency.  The Euro is not safe but how you manage that and local spending is beyond my purview.

[MachineGhost]

I believe that narrower rebalancing bands are necessary when rates are so low because it's harder for bonds to move up to 35% of a portfolio's value. But I'd be happy to have people poke holes in that idea.

Pure rubbish!  See this: https://en.wikipedia.org/wiki/Bond_convexity

[economicsjunkie]

I believe that narrower rebalancing bands are necessary when rates are so low because it's harder for bonds to move up to 35% of a portfolio's value. But I'd be happy to have people poke holes in that idea.

Pure rubbish!  See this: https://en.wikipedia.org/wiki/Bond_convexity

No it's not rubbish! There is less room upwards the lower rates go.

Yes, the bonds become more sensitive to rate changes, but on the other hand the rates can't change as much anymore. Take the extreme example where rates are 0.01% How much do you gain from a move to 0%? Maybe 0.3% or so? Whopeee!

Unless of course you assume rates for 30+yr bonds can go negative to a significant degree and comfortably stay negative for decades to come. I'll believe that when I see it.

[barrett]

I believe that narrower rebalancing bands are necessary when rates are so low because it's harder for bonds to move up to 35% of a portfolio's value. But I'd be happy to have people poke holes in that idea.

Pure rubbish!  See this: https://en.wikipedia.org/wiki/Bond_convexity

I understand the general idea behind bond convexity. But I believe I am correct that a 30-year bond yielding .89% has little chance of breaching a 35% band. It's not an impossible event, merely a low-probability one.

belgo, do you hold individual bonds or a bond fund? If it's individual bonds, their duration will drop over time (as opposed to a fund that targets a certain steady duration). So you could try dialing back your bond allocation to 20% or so, but you would need to keep an eye on the duration of the bonds. I believe what MG is alluding to in his first post is that your bonds may be more volatile than your stocks and gold right now.

[belgo]

Thanks a lot for the replies guys.

Barrett, I specifically opened an account at a German broker in order to be able to buy the bond directly.  It is a crazy situation here with a lot of the bonds already at negative yields. And I am wondering to which extent it is artificial as the ECB is buying billions of bonds every month as part of its efforts to kickstart the economy. But some day they will stop buying.

About the 35% band being a small probability, am I correct that it is maybe not that remote as the bonds traded at 0.5% yield or 10% higher a year ago?

About Germany not being a good credit risk..?  I think it's as safe as it gets in euro and even if Euro would collapse the bunds would become Deutschmark again I guess. 

For the gold section I am in an ETF but am looking at a company called bullionvault in the UK as a physical alternative. But somehow sending a lot of money to such a company that will hopefully do what it claims on its website makes me hesitate as well (e.g. fraud risk).

[MachineGhost]

I understand the general idea behind bond convexity. But I believe I am correct that a 30-year bond yielding .89% has little chance of breaching a 35% band. It's not an impossible event, merely a low-probability one.

belgo, do you hold individual bonds or a bond fund? If it's individual bonds, their duration will drop over time (as opposed to a fund that targets a certain steady duration). So you could try dialing back your bond allocation to 20% or so, but you would need to keep an eye on the duration of the bonds. I believe what MG is alluding to in his first post is that your bonds may be more volatile than your stocks and gold right now.

Oh, you're thinking about a lower zero-bound; I'm thinking about volatility.  It's been discussed elsewhere that there is no lower zero bound, so it won't cap the bond returns.  Yields can easily go negative -- there's nothing stopping that.  In the PP we care less about the reasons for the level of yields and more that there's room to be volatile up or down.

If you're concerned about the level of yields, then you're concerned about "valuation" and that gets into "market timing" issues.

[MachineGhost]

Unless of course you assume rates for 30+yr bonds can go negative to a significant degree and comfortably stay negative for decades to come. I'll believe that when I see it.

You're making an a assumption that investors are going to act like 0% is some kind of hard limit for price moves.  It hasn't for 0-15 year bonds, why would 20, 30 or 40 be some kind of exceptional exception?  40 year is already at .33% and dropping fast.  If your government is completely FUBARed, there must be some really awesome level of optimism that at least one person thinks theres still going to be .33% worth of economic growth over the next 40 years.

[belgo]

The Swiss 10y bond yields -0.34% at the moment... but does that mean one expects the Swiss economy to be depressed for years to come? It's a very strong economy only suffering from a currency that is overvalued.

[economicsjunkie]

You're making an a assumption that investors are going to act like 0% is some kind of hard limit for price moves.

There has never been a 30yr bond that has gone negative as far as I know. So I am saying it is unlikely to happen. That doesn't mean it's impossible, nor does it matter for the purposes of this discussion what's possible or impossible as much as what's likely or unlikely.

The question is how much punch can one reasonably expect from bonds in the European PP at that stage. That's quite frankly something I too would be concerned about if I were in Europe. In fact, since none of the ECU governments have the power of sovereign currency issue (unlike UK, Japan, US, Switzerland, etc.), I would also be somewhat concerned about credit risk.

It hasn't for 0-15 year bonds, why would 20, 30 or 40 be some kind of exceptional exception?

Because yields on longer bonds are usually higher than yields on shorter bonds, barring some temporary exceptions.

[economicsjunkie]

The Swiss 10y bond yields -0.34% at the moment... but does that mean one expects the Swiss economy to be depressed for years to come? It's a very strong economy only suffering from a currency that is overvalued.

It could just mean that there is a huge demand for Swiss net financial assets by international and domestic investors and the government isn't satisfying this demand, unless -0.34% is their desired policy rate for 10 year bonds, in which case there's technically no problem here, OR the people in charge have no clue what they are doing and what any of what I just wrote means. (Although I prefer to leave the possibility that people in charge are completely incompetent as last resort only )

Excess reserves in the Swiss banking system might give you a clue.

[barrett]

About the 35% band being a small probability, am I correct that it is maybe not that remote as the bonds traded at 0.5% yield or 10% higher a year ago?

Well you can run a few "what if?" scenarios to really grasp what has to happen to trigger a rebalance with the 15/35 bands. Say, for example, you start with $100,000 and you set up a 4X25. If bonds move up 40% AND nothing else changes, you then have $35,000 in bonds and $25,000 in each of the other three assets for a total of $110,000. 40% would be a massive bond move but your bond allocation would still be at slightly less than 32%. Obviously you can run a million scenarios like this (bonds go up X amount, stocks go down X amount), but you get the idea.

Man, has anyone checked out what's going on in Japan? Here's a good link:

http://www.investing.com/rates-bonds/ja ... bond-yield

So, less than two months ago, the Japanese long bond was about where the German Bund is now. The yield has plunged to .292 in less than 60 days. Can one of you bondy, mathy types run some numbers on what kind of bond returns a Japanese investor would have gotten so far this year? If you don't want to look at the link, the numbers for the 30-year Japanese bond are:

12/31/15 - Yield of 1.297%
4/20/16 - Yield of .292%

Looks like we got ourselves some good old bond convexity goin' on over there! (I can see Alaska from my window here in CT! ))

[Pet Hog]

So, less than two months ago, the Japanese long bond was about where the German Bund is now. The yield has plunged to .292 in less than 60 days. Can one of you bondy, mathy types run some numbers on what kind of bond returns a Japanese investor would have gotten so far this year? If you don't want to look at the link, the numbers for the 30-year Japanese bond are:

12/31/15 - Yield of 1.297%
4/20/16 - Yield of .292%

Using this bond price calculator I get a return of 28.83% for a bond with a 30-year maturity (face value = $100; coupon = $1.297; periods = 30 years, annual; yield = 0.292%).  Seems intuitive: for an extra 1% yield spread out over 30 years you might expect a roughly 1 x 30 = 30% rise in bond value.

[belgo]

OK. My conclusion for the moment is that the German bund yield can still go lower and that I should stick to the PP theory. But psychologically it will be tough, it can easily lose 20% in value with only a modest interest rate increase. Not sure the other assets will go up 20% when that happens.

[MachineGhost]

[MachineGhost]

OK. My conclusion for the moment is that the German bund yield can still go lower and that I should stick to the PP theory. But psychologically it will be tough, it can easily lose 20% in value with only a modest interest rate increase. Not sure the other assets will go up 20% when that happens.

No, the assets don't all respond at the same time.  The worst underwater period for the PP was 108-weeks in a row, so be forewarned.

I do think the "valuation" problem with bonds is something worth working on.  But a lower duration doesn't have enough volatility even if you increase the weight to come close to matching the volatility of stocks and gold.  It would have to be leveraged somehow.  Much more work needs to be done on duration vs volatility.  i.e. Is a low duration, high net volatility the same as high duration, high volatility?

[economicsjunkie]

So, less than two months ago, the Japanese long bond was about where the German Bund is now. The yield has plunged to .292 in less than 60 days. Can one of you bondy, mathy types run some numbers on what kind of bond returns a Japanese investor would have gotten so far this year? If you don't want to look at the link, the numbers for the 30-year Japanese bond are:

12/31/15 - Yield of 1.297%
4/20/16 - Yield of .292%

Using this bond price calculator I get a return of 28.83% for a bond with a 30-year maturity (face value = $100; coupon = $1.297; periods = 30 years, annual; yield = 0.292%).  Seems intuitive: for an extra 1% yield spread out over 30 years you might expect a roughly 1 x 30 = 30% rise in bond value.

Bear in mind that the correct bond to buy for the Japanese PP is not the 30 but the 40 year government bond (http://www.investing.com/rates-bonds/ja ... ming-chart) which would have yielded more like 42% in that same period.

[barrett]

Thanks for running the numbers, guys. Isn't a 40-year bond actually TOO volatile for the PP? Or, put another way, isn't an allocation of 25% too high when rates are this low? It seems to me that the bond quadrant in Japan and Europe currently has more volatility than the stock and gold quadrants. Folks have written on here about not wanting to hold EDV for that reason. Ryan Melvey proposed that one could just hold a lower percentage in bonds when duration is higher.

It's really fascinating to see this Japanese bond move in real time. Bonds with low yields really can lift a portfolio. But I would still go with narrower rebalancing bands.

[Ad Orientem]

Although in theory the PP can work with any reasonably good sized economy, there are some weak spots. First, gold is geared to the US Dollar, not the Euro, Pound, or Yen. So if your local non USD currency sees a sudden spike in inflation, gold may not react in the violent way that it does when the Dollar is weak. And secondly the PP was designed to work in an environment where there is sovereign control over the currency. This is not the case in the Euro Zone.

I am not going to say that the PP won't offer you some protection in Europe, but I don't have the same level of confidence in it that I do with a US Dollar PP.

[drumminj]

So if your local non USD currency sees a sudden spike in inflation, gold may not react in the violent way that it does when the Dollar is weak.

If measured in USD, sure.  But in local currency you should.  And presumably local currency is what one is most concerned with?

[MachineGhost]

Thanks for running the numbers, guys. Isn't a 40-year bond actually TOO volatile for the PP? Or, put another way, isn't an allocation of 25% too high when rates are this low? It seems to me that the bond quadrant in Japan and Europe currently has more volatility than the stock and gold quadrants. Folks have written on here about not wanting to hold EDV for that reason. Ryan Melvey proposed that one could just hold a lower percentage in bonds when duration is higher.

It's really fascinating to see this Japanese bond move in real time. Bonds with low yields really can lift a portfolio. But I would still go with narrower rebalancing bands.

Generally, the longer the duration, the higher the volatility.  But does a 30-year or 40-year bond with a yield of .29% or 33% have a lot of duration?  I rightly don't know -- never calculated it.  Let me go do it....  Wow, a 30-year JGB has an effective duration of about 29 years and a 40-year is 37.5 years!  So you are correct, 25% would be way too much since we want to more or less have the same volatility contribution coming from all the assets in the PP.  The question is: is that true?

This could save the correlated risk parity concept from irrelevance because we do know that durations don't stay fixed.  T-Bonds had 6.89x less duration in 1980 than now and 12.24x more duration in 1945.  But does that translate into higher or lower volatility?  My gut feel says that volatility is extrinsic and that may or may not match the duration.  Inflation expectations in 1980 were undoubtedly out of this world despite the low duration, so I would expect a seriously high level of volatility.  But I've never looked at the data to see.

P.S.  EDV has about a 25yr duration.

[Ad Orientem]

So if your local non USD currency sees a sudden spike in inflation, gold may not react in the violent way that it does when the Dollar is weak.

If measured in USD, sure.  But in local currency you should.  And presumably local currency is what one is most concerned with?

Actually what you will see is a movement based on the depreciation of your local currency relative to the dollar. If it is a major currency like the Euro, yes, I would expect some upward movement. But again, it is unlikely to be as violent as it would be if the dollar was showing weakness.

[economicsjunkie]

Thanks for running the numbers, guys. Isn't a 40-year bond actually TOO volatile for the PP? Or, put another way, isn't an allocation of 25% too high when rates are this low? It seems to me that the bond quadrant in Japan and Europe currently has more volatility than the stock and gold quadrants. Folks have written on here about not wanting to hold EDV for that reason. Ryan Melvey proposed that one could just hold a lower percentage in bonds when duration is higher.

It's really fascinating to see this Japanese bond move in real time. Bonds with low yields really can lift a portfolio. But I would still go with narrower rebalancing bands.

For the risk free bonds to deliver the most punch more volatility is precisely what you WANT in the PP!!

EDV is not just a longer duration investment, it's different in other aspects, such as tax treatment, so there are other reasons why it's not desirable in the PP as far as I know.

[belgo]

I read that the EU PP actually yielded +2.73% last year with inflation at 0% with hardly any change in the value of the 30 bund whereas the US PP was negative.  It shows that the two do not need to move in parallel as the economies are not in parallel either.