30 year or ladder

[Lonestar]

Based on recent posts I'm going to sell my TLT and buy treasuries either through TD or Fidelity.  I can see advantages to each.  Fidelity for convenience, TD for lack of counter party risk.

If one were to buy through Fidelity (on the secondary market), a 25 to 30 year maturity 5 year ladder could be used.  It appears that the standard way to rebalance is to buy 30 year maturities and sell at the 20 year mark.  Has there been any backtesting on setting up a 5 year ladder and selling the shortest maturity each year (reinvesting in a 30 year maturity).  Seems like this would provide the benefits of the long bond, yet provide some safety in case of a rise in rates. 

[MediumTex]

I am pretty sure that this exercise would not be worth the trouble.

As much as possible, we want all the buying and selling in the PP to be driven by rebalancing events.  That's where we get our gains.

With that said, I'm pretty sure the plan you are describing wouldn't hurt anything (assuming you were still selling when a bond hit 20 years to maturity), but it sounds to me like a more complicated path to the same destination that the basic HB PP recipe allows you to reach with great simplicity.

[Lonestar]

I am pretty sure that this exercise would not be worth the trouble.

As much as possible, we want all the buying and selling in the PP to be driven by rebalancing events.  That's where we get our gains.

With that said, I'm pretty sure the plan you are describing wouldn't hurt anything (assuming you were still selling when a bond hit 20 years to maturity), but it sounds to me like a more complicated path to the same destination that the basic HB PP recipe allows you to reach with great simplicity.

The plan would never have a bond with 20 years to maturity.  The shortest bond on the ladder would be a 25 year maturity and at the end of that year you would sell it and buy a 30 year maturity.  One sale/one buy per year.  Really not very complicated.

Hope I'm explaining this well enough.  You would always hold 5 bonds of equal amounts with maturities of 25 years through 30 years.  A classic bond ladder.

It would provide most of the benefits of a single 30 year bond but would offer some protection against interest rate increase.

[MediumTex]

I am pretty sure that this exercise would not be worth the trouble.

As much as possible, we want all the buying and selling in the PP to be driven by rebalancing events.  That's where we get our gains.

With that said, I'm pretty sure the plan you are describing wouldn't hurt anything (assuming you were still selling when a bond hit 20 years to maturity), but it sounds to me like a more complicated path to the same destination that the basic HB PP recipe allows you to reach with great simplicity.

Why not just buy all 30 year bonds and sell them when they all reached 25 years?

As far as protecting against rising rates you have that elsewhere in the portfolio.  There is no need to make any changes to the long term bond allocation to prepare for that contingency.

If you are setting up a system that requires you to sell one bond a year and buy one bond a year, then if you have capital gains in a long term bond you will be forced to recognize it when you could have easily waited years under the basic HB PP recipe before recognizing the same gain.

The plan would never have a bond with 20 years to maturity.  The shortest bond on the ladder would be a 25 year maturity and at the end of that year you would sell it and buy a 30 year maturity.  One sale/one buy per year.  Really not very complicated.

Hope I'm explaining this well enough.  You would always hold 5 bonds of equal amounts with maturities of 25 years through 30 years.  A classic bond ladder.

It would provide most of the benefits of a single 30 year bond but would offer some protection against interest rate increase.

Why not just buy all 30 year bonds and sell them when they all reached 25 years?

As far as protecting against rising rates you have that elsewhere in the portfolio.  There is no need to make any changes to the long term bond allocation to prepare for that contingency.

If you are setting up a system that requires you to sell one bond a year and buy one bond a year, then if you have capital gains in a long term bond you will be forced to recognize it when you could have easily waited years under the basic HB PP approach before recognizing the same gain.

I don't think doing this will do any harm, but between the potential for unnecessary recognition of capital gains and updating the long term treasury ladder every year (in addition to whatever changes may have been occasioned by hitting a rebalancing band), I wouldn't do it.

[rickb]

I think there might be a mistaken notion here that it's beneficial somehow to rollover into a new bond if interest rates go up.  An example might help.  Say the 25 year bond in the ladder is paying a 3% dividend.  If interest rates go up to 4%, if you sell this bond you'll replace it with a 30 year bond paying 4%.  If the face value of the 3% bond is $10,000 it pays $300/year.  If prevailing rates for long bonds increase to 4% this $10,000 bond will only be worth about $7500 (slightly more, but it will be very close to this amount).  If you sell this bond and replace it with a $7500 bond paying 4% you're still getting $300/year.  The point is the value of long bonds changes enough to keep the dividend payments about the same.  If interest rates are going down, this means the value of the bonds goes up - but if interest rates go up, the value of the bonds goes down (enough to keep the dividend payment the same).

If the bonds are held in a taxable account you might do this to realize a capital loss, but it doesn't change the value of the bonds you're holding or the total dividend payment you're getting.  If the bonds are held in a non-taxable account, there's really no point.

[TripleB]

Hope I'm explaining this well enough.  You would always hold 5 bonds of equal amounts with maturities of 25 years through 30 years.  A classic bond ladder.

It would provide most of the benefits of a single 30 year bond but would offer some protection against interest rate increase.

Yes it would offer some protection against rising interest rates, at the cost of reduced protection against lowering interest rates. Since the long term bonds in the PP are specifically used to provide protection against deflation, I want the bonds to be as long as they can be. If there were 40 year US treasuries, I would buy them.

If there were no transaction costs, I'd sell all my 30 year bonds at the 29 year mark and buy new 30 year ones.

If you're worried about rising interest rates, then instead of going through the hassle of creating a complicated bond ladder, just change your allocation from 25% bonds, 25% cash to 24% bonds and 26% cash. It would likely perform nearly identical to a 5 year bond ladder, which is to say, virtually no different from 25/25 for better or for worse.

[Lonestar]

Thanks for the input.  All very good ideas and once again, tinkering with the PP probably defeats it's overall usefulness.

I can see now why the idea would be of no benefit even if held in a tax deferred account.  Like most, I guess I'm just too predisposed to "looking for a better way".

[6 Iron]

What about holding 50% in PLW 1 to 30 year Treasury Ladder ETF as both the cash and LTT holdings.

http://www.invescopowershares.com/produ ... W#freqdist

I have thought about this, as the simplicity is appealing. Downsides include a total asset volume of only 280+ million with an occasional non-trivial bid/ask discrepancy, an expense ratio of 0.25 percent, and the third party issue. It has an effective duration of 11 years... For me, though, the biggest issue is psychologically the joy of having a dedicated cash pool available to buy an asset that is having a fire sale.

[D1984]

What about holding 50% in PLW 1 to 30 year Treasury Ladder ETF as both the cash and LTT holdings.

PLW (and the Ryan 1-30 year laddered Treasury index for 2006-2007 since it is an approximation of the same concept and predates the ETF by almost two years) does seem to perform fairly similarly on average to a 50/50 mix of LTTs and STTs. My only concern would be that in a rapidly rising yield environment like the late 60s, early 80s, or 1994 it might overall underperform the LTT/STT barbell. If rates rise and you hold the standard PP mix of STTs/cash and LTTs then the former can act as "dry powder" for rebalancing/buying other assets (since STTs/cash won't be affected much or will even be helped by a rise in rates while the LTTs get hurt by it...but it is this loss (as yields go up, prices of existing LTTs go down) in the LTT portion that lets you rebalance into them "on the cheap" and buy them when they are down. Holding an ETF that has 1/30th of its maturity in each year and only rebuys one 1/30th of itself (as the one-year bond matures, the 30 year bond becomes a 29-year bond, the 29-year bond becomes a 28-year bond, etc, and the fund buys another 30-year bond at current rates with the cash proceeds from the maturing one-year bond) each year might not allow you to get the full effect of volatility capture rebalancing gains; you might get stuck with an ETF that has most of its assets locked up in 1 to 29 year bonds paying (on original purchase price) from 0% (on the short end of the maturity spectrum) to around 3% (on the long end) when current bonds-both LT and ST-are paying much more.

To be fair, the a hypothetical situation like the one above might not turn out too badly since the ST and LT bonds that had the low yields would be "repriced" at current market yields when/if rates rose (even though the ETF would only be buying/selling about 1/30th of its bonds each year they would still lose value as rates rose and this should theoretically be reflected in the NAV and price of the ETF which would both drop which would (hopefully) trigger rebalancing bands and let you buy the whole ETF "on the cheap." I just don't know if this would be enough to offset the volatility capture you'd lose by not holding the LTTs and STTs separately.

You might even combine stocks and gold by holding 50% in GDX

Unless I am misunderstanding you (please correct me if so) you are saying hold GDX (gold miners) as 50% of the PP in lieu of 25% gold and 25% stocks. BAD IDEA! Check out NEM, ABX, GFI, INIVX, SGGDX, and USAGX vs SPY or VFINX from 1/1/1995 to 12/31/1999 (the late 90s bull market) and you'll see exactly what I mean. The total stock market portion of the PP carried it during most of this time and gold stocks (at least in USD) got killed. Were things any different in pounds sterling?

As a somewhat "leveraged to the price of gold" (except in extreme deleveraging situations like 2008) replacement for a SMALL portion of the gold section of the PP gold stocks might be considered OK but IMO replacing all or most of the regular stock index (or even a mixture of the stock and gold portions....these miners got killed worse than the price of gold in the late 90s as gold prices fell roughly 25%) with them isn't a useful idea.

Non-gold mining stocks (in the form of the 25% of the PP that is dedicated to a TSM index) are there for one reason-to give the PP good returns in times of non-inflationary prosperity. Gold is there because it shines in times of stress, currency devaluation, hyperinflation, high inflation (at least when real rates are lower than CPI...like most of the 1970s), and moderate inflation (circa 2001 to present) when real rates on cash are negative most of the time and thus the opportunity cost of owning "cash" (one form of money) vs "gold" (another form of money) are skewed in gold's favor. The two are there for very different purposes and thus conflating them by owning gold stocks in place of both seems rather unwise.

[beafet]

What about something like 45% PLW, 5% EDV, 25% VTI, and 25% IAU??