Something’s Always Going Up

[jk]

Re: Something’s Always Going Up

a different slant on browne allocation:

if you have a tax sheltered account use zero coupon treasuries. if, e.g., gary shilling's prediction of a 3% 30year bond were to come to pass, 30yr bonds purchased now would produce a total return of 50%, while 30 yr zeroes would produce 80%. if you believe that inflation is the more likely outcome, you can allocate a smaller portion of your capital to zeroes than you would to 30yrs tlt's, and then a larger amount to e.g. pm's. part of browne's notion was to have investments leveraged to different scenarios, with the idea of losing less on the losers than you gain on the winners.

[Finster]

Re: Something’s Always Going Up

a different slant on browne allocation:

if you have a tax sheltered account use zero coupon treasuries. if, e.g., gary shilling's prediction of a 3% 30year bond were to come to pass, 30yr bonds purchased now would produce a total return of 50%, while 30 yr zeroes would produce 80%. if you believe that inflation is the more likely outcome, you can allocate a smaller portion of your capital to zeroes than you would to 30yrs tlt's, and then a larger amount to e.g. pm's. part of browne's notion was to have investments leveraged to different scenarios, with the idea of losing less on the losers than you gain on the winners.

Indeed, you can think of the zeroes as a kind of supercharged long bond. Basically they are a bond of thirty year duration, while a treasury with a maturity of thirty years would have (depending on the rate), say, a duration of ten or so years. In rough terms, an allocation of 25% cash and 25% thirty year treasuries would be similar to having 35% cash and 15% thirty year zeroes.

But there is no free lunch here. If you see inflation as the bigger risk, either way you'd presumably allocate more to hard money and less to the bonds.

[Jim Nickerson]

Re: Something’s Always Going Up

Indeed, you can think of the zeroes as a kind of supercharged long bond. Basically they are a bond of thirty year duration, while a treasury with a maturity of thirty years would have (depending on the rate), say, a duration of ten or so years. In rough terms, an allocation of 25% cash and 25% thirty year treasuries would be similar to having 35% cash and 15% thirty year zeroes.

But there is no free lunch here. If you see inflation as the bigger risk, either way you'd presumably allocate more to hard money and less to the bonds.

Finster,

There is something important in the terminology you are using, which I am missing. I am not grasping how a 30 year bond with some rate translates into a duration of 10 or so years. I would say I understand "duration," but I am not understanding it here. Do you mind educating me?

[Finster]

Re: Something’s Always Going Up

Finster,

There is something important in the terminology you are using, which I am missing. I am not grasping how a 30 year bond with some rate translates into a duration of 10 or so years. I would say I understand "duration," but I am not understanding it here. Do you mind educating me?

A full treatment would require some fairly heavy math, but the basic idea is simple. A thirty year bond pays interest throughout its life time of thirty years, and then returns the principal at the end of its life. The thirty year period is the "maturity" of the bond, because that's when the bond matures. The term "duration" reflects a sort of weighted average of how long it takes to get the total amount of money from the bond. Since you receive regular payments during the life of the bond, the duration is shorter than the maturity. It reflects not just the nominal lifetime of the bond, but also the fact that you receive an income stream long before the maturity date.

This is why the "duration" of a zero-coupon thirty year bond equals its maturity of thirty years. There are no payments before the thirty year period expires. Everything comes all at once at the end of the thirty year period.

Duration is a useful concept because it distills the total effect of the time of the payments of the bond into a single number. Two bonds of the same maturity may have different price changes in response to a change in interest rate. But two bonds of the same duration, under the same change of interest rate, will change in price the same amount.

[Jim Nickerson]

Re: Something’s Always Going Up

A full treatment would require some fairly heavy math, but the basic idea is simple. A thirty year bond pays interest throughout its life time of thirty years, and then returns the principal at the end of its life. The thirty year period is the "maturity" of the bond, because that's when the bond matures. The term "duration" reflects a sort of weighted average of how long it takes to get the total amount of money from the bond. Since you receive regular payments during the life of the bond, the duration is shorter than the maturity. It reflects not just the nominal lifetime of the bond, but also the fact that you receive an income stream long before the maturity date.

This is why the "duration" of a zero-coupon thirty year bond equals its maturity of thirty years. There are no payments before the thirty year period expires. Everything comes all at once at the end of the thirty year period.

Duration is a useful concept because it distills the total effect of the time of the payments of the bond into a single number. Two bonds of the same maturity may have different price changes in response to a change in interest rate. But two bonds of the same duration, under the same change of interest rate, will change in price the same amount.

Good, Finster, I think I grasp now what duration means in this context, IF that underlined above means when all the interest payments add up to the amount loaned (paid) for the bond, and I think I know bond math is complicated, I just don't know (nor actually care, you know like ignorance and apathy) how complicated.

[jk]

Re: Something’s Always Going Up

Good, Finster, I think I grasp now what duration means in this context, IF that underlined above means when all the interest payments add up to the amount loaned (paid) for the bond, and I think I know bond math is complicated, I just don't know (nor actually care, you know like ignorance and apathy) how complicated.

not quite right, jim. the total amount received is what is referenced, not the amount paid. this is more than the total paid. assume you purchase the bond at par, i.e. a $100 bond with rate of return x% and maturity m years, purchased for $100. you will receive x dollars per year til maturity, and then your $100 at the time of maturity, for a total = 100+mx.

the notion of duration is based on the fact that at the end of year 1 you've received $x, at the end of year 2 you've received a total of $2x and so on. at the end of year m, you've received all you're going to get, which is $mx+100. now go back and notice that at the end of year 1 you've received a fraction of your total return which equals x/(mx+100). etc.

[Jim Nickerson]

Re: Something’s Always Going Up

not quite right, jim. the total amount received is what is referenced, not the amount paid. this is more than the total paid. assume you purchase the bond at par, i.e. a $100 bond with rate of return x% and maturity m years, purchased for $100. you will receive x dollars per year til maturity, and then your $100 at the time of maturity, for a total = 100+mx.

the notion of duration is based on the fact that at the end of year 1 you've received $x, at the end of year 2 you've received a total of $2x and so on. at the end of year m, you've received all you're going to get, which is $mx+100. now go back and notice that at the end of year 1 you've received a fraction of your total return which equals x/(mx+100). etc.

I hate to beat dead horses or anything, but I am still not clear on this.

Take a 30 year bond of $100 @ 10%. Each year one would receive $10 in interest. After 10 years one would have then received $100, which equals the amount loaned by buying the bond. Would the "duration" of said bond then be 10 years, and I understand that one would still receive 20 more years of interest plus the original (no doubt badly inflated) $100 at the end of the 30th year. If what I just wrote is the same as you explained, then to me this is clearer, god, I hope this is correct. I do not see in your explanation where the concept of "duration" was explained. Thank you, jk.

[jk]

Re: Something’s Always Going Up

I hate to beat dead horses or anything, but I am still not clear on this.

Take a 30 year bond of $100 @ 10%. Each year one would receive $10 in interest. After 10 years one would have then received $100, which equals the amount loaned by buying the bond. Would the "duration" of said bond then be 10 years, and I understand that one would still receive 20 more years of interest plus the original (no doubt badly inflated) $100 at the end of the 30th year. If what I just wrote is the same as you explained, then to me this is clearer, god, I hope this is correct. I do not see in your explanation where the concept of "duration" was explained. Thank you, jk.

i don't think the duration of your example is 10years. yes, in the first 10 years you get $100, but you also get $100 in the second decade and $100 in interest in the 3rd decaded and then the $100 pricinple. so you receive a total of $400 back. you have to calculate the present value of those payments, discounting by the bond's yield to maturity.


there is a good explanation with illustrations at:

http://www.investopedia.com/universi...ancedbond5.asp

another explanation and examples are at:

http://invest-faq.com/articles/bonds-duration.html

frankly, i have never calculated a duration in my life. if i had to guess at the duration of your example, i might guess 13-14 years, but i don't have a clue whether that's right, and don't want to bother finding out. but i'm pretty sure it's more than 10.

[Jim Nickerson]

Re: Something’s Always Going Up

there is a good explanation with illustrations at:

http://www.investopedia.com/universi...ancedbond5.asp

another explanation and examples are at:

http://invest-faq.com/articles/bonds-duration.html

frankly, i have never calculated a duration in my life. if i had to guess at the duration of your example, i might guess 13-14 years, but i don't have a clue whether that's right, and don't want to bother finding out. but i'm pretty sure it's more than 10.

"I see," said the blind man. I do see. Gross has often written about "duration," and I have just blown over it without understanding what he was really writing. Thanks for pointing me.

[Finster]

Re: Something’s Always Going Up

"I see," said the blind man. I do see. Gross has often written about "duration," and I have just blown over it without understanding what he was really writing. Thanks for pointing me.

I think JK is steering you right, Jim. Actually calculating a duration is kind of complicated except in the simple case where the entire bond principal and interest is paid at one time. That's zero coupon bond. For a thirty year zero coupon bond, the maturity and the duration are both thirty years. If there are coupon payments which occur earlier (for example, annually, semiannually, or whatever) over the life of the bond, the 'center of mass' of those cash flows is then moved sooner. In other words, the duration of a bond that pays interim interest is generally less than its maturity.

[WDCRob]

Re: Something’s Always Going Up

Have gone back and read through this old thread again and am curious about opinions on what people consider a minimum allocation to the 4 broad categories described in Finster's opening post? Assuming you're actually interested in the asset allocation model in the first place I mean.

I understand that you want to maintain a position in 'everything' because it's impossible to know the future. But if one expects that one segment of the allocation is overdue for a real beating what would most people consider a floor in terms of the allocation there?

i.e. - where does aggressive end and reckless begin?

[FRED]

Re: Something’s Always Going Up

Have gone back and read through this old thread again and am curious about opinions on what people consider a minimum allocation to the 4 broad categories described in Finster's opening post? Assuming you're actually interested in the asset allocation model in the first place I mean.

I understand that you want to maintain a position in 'everything' because it's impossible to know the future. But if one expects that one segment of the allocation is overdue for a real beating what would most people consider a floor in terms of the allocation there?

i.e. - where does aggressive end and reckless begin?

Recall that from a Next Bubbles theory perspective, stocks in companies that benefit from the next asset price inflation do not tend to correct while everything else is tanking. Think home builder stocks during the 2001 - 2002 bear market in everything else.

[bart]

Re: Something’s Always Going Up

My test would be whatever allows you to sleep well and avoid significant worry.

Another way to look at it is to focus in on sectors. If one is only in US stocks at 50%, that's hugely different than a balance in emerging markets, ag stocks, infrastructure plays and "basic" foreign and commodity stocks.

[c1ue]

Re: Something’s Always Going Up

Rob,

The choice is dependent on what you hope to gain balanced vs. what you can withstand losing.

There hasn't been much talk on iTulip on this subject, but I'd look around on the Internet for discussions on risk adjusted gain.

To summarize: actual gain is easily understood - how a given asset has changed vs. some stable metric.

Risk adjusted gain then adjusts the actual gain vs. the amount of risk taken.

The concept is similar to what "Black Swan" talks about - given sufficient numbers of managers and irrelevant to method, some subset of managers will have done well. The question is whether one or more of the methods used are superior or whether the timing was right/lucky.

Therefore understanding how much risk you undertake in a given investment is just as important as how much you actually gain.

In other words, buying 10 Y2K internet stocks and gaining 15% is not the same as buying a Treasury bill in 1987 with a 14.5% interest rate.

Ultimately you'll have to experiment to understand what amount of risk you can withstand; at least for me it was the case.

The relevant risk adjusted gain for each sector, plus your desired gain, then gives you an idea of allocation.

For example if your goal is higher gain than the 'safe' sectors, then you'll have to allocate more into the 'unsafe' sectors.

[WDCRob]

Re: Something’s Always Going Up

Thanks for the thoughtful responses all. Much appreciated.

Oddly (for me at least) my appetite for risk seems to be very high right now and I'm trying to get a handle on what that should mean for the next few years.